Radio Boulevard
Western Historic Radio Museum


Navy Department - Bureau of Ships

Model ATD
 
Aircraft Radio Transmitting Equipment

Contractor:  Bendix Radio Division of Bendix Aviation Corporation

Manufactured by: Sparks-Withington Co. (Sparton)  -  SN: 2104

ATD History and Design Basics, Mechanical Construction Details,
Electronic Requirements, Using an ART-13 AC Power Supply to Power the ATD
Getting the ATD to work with a 50Z Antenna Load, Series Antenna Capacitors
Operating the ATD "On the Air," The Missing RF Amp Meter Solution
Dynamotor Details, An "Easy-to-Build" 80 Amp Power Supply To Run the Dynamotor
More ATD Parts and ATD Accessories Found, Boosting the ATD Output using a Linear Amp
Building New Cables for Dynamotor, Pilot's Remote and Remote Indicator
Cannon Part Numbers for ALL of the Connectors used with the ATD

 

by: Henry Rogers WA7YBS


Navy Department-Bendix Aviation Corp.  Model ATD Aircraft Transmitter
  Type CQW-52253   SN: 2104   Mfg by: Sparks-Withington Co.

 Now equipped with four functioning 3000kc to 9050kc Tuning Modules for 80M and 40M operation. Note the HV doorknob capacitors providing 37pf of series C to allow proper 50Z antenna matching. A Shure Bros 102C carbon microphone and Telephonics CTE26003A flame-proof hand key. Though it probably isn't that unbelievable, the ATD has 50 switches and adjustments, 40 adjustments (16 interior adjustments and 24 exterior adjustments) just in the four Tuning Modules and 10 more on the exterior and interior of the main frame.      2025 photo

This write-up is a little bit different. All of the repairing, rebuilding and restoration tasks are presented in a "day-by-day, problem-by-problem" journal format that also has included the solutions to the multitude of minor issues encountered. Of course, the ability to easily apply the voltages required at the current levels necessary for proper operation is one of the most difficult aspects of setting up a vintage aircraft transmitter to function effectively in a home environment ham shack. Though a homebrew ART-13 AC power supply can operate the ATD, I've found that it's also not too difficult to set-up and use the ATD dynamotor to provide the proper voltages at the necessary current levels. However, the dynamotor's 50A+ "instantaneous starting current" can present problems. Fortunately, I found an easy solution and plans are included for an "easy-to-build" 80 AMP +28vdc power supply that can provide all the current necessary for dynamotor operation. Antenna matching solutions are easy to accomplish. There's also the fairly low 40W RF output power that makes reception of the ATD's signal a challenge for many of the mil-rad net participants. How to boost the power output using a linear amplifier is also covered. Building proper cables can be a challenge and finding connectors can also be a problem if you don't have the correct part numbers. Building shielded cables (necessary) is covered in detail and Cannon part numbers for ALL of the connectors are provided.


Navy Department - Bendix Radio Div. of Bendix Aviation Corp. - Model ATD Aircraft Transmitter
 


RCA Mfg. Co., Inc. - ATB Transmitter

ATD History - The story goes that, around 1940, the Navy was interested in purchasing a newly designed, modern aircraft transmitter, mainly for their two-seater aircraft. Supposedly a competition ensued between some of the manufacturers vying for a Navy contract. Eventually, three different types of aircraft transmitters were accepted and produced for the Navy (some competition,...since all three transmitters were given contracts.) RCA Manufacturing Co., Inc. provided the very small ATB transmitter. With only two selectable channels and only about 3.5 watts output, RCA had decided that their transmitter would be used in a two-seater for plane-to-plane communications. Bendix Aviation Corp. provided the ATD, a slightly more powerful transmitter with motor-drive selecting of four frequency channels that were pre-flight tuned. However, the ATD was still basically for the two-seater aircraft but the 40-50 watts output could provide air-to-ground communication.  >>>


Collins Radio Co. -  ATC Transmitter

>>>  Collins Radio Company provided the ATC, a versatile and powerful transmitter that could be used in almost any installation. At 100 watts output with ten HF and one LF selectable frequency channels that motor-drive auto-tuned the transmitter to preset antenna matching, it really wasn't much of a surprise that the Navy favored the ATC transmitter. It became the famous ART-13 and well-over 100,000 were built during the war and thousands were used post-WWII by various airlines.

But, what about the Bendix ATD and the RCA ATB? The ATB had multiple contracts issued during WWII but post-war it didn't provide a usable transmitter for commercial needs or for hams. Most were scavenged for parts since a 3.5 watt transmitter was virtually useless for anything else. There was only one ATD contract issued (dated May 20, 1940) but it seems that several orders were built on that one contract during WWII. Neither the ATD nor the ATB are encountered very often nowadays when compared to the very common, very powerful, excellent performing and multi-usable ATC/ART-13.

ATD Transmitters Post-WWII and Today - The total number of ATD transmitters built was probably much less than 5000 (serial numbers go up to just over 4000 or so.) Many were never used and were still essentially new in their crates (packed with all their accessories) after WWII ended. Many of these "new in the crate" ATDs were sold surplus after the war. However, the ATD transmitter didn't even start out as an up-to-date modern design so it wasn't suitable for use by post-WWII airlines like the ART-13 was. The ATD had a multitude of tiny-sized knobs and several "hidden" adjustments that made "in-flight" retuning impossible.   >>>

ATD Components Shown in Artwork

A - ATD Transmitter

A - 200kc -500kc Tuning Module

B - LF Antenna Tuner

C - Dynamotor

D - Pilot's Channel Indicator

E - Pilot's Remote Control

F - Dynamotor Battery Cable

G - Transmitter Power Cable

H - Remote Cable

I - Indicator Cable


B&W Artwork from the ATD Manual

>>>   The ATD wasn't practical for ham radio either. It really didn't provide much RF power and several Tuning Modules were necessary to cover the 160M, 80M, 40M and 20M ham bands. Working on an ATD was difficult due to its multitude of relays and other "moving" contacts that were always having "contact" problems. The dense packaging of the main frame made maintaining an ATD as an operational transmitter a real challenge.

Today, there seem to be three types of ATD survivors,...the "found in the crate" mint condition transmitters that have been kept in NOS condition and have never worked on. These types of transmitters, despite their pristine physical condition, rarely function "out of the box." At the other end of the condition-spectrum are the non-functional, hacked-up "junkers" that were the unfortunate victims of the readers (and believers) of the Surplus Conversion Manuals put out by CQ magazine in the 1950s.

But, many ATD transmitters, while maybe not mint, are still in pretty good shape. These "less-than-perfect" transmitters haven't been hamster-hacked and they haven't been converted into something else that doesn't function. The following sections of this write-up are a detailed account of what to expect from a maybe not perfect,...but a pretty nice ATD. One that obviously wouldn't work "as-found" but looked like it might be coaxed into becoming a functional transmitter capable of producing a good quality signal for communications with the various "military radio" ham nets. The obvious problems are that only one of the tuning units matches the main frame serial number (so three tuning units are replacements for some reason,) the RF Amp meter is gone and all of the tubes except the 12SL7GT are gone. All of these clues are indications of a non-working unit that supplied a RF amp meter and some tubes for some other projects but was kept in relatively good condition,...well, except for the usual moving and storage cosmetic problems.

So, let's see what it takes to get this relic of early WWII actually producing some RF,...

 

ATD Basic Design and Construction

The ATD is an aircraft transmitter that was designed to be powered by the aircraft's battery-charger system running the ATD dynamotor. The transmitter's output signal was fed to a relatively short antenna that usually ran from a post in front of the cockpit to the vertical stabilizer (usually a 25 ft "T" antenna.) The transmitter was setup to operate "remotely" by the pilot of a two-seater airplane. The two-seater airplanes at this time had a front-seated Pilot and then the second person was seated about eight feet behind the pilot at the rear of the cockpit. The second person (called the "Operator") in the aircraft could operate the transmitter but transmitting channels were selected by the Pilot.

If the Pilot was incapacitated for some reason, then Operator could cut the safety wire on the LOCAL-REMOTE switch to allow placing the transmitter in LOCAL. Then the Operator could control the transmitter directly, but normal operation of the ATD was by the Pilot through the Pilot's Remote Control Unit Type CRR-23280 that was mounted to the right side of the Pilot's seat. The Remote had a channel selector knob, a large flat button-type of hand key for CW, a MIC input jack, a three-position PHONE/STANDBY-OFF-CODE toggle switch (turned the ATD "on") and a TRANSMIT indicator (this was a visual "pop-up" indicator so the Pilot would know that the ATD was powered since he didn't have direct visibility of the transmitter.)

In addition to the Pilot's Remote Control Unit there was also a Remote Channel Indicator Unit Type CRR-55079 that was mounted in an easily visible area in the rear section of the cockpit. Depending on the aircraft installation, sometimes the ATD transmitter was mounted under the Pilot's seat facing aft (towards the Operator.) Depending on the Operator's seated position, sometimes the ATD panel wasn't totally visible. The Remote Channel Indicator allowed the Operator to see what channel the Pilot had selected. Although the Operator could use the transmitter, it was controlled by the Pilot as far as applying power, selecting mode of transmission and selecting the transmitting channel.

The multitude of adjustments that were necessary for each of the four channels to operate properly made "pre-flight radio setup" a prerequisite. Since any significant frequency adjustment of any channel required removing that channel's individual Tuning Unit from the main frame for resetting four internal switches followed by retuning the transmitter, any readjustment of control settings during the flight wasn't practical and essentially couldn't be accomplished. The manual implies that neither the Pilot nor the Operator were necessarily "radiomen" but they could "operate" the transmitter within the limitations of the pre-flight setup. The fact that the REMOTE-LOCAL switch on the ATD was "safety-wired" to REMOTE implies the Operator in the airplane wasn't a trained radioman.

The frequency selection of the four channels was motor-driven based on the position of the channel switch on either the transmitter or its remote. A chain-driven rotating "letter wheel" (on the ATD) would display which channel had been selected. The selection of a channel's tuning module was accomplished by the motor driving a gearbox that slowly rotated a shaft that had four cams. The four cams with "roller tips" on the shaft were spaced at 90º increments and would individually would push down a specific rectangular push-rod for each channel as the shaft slowly rotated. The push-rod had several metal pins (~.25" diameter) that would push down the movable selector contacts on the back of the selected Tuning Unit to connect the various frequency controlling circuits and the antenna matching circuits for the channel selected. Unselected channel contacts were grounded by way of the spring-loading of the contact arms contacting metal grounding pins. The motor-drive used a forward and a reverse winding that was timed to use just a momentary "reverse" connection to stop the motor after the forward winding was disconnected. This allowed precise position stopping at the selected channel with no "over-shoot." There was a manual channel select knob on the right side of the transmitter. If the handle of this knob was folded out, that disengaged the motor drive gearbox and the main shaft could then be rotated using this knob. When finished with the manual adjustment, the knob was folded back, "rocked" slightly and then the manual control being "spring-loaded" would "drop into position" to engage the motor drive gearbox and to be in the correct position displaying the correct channel.    >>>

>>>   The ATD was supplied with five tuning modules that covered various frequency ranges. The LF/MW modules allowed the ATD to operate 200kc to 540kc and 540kc-1500kc. There was another MW module that tuned 1500kc to 3000kc. HF was tuned using the 3000kc to 9050kc module and the 9050kc to 15,800kc module. All ATD tuning adjustments are inside and on the panel of each of the modules specifically for the channel selected and placement of the module in the transmitter main frame.

The ATD uses a 6L6G tube as the Master Oscillator. The Audio Driver and MCW/Sidetone oscillator functions are both accomplished by another 6L6G. The Power Amplifier tube is an 814/VT154 with +1000vdc on the plate. The two modulator tubes are a pair of 6L6G tubes provided push-pull plate modulation of the 814 PA. A VT-150/0D3 tube is used for a +150vdc voltage regulator for the MO screen voltage. A 12SL7GT is hidden away behind a shield in the right side of the transmitter. The 12SL7GT is the speech amplifier and it uses one triode if a dynamic microphone is used (250Z) or it uses both triodes if a carbon mike is used.

There are a multitude of relays in the ATD. The sending relay switches the antenna from receiver to transmitter with actuation of the sending key or PTT. The Phone relay selects many of the functions necessary for voice or MCW operation. CW selection places the +1000vdc on the 814 plate and +380vdc on the plates and screens of the other tubes but allows the telegraph key to work the keying line (PTT.) There's another relay that places a capacitor on the PA grid to keep it from oscillating when the TUNE mode is selected (it's switched out with PTT or keying.) There's a connecting strap between the TUNE and OPERATE switches so they will function together. The TUNE switch is isolated since it connects the +1000vdc when OPERATE is selected.

The Grid Current Meter measures the MO plate output by its connection to the PA grid. This meter has a switch that allows it to also read battery voltage (the aircraft battery buss voltage.) The PA plate meter reads the plate current (cathode current) on the 814 tube. The RF Amps (Antenna Current) measures the actual RF current to the antenna (fs=5A.) Power output is normally 50 watts in CW and 40 watts in Phone.

There are three data plates on top of the ATD and each tuning module also has its own data plate. The data plate shown to the upper left is for "MODEL ATD" and lists the various accessories. The data plate to the right is for the CQW-52253 transmitter itself. The letter prefix in the "TYPE" identification denotes the builder, in this case, Sparks-Withington Co., whose trade name was "Sparton" (the manual indicates CRR for the letter prefix indicating Bendix Radio as the builder and contractor.) Note that the Acceptance Tag, also shown to the right, is unstamped indicating this ATD was probably still in its crate at the end of WWII. There's a metal Overall Calibration Chart on top of the ATD that has information on both sides. It stows in a holder with two securing clips. The information on the chart provides approximate settings for all of the tuning modules at various frequencies.

The dynamotor provides +380vdc at 400mA and +1000vdc at 150mA. Input voltage is +28vdc and the current required depends upon the load but could range up to 19 amps in some modes and to certain types of antennas (this is the "up and running" current but the actual "instantaneous starting current" is over 50 amps.) The top cover that allow access to the tubes has an interlock switch that only allows the +28vdc to go to the dynamotor when the ATD lid is closed. The dynamotor also has an interlock on its top cover. The Cannon box connectors on the left side of the transmitter are for the dynamotor, the remote and for the channel indicator. The remote was operated by the pilot and had inputs for a mike and a key along with the ability to select channels. No meters or other indicators were on the remote.

Mechanical Construction - The ATD is build around an aluminum framework that is spot-welded together to create a three section open box with a right side, a center section and a left side. The center section has a tube shelf at the rear, a bottom and an upper support for the motor drive shaft and the channel switching push rod assembly. Aluminum panels are screw-mounted to the two sides, the top and the bottom and these panels are painted black wrinkle finish (except for the bottom cover.) The front panel on the left is smooth flat black finish and contains the RF Amp meter and input jacks along with the Tune/Operate switch. The right panel is also smooth flat black finish and has the PA current meter and the PA Grid current meter that also measures Battery voltage along with the channel selector and channel indicator wheel. Mode selector switches and the OFF switch are also on the right side. Also on the right side but inside the main frame are the Sidetone volume control and the Microphone type selector switch. The rear part of the top cover is hinged and lifts up to allow access to the tubes. However, the 12SL7GT tube is located behind the right side panel and its only easy access requires removing a shield that is located on the inside of the center section (right side) with the Channel D module removed. The channel tuning modules contain all of the coils and capacitors necessary for the MO and the PA tuning. A lot of fiberboard (Garolite) construction is used and even the switch shafts are actually made out of fiberboard. The mounting rails have integral rubber shock mounts. Total weight of the ATD is 70 pounds. The external dimensions are 23"W x 11"H x 13.75"D. Electronic Requirements - The ATD uses (1) 814, (4) 6L6G, (1) 0D3 and (1) 12SL7GT tubes. The 814 plate requires +1000vdc. This is the only place that the +1000vdc is used. The two modulator tubes, 6L6G, actually have +380vdc on their plates. The maximum rating for the plate to cathode voltage on a 6L6G is around +380vdc so that was the limitation of using 6L6G tubes as the modulators. The +380vdc is also used on all four of the 6L6G screens and plates (with various divider or load resistors) and also on the 814 screen. The 0D3/VT-150 provides +150vdc for just the MO screen voltage. +28vdc operates the tube filaments and the relays.

The dynamotor runs on the +28vdc aircraft battery+charger buss. The +28vdc battery buss is also routed directly to the transmitter through the dynamotor and ATD power cable. When the ATD is powered up, it routes the +28vdc back to the dynamotor relay via the PTT line to power up the dynamotor to supply the +380vdc and the +1000vdc. CW operation will have the dynamotor running while the transmitter is keyed. In Phone, the PTT turns on the dynamotor for operation.

AC Power Supply Operation - To run the ATD will require +1000vdc at about 150mA, +380vdc at about 400mA and +28vdc at maximum of about 20 amps for the dynamotor (this doesn't include the instantaneous starting current,) but only about 5 amps are required if the +28vdc just powers the transmitter. It's convenient that these voltages are very close to those required by the ART-13 transmitter. This makes operating the ATD using an ART-13 AC power supply a real possibility. The current requirement of 400mA on the +380vdc line might be a problem in some ART-13 power supplies (it depends on the design.) My "Homage a le Valve" ART-13 AC PS provides +1050vdc at 250mA, +420vdc at 250mA and +28vdc at 10A. I'll have to do some slight modifications to the +400vdc supply but whether the ATD actually requires almost half an amp on the +380vdc line will have to be determined when I actually get the transmitter operational. If the +380vdc actually does require 400mA then a separate +380vdc with higher current availability would be required. I do have the HP 713B that could supply +380vdc at 500mA, so that's a possibility,...if it becomes necessary.

 

ATD Aircraft Transmitter SN: 2104 - Problems and Solutions

This is more-or-less a chronological listing of what it took to get the ATD to actually produce a RF signal into an antenna. Dates are at the end of each paragraph.

Missing Antenna Current Meter - This is a 5 amp FS meter that has an internal shunt that allows connecting the RF directly to and through the meter out to the antenna post. The meter is recessed behind a standard bakelite meter flange, glass and partial housing. About 1" of the housing allows the actual Antenna Current meter to mount behind the empty meter housing and that gives the actual meter the spacing necessary. It's possible that the actual meter mounted with four standoffs but what the meter itself mounted to is not known and there aren't any photos available of an original installation. I assume that fiber board was used. The actual meter's flange has to be cut on the right side to provide proper clearance to "center" the meter mounting and have the meter scale centered in the front meter housing (it doesn't seem likely that the installation would require modifying the meter flange - but it does!) Also, the meter I have is out of an ART-13, so the scale is marked "ANTENNA CURRENT" while the original meters were marked "RF AMPERES."  8-17-24     UPDATE: W7MS, who also owns an operational ATD transmitter, is going to donate a 5A RF current meter out of a "parts set" Bendix TA-12. This meter has the proper spacer for recessed mounting but uses only the three meter flange screws. The scale will have to be changed since it's a white scale. The ART-13 meter scale should fit. According to Mike-MS, the four outer perimeter holes were for mounting a metal cup-like shield that covers the meter (actually only three screws were used since the top-left screw threads into the main frame.) I'll pick up the Bendix meter at the SNARS Fall Cabela's Swap Meet a few weeks away.  NOTE: The manual doesn't provide any photographs or even drawings of how the Antenna Current meter mounted. The manual has absolutely no written information on the Antenna Current meter either. Only inspecting another ATD that has its complete Antenna Current meter would provide the information BUT that requires someone willing to remove the numerous screws to take the top off and photograph the meter and then put the top back on and thread in and tighten all of the screws.

The Plate Current Meter had the glass pushed in. This required complete removal of the meter to fix this problem. It's more difficult than it sounds because the ATD right-side panel can't easily be dismounted. Dealing with the nuts and washers that mount the meter requires some special tools. Once the meter was out, the housing and the glass were cleaned. The meter zero mechanism needed to be reassembled and then glued. The retaining ring for the glass had to be installed by pushing it evenly down to be against the backside of the glass. Then Cyno-A glue was sparingly applied to keep the ring in place. The meter was reinstalled. The one nut and washer on the lower inside has to have the washer super-glued to the nut and then, using a special tool that holds the nut/washer, it can be held in place and threaded by turning the screw in the front. The other two nuts are fairly easy to access and remount.  8-17-24

Thumb Screw for Ch.B Osc Adjustment Lock was missing. I found it inside the transmitter during disassembly. Luckily, it just screwed back in place and works correctly.

Missing and Broken Fasteners - Several screws, lock washers and nuts missing. One broken screw on the Antenna-Sending relay board (K-103.) Enough of the threaded portion was left to allow gripping with flush type side cutters to get the piece unthreaded and removed so a new 4-40 BH screw could be installed.  NOTE: I didn't notice until I doing the final assembly and mounting the rear panel that the lower-left corner 6-32 BH screw had been broken off and the remains were deep in the threaded hole. I may fix this later. I don't want to do any drilling, screw extraction and re-tapping with all of the tubes installed.

Disassembled down to Main Frame - Cleaned spider webs and dead insects out of the left side area. Right side was not as bad. Lots of spider webs though. Some minor corrosion. Cleaned panels with Glass Plus. Minor straightening, especially on corners. Checked the 2 watt resistors that are underneath (after dismounting the bottom cover) and all checked within 20% (even though marked 10%.) Every cavity of the main frame has insect debris but it brushes out easily with a 1" wide soft paint brush or 1/2" paint brush for smaller areas. Cleaned every part of the main frame, first with WD-40 and then with Glass Plus. 8-17-24

Online Manual - I downloaded the only ATD manual that I could find online. This online manual isn't complete. It has most of the text but not all of it. The multitude of mechanical drawings that are in an original "hardcopy" manual are NOT included in the online manual. The important complete "parts list" isn't included (a partial transmitter electronic components list is included.) The installation set-up and operating instructions along with the overall schematic are all that's included in the online manual. It's enough information to get an ATD up and running but, as I was to find out, the complete parts list is important and that isn't provided.

Unobtainium Power Connector, but,...it's only unobtainium if you use an incorrect Part Number - The box connector is a Cannon NK-M8-32S, so I assumed the mating connector plug would be Cannon NK-M8-32P but I can't find any source that has "32" size connectors. I assumed that only a "P" substituted for the "S" would indicate a "plug" but that's not the case. I was using an incorrect part number and that's why I never could find the correct plug. The actual part number is Cannon NK-M8-23C 11/16" - I found correct part number when I was able to look at the "complete" parts list in an original "hardcopy" manual.

Temporary Solution for the "No Connector" Problem - Until a proper connector can be located, a temporary method to apply power is to use "push-on" female socket connectors. The pins are .125" diameter and that's the same size as the tube socket receptacle pins on a five pin tube base. A fiber-type five pin tube socket has receptacles that almost surround the tube pins. These can be removed and used to plug onto the power connector pins. Then proper size wires can be used for a cable. The pins will have to be insulated. When building the cable, it will have to be sleeved with copper braid for shielding. This cable should be at least six feet long so the PS (or dynamotor) can set on the floor if it needs to. Hopefully, this will be just a temporary solution and the correct plugs can be found. 
 

CONNECTOR UPDATE: In August 2025, I found a complete, original manual for the ATD. This manual had the proper identification for ALL of the connectors. I was completely surprised that, when using the correct part number, the power connectors were found immediately on eBay. I had to use two different sellers but they each had one NOS Cannon NK-M8-23C 11/16" (right-angle) connectors at reasonable prices ($5 for one, $20 for the other.) These are both right-angle connectors where originally only the Dynamotor side was a right-angle and the Transmitter side was straight. However, right-angles on both sides will work fine (the position of the tubular cable exit can be rotated, if necessary.) The part number for "straight" connector is NK-M8-21C 11/16"


View with Rear Cover Removed showing the Tubes


Homage a le Valve - ART-13 AC PS uses 866A and 5U4GB Rectifiers

Using the ART-13 AC PS to Operate the ATD - To operate the ATD using the "Homage a le Valve" ART-13 AC PS (without permanently modifying the PS's ability to also power an ART-13) will require using a special "ATD to ART-13 PS" cable that has to be built up. In the cable, ATD pin 1 (+28vdc) will be connected to ART/PS pin 3 (+HV/+LV relay) and ATD pin 4 to ART/PS pin 8 (chassis connection with PTT.) This allows the ATD's PTT or CW selection to turn ON the +HV and +LV transformer primaries providing these voltages in a manner similar to the dynamotor operation. In the cable, ATD pin 8 will be connected to ART/PS Pin 10 and will carry the +1000vdc. Additionally, -HV on ART/PS pin 2 should have a jumper to pins 5 or 9 to provide chassis-ground for the +1000vdc. ATD pin 2 will be connected to ART/PS pins 5 and 9 and will be the chassis-ground connection that will also ground the shield of the cable. ATD pin 5 will be connected to ART/PS pin 1 and will carry the +380vdc. ATD pin 3 will be connected to ART/PS pins 6 and 4 and will carry the +28vdc. All of these "cross connections" can be accomplished in the special-build power cable used between the ART/PS and the ATD. The cable will have soldered-on ring lugs on the ART/PS side and will have push-on pin female sockets for the ATD side. The wires for chassis-gnd and for +28vdc will be 14 gauge and the other four wires will be 16 gauge. Also, to eliminate the need to close ART/PS pins 7 and 5 to turn ON the +28vdc at pin 4 and pin 6, a jumper from pin 7 to pin 5 on the ART/PS rear terminal strip can be installed. This would result in the PS +28vdc relay being "on" during PS operation but that's the way it operates when powering an ART-13, so the jumper from pin 7 to pin 5 is an easy to install solution. The "jumper mod" is easily be removed when the ART/PS is used to power an ART-13. Of course, to power an ART-13, the ATD cable would be removed and the ART-13 cable installed.  NOTE: Although I did operate the ATD "on the air" powered by the ART-13 AC Power Supply, I was not entirely satisfied with that set-up. Ultimately, I decided to concentrate on using the original Dynamotor as the +HV and +LV voltage source.
Logging Chart for Ch.B missing -  The other three are present so I can copy one of those onto a manila folder paper for the correct color. I probably will have to glue to something for the replica to be the proper thickness. These tags are mounted with two 4-40 screws and nuts - requires that the module be "out of the transmitter."

Tested the Two Audio Transformers - I used the schematic to determine where to access the primaries and secondaries for each transformer. Both the audio driver transformer and the modulation transformer tested okay. 8-18-24

Cleaned All of the Relay Contacts and the Contacts in the Main Frame for the Tuning Units - I used DeOxit applied with a saturated piece of paper that was drawn through the relay contacts to clean them. The stationary Tuning Unit contacts were cleaned with DeOxit applied with a small paint brush and then wiped with a paper towel (later these Tuning Unit points were cleaned more aggressively.8-18-24

Cleaned and Inspected Channels A through D Tuning Units - Mildly dirty, a few broken or missing screws. Slightly bent sheet metal on Ch. D air variable-C but it seems to have been that way since the original assembly. There's a flex-coupler on the shafts that compensates for the misalignment. Originally, the ATD came with a 540kc-1500kc module, a 1500kc-3000kc module and two 3000kc-9050kc modules installed. Mine has two 540kc-1500kc modules and is missing the 1500kc-3000kc module. One of the 540kc-1500kc Tuning Units has the only serial number that matches the main frame serial number. UPDATE: Jun 3, 2025 - I've removed the two "useless" 540kc to 1500kc Tuning Units and replaced them with two more 3000kc to 9050kc Tuning Units.

Channel C Tuning Unit - Broken screw head on rear bottom insulator panel (6-32 BH.) Removed the broken screw by removing the rear nut and washer so it could be pushed out of the hole. Installed new 6-32BH machine screw and installed the washer and nut.


ATD with the Bottom Cover Removed


The Motor Drive Shaft and Cams with Roller Tips for Changing Channels

Low Voltage Test - Applying +27vdc to the ATD - I used an old Meanwell +24vdc 13A power supply with 14 gauge wires going to the Cannon NK-M8-32S connector. Pin 2 is chassis ground and pin 3 is +28vdc in. I used high-quality Allpoints 85-1075 push-on connectors that were soldered to the two wires and then pushed onto the proper pins of the box connector. I installed all of the tubes, (1) 814, (4) 6L6G, (1) 12SL7GT and (1) VT-150. I switched OFF to PHONE-STANDBY position and all of the tubes illuminated (except the VT-150, of course) and the BATT position for the GRID Meter showed +27vdc when actuated. I confirmed that the channel selection worked fine in manual. I then tried the motor drive channel selector. It rotated very slowly and the forward-reverse contacts (S-101) started to smoke. The load of trying to rotate the gear work was too much. I sprayed the gears and bearings with WD-40 (a quick expedient for penetrating oil.) I also cleaned the contacts on the forward-reverse switch. Now the motor drive ran quite fast with just a very slight, normal load. But, although the motor would stop on the correct selected channel, it wouldn't reverse direction if a "lower" letter was selected, it would just continue to rotate forward and then stop on the correct channel selected. There's a cam that should actuate the forward-reverse switch that must be in the right position. Also, the letter-wheel is slightly off, although it is within the window of the front panel, it isn't in the center of the window but the cams for the channel select are exactly in the correct position. This might just be a simple adjustment of the letter-wheel position. I also checked the operation of the keying relay by operating the KEY TEST switch and the K-103 relay operates correctly. 8-19-24

Motor Drive Channel Select - Read the Manual! - I've re-read the manual several times and no where does it say that the motor runs in reverse if a "lower" letter is selected. I checked the sequence of the forward and reverse contacts and it looks like it does exactly what's described in the manual. The motor-drive is working great and stops perfectly on whatever selected channel is set. I think the "reverse" operation on "lower" letters (an assumption on my part) isn't really what the "reverse" is for. I now think that the "reverse" motor winding actually seems to be for "stopping" the motor exactly in the correct position for the channel selected with no "over-shoot."   8-20-24

Restoring the Antenna-Gnd Mount - This piece bolts to the left side of the transmitter and holds the insulator that has the Antenna input terminal and the Receiver terminal. All three connectors are RAJAH types, like spark plug tips. The Ground terminal is connected to the mount and to the side panel for the chassis connection. The paint on this metal mount is very, very thick, like Japan enamel. On this ATD, the mount's paint was chipped and some paint was missing. There also was some corrosion on the metal mount. The engraved nomenclature looked like it was done after the paint job since the engraving into the metal is fairly shallow. To repaint, I had to remove the old paint. This proved somewhat difficult. Methylene-chloride stripper would barely do anything. I chipped and scraped and then applied more stripper and left that on the mount for one hour. This time I used a metal bristle brush to remove the softened paint. This worked fairly well but I still had to sand the remaining paint off of the mount. I went over the engraving with a diamond tip graver to get it a bit deeper. It's just the way the original was made however, that is, the engraving done after the paint job resulted in deep debossing that was filled with white paint. However, the depth was almost all paint! Now that's gone. If the new paint job fills the engraving, I'll just leave it that way. The connections are pretty obvious anyway.  8-20-24 - I painted the mount with several coats of Krylon Gloss Black. This really didn't fill up the engraved nomenclature so I used various graving tools to "dig out" the paint. This left the nomenclature silver and easily readable. In two places, the "D" in GND and the "AN" in ANTENNA, were corroded to the point that the engraving wasn't present at all, so these two areas were engraved by approximation of what should have been there. Looks fine from a distance and acceptable close-up. 8-21-24

Cleaned Antenna-side panels, remounted and resoldered Antenna buss wire. Until I get the RF Amp meter from Mike 'MS, I'm just going to connect the RF output directly to K-103, the sending/antenna relay. That way I can proceed with testing the ATD. The RAJAH bakelite spark plug terminal push-on connectors arrived today. These will be used to connect the antenna and ground to the ATD.  8-22-24


Side showing Antenna-Receiver-Gnd Mount


ATD Dynamotor - TYPE DMDFX618 sn: 260453
Mfg. by: Continental Electric Co., Inc.

Making the Power Cable - Chassis-Ground and +28vdc are 14 gauge wires (+28vdc is Red and Gnd is Blue.) PTT is +28vdc Red 16 gauge and switched to ground is Brown 16 gauge. +380vdc is 16 gauge Orange. +1000vdc is 16 gauge Blue. The cable is approximately six feet long. The wires were slightly twisted together and then wrapped with black electrician's tape. Then a braided copper shield was harvested from RG-8U coaxial cable and sleeved over the ATD cable. A small copper braid was soldered to each end for a ground connection on the PS and on the ATD. Then the cable was again wrapped with black electrician's tape. Ring lug terminals were soldered to the PS side of the cable. The ATD side has Allpoint push-on connectors for +28vdc and Ground. I harvested the receptacle pins from a fiber-type five-pin tube socket. These can be straightened and then function quite well as push-on connectors. These were installed on the four 16 gauge wires. All of the push-on connectors were sleeved in fairly thick rubber tubing for insulation. The color of the wires is used for ID purposes. The two blues and the two reds are different shades and different gauge wire so there's no confusion.  8-23-24

ATD Dynamotor Arrives - I found an excellent condition (NOS, actually) ATD dynamotor for $70. It arrived today. I don't plan on running the ATD on the dynamotor just yet. If and when I move the ATD out to the shop, that's where I have the large PP-1104-C power supplies (three of them) for running high-current devices. I'll have to either modify the power cable I just made or build another. But first, the ATD needs to be running correctly on the ART/PS and operated "on the air" from upstairs. Before I actually used the dynamotor I would have to "service" it. After all, the grease in the bearings is over 80 years old.   8-23-24   Well, that plan changed,...read Dynamotor section further down.

More ART/PS Changes - I checked the schematic and updates that I had made to this ART/PS a few years ago and found that I had changed the +400vdc to a pi-network filter to up the voltage and it was then running about +420vdc, too much for the ATD. I decided to put the +400vdc back to a similar configuration that I had before but without the input capacitor, in other words, a choke-input with dual filter chokes. This should produce about +380vdc. The choke input filtering gives better stability and at least a 10% reduction in output voltage. Test to confirm.  I also brought in from the shop a roll-around small furniture dolly in order to have any easy way to move this ART/PS that weighs about 80 pounds. 8-24-24

+380vdc Test - I put 5000Ω load on the +LV output and the resulting voltage was +380vdc. That's not a "full load" but I can't really find what the current draw on the +380vdc is. The dynamotor data plate indicates the +380vdc is rated at 400mA but I don't know if that's how much current availability the dynamotor has or if that's the +380vdc current load of the ATD.  8-25-24

+28vdc "ON" Relay - In looking at the ART/PS, the wire from the +28vdc "ON" relay (ART-13 pin 7) that connects to chassis-gnd to power the relay is the next terminal adjacent to the chassis-gnd terminal. Rather than have a second jumper, I just moved the ring lug over to the chassis-gnd terminal to provide the ground return for the +28vdc "ON" relay. Now there's only one jumper and that connects -HV to ground (ART-13 pin 2.)  8-25-24

Set-up and Operational Test - The manual has a step-by-step procedure for testing and set-up. This is probably the easiest way to find any additional faults. I had to move the ART/PS up onto a small table next to the test bench since I really couldn't see the output voltage meters or the indicator lamps with it on the floor. Out of the cabinet it probably only weighs about 65 pounds. 8-26-24

Throttle Switch and PTT Line Problems - The PTT line and the Throttle Switch wouldn't "lock" the sending relay K-103 in the transmit position. This problem was caused by K-101A on the Phone Relay having dirty contacts. I used DeOxit soaked paper pulled through the contacts to clean them. Then the PTT and Throttle Switch both worked correctly.  8-26-24


Side View showing Motor Drive and K101 Phone Relay


540kc-1500kc Tuning Unit showing setup switches inside. Note that four switches inside the unit must be set for proper operation.

Channel Select Contacts - The manual indicates that these can be cleaned with "sandpaper" which sounds a little harsh. These contacts were black with oxidation. Usually "black" indicates silver oxide which is very conductive. I used a brass bristle brush to clean the black oxidation off anyway. The brass brush only removes the oxidation and doesn't harm the plating on the contacts. I cleaned both the modules and the main frame contacts.  8-26-24

More Testing Reveals Another Problem - The +380vdc under load measures +360vdc which is close enough. VT-150 glows purple when transmitter is operating. So far all testing is in the "TUNE" mode which prevents the +1000vdc from going to the 814 plate. I setup one of the 3000kc-9050kc modules for 4000kc operation. I connected a 50Z dummy load to the ANTENNA-GND terminals. There are four switches inside each of the four modules that have to be set correctly and then there are six front panel switches or controls have to be set correctly for the frequency of operation on each module. The ATD was set for CW operation. When the TEST KEY is switched, some PA grid current should show, however, the meter actually went negative. By tuning the MO PLATE, I could just get the meter to read "0." I happened to look at the 814 when I pressed the TEST KEY and the bottom half of the tube turned purple inside. This is a sure sign of a "gassy" tube, so no wonder the PA GRID current wasn't measuring as expected. I thought I had an 814 in my tube stock but I couldn't find it (if I ever had it.) I had ordered two of the cheapest 814s off of eBay. Neither of the tubes had been tested but the sellers thought the tubes were NOS. The first 814 had an open filament (got my money back on that one.) The second tube didn't show much emission in the TV-7 and did give a slight indication of being gassy but I thought it might work in the ATD (I didn't ask for a refund since a couple of weeks had passed before I actually tested the tube in the transmitter.) This second 814 was in the original box with the original wrapping and original paperwork,...naturally the seller thought the tube was NOS so it wasn't tested. But, it's gassy, and the purple glow was only with the +360vdc screen voltage applied (wonder what it would have done with the +1000vdc plate voltage applied?) The only way to proceed is to buy tubes that are tested and guaranteed. These two "cheapies" were a waste of time. I ordered another 814 that the seller provided the TV-7 test data information on. We'll see what happens with this tube when it arrives. Out of curiosity, I'm sure the "cheap-o" prices might be of interest, right? The open fil 814 was $15 and the second gassy 814 was $18. The tested 814 was $39. So, luckily, the 814 isn't an expensive tube but I wasted a lot of time and that's what was aggravating.    8-26-24

Making Replica Logging Charts - To make a replica logging chart I first copied an original logging chart using a laser printer and copied it onto manila folder heavy paper. I then used Titebond glue and added four more manila paper pieces for the correct thickness. This sandwich of glue and paper was compressed until the glue dried. I should have sprayed the laser copy with Matte Clear Krylon, but I didn't. Also, I should have punched the mounting holes in each layer before gluing, but I didn't. Since Krylon wasn't used, the durability of the laser printing isn't very good. Since I didn't pre-punch the holes I tried drilling the holes but that forces cut paper into the sandwich of paper and that spreads the edges. So, the replica wasn't perfect. It still looks just like the originals, at least from a distance of a few feet. It's really very easy to make another one after I buy some Clear Krylon and I'll have plenty of time since I have to wait for the new 814 to be delivered.   8-27-24

The New 814 - The new 814 arrived on Friday 8-30-24. I tested it in the TV-7 and this one actually tested good. I couldn't get the old 814 out of the ATD unless I took the rear cover off. Luckily, I had only put about six screws in since I thought I might be taking it off again. With the new 814 I now had about 15mA of PA grid current that would easily peak with the MO Plate control as described in the test procedure. Next was to go to OPERATE. I had about 80mA of PA current showing but the Plate tuning brought that down to 40mA at resonance. I tried several adjustments but it acted like the antenna wasn't connected but the dummy load RF meter did show a few watts of output. I tried connecting two dummy loads in parallel to have about 25Z load, this didn't help, still about 30mA to 40mA at resonance. I checked the manual procedure and it indicated that without an antenna the PA current would resonate at 30mA, which it did. I then found in the front part of the manual that the expected antenna was a 25 ft "T" antenna and the antenna resistance at 3mc was expected to be just 1.3 ohms with a capacitance of 80pf. The 50Z dummy load (or 25Z for two dummy loads in parallel) must look like there's no antenna connected. But, reading a little further, there's a section on loading "abnormal antennas" that indicates that the series capacitor can be switched out or the taps on the antenna L can be changed to allow matching unexpected antenna impedances.  8-30-24

The ATD has RF Output Now,...Sort of - Taking the manual's indication that abnormal antennas may require adjusting the tuning module differently, I tried a few other setting. The best results were with switch J set to 5 (this shorts out coil a portion of L603) and then place switch K to short out the Antenna Series Capacitor. I then set Control E to position 7 (L606 Antenna Load.) These setting allowed me to adjust the PA Plate current to just over 100mA. Control C PA Tune worked just like it should, that is, dipping the plate current at resonance. Now this was using the Harrison dummy load (47 ohms) which would be presenting a pure resistance antenna with no reactance. Apparently, the 25ft "T" antenna that was "normal" had a much higher C value that required more L in the tuner. Since I use an external matching device that presents a 50Z load with minimal reactance, I should get the same results running to the antenna. 8-31-24

PHONE Operation and CW "On the Air" with Collinear Array on Harmonics - First I tried the Sidetone. I used 600Z phones but I didn't hear anything out of either jack. Next, I switched to PHONE and plugged in a Shure 102C carbon mike. I've used this mike many times and it has a good response. I could just barely tell that the PA was being modulated. I could see the plate current meter move a little. I had the signal (still on the dummy load) tuned in on the bench monitor receiver. I could hear some modulation but it was probably only 30%, barely detectable. The +380vdc is running at +340vdc under load. I didn't notice any change in the voltage as I spoke into the mike. The manual indicates that the +LV can be anywhere from +300vdc to +400vdc and since the dynamotor would vary in speed with the charging rate and engine speed, it would be expected that the dynamotor voltage would vary some. I tried connecting the ATD to the Collinear Array through the Viking Matchbox tuner. The tuner didn't show any change in the very high SWR. I added an auxiliary variable condenser on the Antenna output to ground (like the ART-13 uses) and I ended up switching the tuner to 40M. Now, I could get the SWR down to 1:1 and the PA current at 100mA. That gave me 35 watts to the antenna. I think if the screen voltage on the 814 was up to +380vdc or so, I'd be able to get more PA current and output. My concern though is that I've tuned the output network of the ATD to a harmonic of 3.974mc which would be 7.948mc and that's how I can get the tuner to achieve a 1:1 SWR on 40M but not on 80M.  8-31-24

Electrolytic Caps were Defective - There are two used in the ATD, C102 is the carbon mike bias voltage filter and C126 is the Driver cathode bias filter to ground. Both caps are can-type electrolytics that are 25mfd at 50vdc and are single capacitors per each can. Luckily, only one wire connects to the positive terminal on each capacitor so isolating the old electrolytic is easy. I used 33uf 50vdc radial type electrolytics. The + lead was soldered to the isolated + wire and sleeved with black plastic tubing. The negative lead was soldered to the ground terminal that the cap's negative terminal was connected to. This is the method used for both capacitors. The modern electrolytics are so small they can be tucked-away and hidden from view. Tested the Sidetone on CW and the output to 600Z phones was clean and loud. I'll have to reduce to Sidetone volume down to about 4, it was on 7 which is maximum (and way too loud for phones.) I tested the ATD on PHONE and listened on the Siemens model 311 monitor receiver. I was using the Shure 102-C carbon mike. Modulation was strong and voice characteristics sound clear but like a carbon mike normally sounds.   9-1-24

Experimenting with Various Antenna Coupling Capacitors - Since I was pretty sure that I had the ATD antenna coupling circuit tuned to 7.948mc instead of 3.974mc, I set all of the internal switches back to where they should be (G, H and J on 2 and K on 1.) Then, figuring that I had reduced the L in the module and got the loading to somewhat work (although on a harmonic,) the Collinear Array and Viking Matchbox must look like a lot of inductance until resonated and by adding capacitance in series, that might work. By switching K to 1, that added a series C already but it wasn't enough for a 50Z load. I tried 100pf first and this allowed the ATD to get kind of close to loading correctly but it required too much L that the module didn't have. I added another 100pf in series to have 50pf in series. This worked very well although I had to switch E to position 4 instead of 5. I could get a little over 100mA of PA plate current. I added another 50pf in series for 25pf. The results were about the same except switch E had to be in position 3. This switch E was adding L to the Antenna Loading which seemed to indicate I was close to finding a range of series C antenna coupling that worked. I think about 60pf or so would allow switch E to remain on 5 and I'd be able to get over 100mA of PA Plate current (later I was to discover that 100mA is way too much plate current.) I have to look through my high voltage capacitors to find a combination that will give me 60pf. OR, I have a box of NOS transmitting type air variables that are very small capacitance, probably 10pf unmeshed to 100pf meshed. Using one of those would allow "dialing in" what C works best.   9-1-24

Showing the electrolytic C102 Carbon Mike bias filter
The Sending Relay K103 is shown to the left


37pf of series C for 50Z antenna with RAJAH connectors

ART/PS Mods for the +LV Supply - I noticed that as I get the ATD closer to working correctly, the load on the +380vdc must be going down some. The voltage is now running +350vdc under load. I can load the ATD to about 110mA of Plate current and the "red line" is 120mA, so it's pretty close. I'm pretty sure that if the 814 screen was +380vdc, I'd have the extra 10mA of Plate current. I could easily remove the input choke from the circuit and then move the electrolytic to be in parallel with the output electrolytic. This would have a filter that is a choke input and a single large electrolytic on the output. I've seen these used before but I'll have to test to see if the filtering would be adequate (I doubt it.) The other test would be to just place a shunt across the input choke and that would create a pi-network filter. Voltage stability might be a problem but it would get the voltage up.   9-1-24

More Antenna Capacitor Testing - I tried the 100pf max air variable but the plates were too closely spaced and it arced at resonance. I have many other wide spaced air variables but the fixed high voltage caps are much easier to use. I tried 60pf and that was too much. I ultimately ended up with 37pf with the E switch on 3. I found that maximum output power doesn't happen with maximum PA plate current. At 110mA, I had about 25 watts output but at about 80mA I had 40 watts output. The SWR at the Matchbox is 1:1 so reflected power is zero. I suspect that when I raise the +340vdc (that's what it measures at 40 watts output) up to at least +380vdc, I should be able to have better efficiency in the 814 with higher screen voltage and I should be able to load up to 50 watts output on CW.   9-2-24

The ART/PS +LV Mod that Worked Best - First I tried connecting the +LV filter as a pi-network but the voltage was too high even under load (+420vdc.) Also, I had almost 20mA of PA grid current because of the high plate and screen voltages. Next, I connected the +LV filter as a single choke with output filter capacitor. Now, I was back to +340vdc under load. I listened to the signal on the monitor receiver and there really wasn't any change in the hum level. Then I thought about the capacitor. I had 20uf made up of two series 20uf caps at 10uf and a second series pair in parallel giving 20uf total. I had some 100uf 450vdc electrolytics that two in series would provide 50uf of capacitance. I added 750K balancing resistors to even out the voltage drop on the series electrolytics. With this hook-up, I had a consistent +360vdc under load in CW.  9-2-24

Using a D-104 Mike? - Why not? I set the mic switch to Magnetic (meaning dynamic) which is a 250Z input (same as the ART-13.) The D-104 is on an amplified base so the Z is much lower than a crystal mike and the gain can be adjusted as needed with the base's gain control. No problem in modulating the ATD. It sounded very good in the monitor receiver. While I had access to the Mic switch, I also set the Sidetone volume down to 4.

ATD Reassembly -  I installed the second 3000kc-9050kc module and set it up on CH. C to the same 3.974mc and made sure it loaded and performed the same as CH. D. I then installed the two 550kc to 1500kc modules but since these really aren't usable "on the air" I didn't do any setup on them. They are in channels A and B. I installed all of the panels and all of the screws with lock washers.

More Antenna Capacitor Testing - I got to thinking that I had some doorknob capacitors that might work nicely for antenna caps. I found two 50pf, one 100pf and several others that were higher values that couldn't be used. By using the 100pf in series with one 50pf, I'd have 37pf and that should work very well. I used the RAJAH connectors and made one end of the capacitors able to plug into the ANTENNA terminal on the ATD using a short flex cable made out of copper braid and the other end of the caps were soldered to the coax center conductor of the ATD coaxial cable. The shield of the coaxial cable also had a RAJAH connector for the GND terminal. I then connected this cap-coaxial stub to a 50ft coaxial cable to run into the ham shack to connect to the Johnson Matchbox and the Collinear Array. This hookup allowed me to use the Drake W-4 wattmeter and the modulation monitor oscilloscope. With careful tuning, I found that the best setting was at 80mA of Plate current and that showed 48 watts on the wattmeter with a 1:1 SWR.


The ATD Works on the Bench using the ART-13 Power Supply. Note the "dummy" ANTENNA CURRENT meter and the two "useless" 540kc to 1500kc Tuning Modules in A and B channels.

At this point the ATD is ready for its "On the Air" debut but I have to move it from the workbench to the ham shack. Not just that, but I have to make room for another station setup in the shack. I'm planning to initially use a Navy receiver with the ATD,...not the ARB,...I've tried that receiver on the air before and the lack of selectivity is a huge handicap for successful communications. I'm going to use the Hammarlund RBG-2 Navy receiver. This was a USN version of Hammarlund's HQ-120X that was specifically built for the Navy with heavy-duty construction, a slightly different tube line-up and different band spread calibration. I've used the RBG-2 on the air several times before and its a very capable receiver and very selective. One note though,...when I am given the Bendix RF Ampere meter, I'll be putting the ATD back on the bench for that installation.
 

Operating the ATD "On the Air"

Setup and Testing - I first moved the ART/PS cabinet to the ham shack table. Then I disconnected the ART/PS from the ATD and moved the PS chassis to the ham shack. I then installed the chassis into the cabinet. I pulled the four tuning units out of the ATD and moved the pieces separately. I connected the ATD to the PS and installed all of the tuning units except Channel B since that was where the motor drive was set. I moved the channel to D and installed the B channel tuning unit. I made up some coaxial cable using RG-8X. The 37pf series capacitor was inline as it had been for testing. I also connected the antenna coax from the RGB-2 receiver to the ATD. I was ready to apply power.

The ATD powered up fine although there isn't any visual indication that the transmitter is "ON." The BATT V switch can be pressed and it should read +28vdc when the ATD is "ON." I was on CW so I switched on the high voltage at the PS. Pressing the KEY, I had about 85mA of PA Plate current. I checked the wattmeter and it read 55 watts. I think because the +HV is actually running about +1100vdc instead of +1000vdc (only 10% high) that's the reason that I have to load to a lower PA current. I reduced COUPLING slightly and dipped the Plate, I still had 55 watts. I reduced the ANT TUNING and dipped the Plate, now I had 50 watts at 75mA of Plate current. I went to PHONE and I connected the amplified D-104. Pressing the PTT, I still had 50 watts output. I reduced the COUPLING slightly and the ANT TUNING slightly and the power output dropped to 40 watts. I looked at the monitor scope as I talked into the mike. Lots of modulation but I don't seem to be going all the way down to cut-off. Negative modulation looks like about 80% maybe. Watching the +360vdc, it drops significantly with the voice peaks,...maybe down to +320vdc. Obviously, the current required for the modulator to really achieve cut-off is more than the ART/PS +LV can supply. I may bring in the HP713B power supply that can provide a solid +380vdc at 500mA and test the ATD in PHONE then. I would then have to do a few mods to the +400vdc supply in the ART/PS to get the current available up to 400mA or more.

More Current Availability for the +380vdc Needed - Even though modulation isn't as potent as it could be, I think with the D-104 mike, the ATD will sound fine. After all, it's just the first "on the air" test coming up. The ART/PS upgrade can be another project in a week or two. I have a hefty 400vac CT transformer that must be good for at least 500mA. It measures as a 5.5" cube and has no other windings other than the 400vac CT and the primary. I'd have to look for a 500mA choke but I should be able to find one. I do have room on the ART/PS chassis if I rearrange some of the components. Not too much work - well, certainly not as much as building a complete PS. But, before I do that, I have to bring in the HP713B and verify that 500mA of available current will cure the wobbling +360vdc supply and allow 100% modulation.

Dynamotor Operation - Of course, an easy solution to the power supply current problem is to use the original dynamotor. I have a very small +28vdc 27amp switching PS that could easily(?) spin the dynamotor and not be loaded down with varying current demands (this power supply alone couldn't supply enough current to even start the dynamotor. At least 45 amps are necessary to just get the dynamotor to spin with no load. I tested using parallel-connected switching power supplies and while 40 amps wouldn't spin the dynamotor, 50 amps did spin the dynamotor. At least 55 amps would provide enough reserve for a fast start-up of the dynamotor actuated by the ATD PTT but more current availability would certainly help. More information in the "The ATD Dynamotor" section further down this page.) The entire PS needs to be shielded and the +28vdc cable from the PS to the dynamotor must also be shielded. I'll only go this route if I can't find an AC PS solution that works. One advantage to dynamotor operation is size and weight when compared to the ART/PS.

Arcing in the Unused Modules - I read in the manual about the unused modules causing problems if they are tuned to harmonics of the frequency being used. The cure is to retune them to a different frequency. I never tuned either of the 540kc-1500kc modules. So, I'll start by tuning one to 600kc and the other to 1100kc. Neither are harmonics to each other and neither are harmonics to 3974kc. I think I'll retune CH C to a different frequency than CH D. Maybe I'll set it up for 40M or easier would be 3995kc, an informal mil-rad net.

When looking a little closer at one of the AM-BC tuning modules I noticed that four screws were missing. Two were completely gone and the other two had the screw head "popped off." These appear to be 5-32 screws but a 6-32FH fits fine and these screws, washers and nuts are only for the Garolite mounts for two of the coils. I installed the four 6-32FH screws using lock washers and nuts.  9-4-24

I thoroughly cleaned the contacts and the levers plus shot some DeOxit into the heads of the levers. I used a brass bristle brush and some DeOxit on a paper towel. I also cleaned the switch contacts G,H, J and K. Before installing the modules, I set one for 600kc and the other one for 1100kc. I set CH D for 3995kc and CH C is set to 3974kc. I tested the ATD and now there's no arcing.

I had KØDWC (located about 2 miles away) listen to the transmitted PHONE signal from the ATD. I could also hear the signal over the telephone although there's a propagation delay of about half of a second that sounds like an echo. Other than a little feedback over the phone and the delay, the audio sounded good and Chuck also thought the ATD sounded good. I'll be giving the "ATD debut" on Sunday's (9-8-24) Nevada Mil-Rad Net on 3974kc at 0700hrs    9-5-24

ATD Debut Ends with a Problem - Propagation conditions were very good but attendance to the Nevada Mil-Rad Net was sparse. Still I got good reports from everyone. Although the +380vdc was running at +320vdc, the modulation levels seemed okay and the reports were that audio quality was good and modulation levels more than adequate. I made several transmissions but none were of any great length. Maximum length was probably about 3 minutes. During the "73 round" I noticed the PA Plate current dropped to about 50mA. A quick look at the PA grid current and it had dropped from 13mA down to 5mA. I still had a carrier according to the oscilloscope. I signed quickly as a "curl of smoke" appeared out of the right side vents. This seemed to indicate a burning resistor associated with a leaky or shorted capacitor. I disconnected the ATD and moved it to the workbench. I had pulled all of the tuning units. I checked the schematic to locate where the most likely resistors involved would be. I pulled the bottom cover and saw quickly that R121 had swollen and had cracks on top of the resistor body. This is the 1000 ohm plate load for the master oscillator although actually the voltage (+380vdc) is routed through R121 for all four channels and then the individual tuning unit selected will involve one of four 300 ohm resistors and then there's a bypass capacitor to ground on each of the 300 ohm resistors. These four bypass capacitors and R121 are always connected to +380vdc whether the channel is selected or not. I measured each bypass capacitor and C117 had a short to ground. The other three bypass caps involved all tested okay. I do remember that the arcing I heard did sound like CH A was involved and maybe it was. I wonder how long C117 had been intermittently shorting before it became a "hard short."
Troubleshooting, Rework and Repair - First, R121 is a huge CC resistor. It's the old style IRC 2W CC resistors that are really quite large. I have a few of this size CC but, of course, not a 1K value. I used two 2K 2W CC made by IRC (later manufacture so these 2W are the same size as the old 1W resistors were) in parallel to have 1K at about 4W. It will do until I can find one of these large IRC 2W carbon resistors in a 1K value. Installation,...easy.

Second, accessing and removing C117 is made impossibly difficult due to the shield around the MO tube. With the 814 removed, one end of C117 is accessible but the other end is behind the MO shield. Although the shield might be removable, the mounting screws have washers and nuts that are covered by several MO components under the tube shelf. The value of C117 is .01uf and it's the standard "postage stamp style" cap with the yellowish-brown square body and the color-coded dots. I had several of these in the parts box. The color-code is probably the old RETMA code and has brown-black-black with a multiplier of red which equals 10,000pf or .01uf. Then there are two green dots for 5% and 500vdc rating. I found several .01uf square-package capacitors. I tested their working voltage up to +340vdc (max that the bench supply can achieve) and all caps handled the voltage fine. I chose the one that looked closest to the original. I lifted one end of the CH A MO plate load 300 ohm 1W CC resistor R148 and found that it too was cracked from overheating due to the shorted C117. I found some old style IRC 1W CC that were 390 ohm which is probably close enough. I couldn't get the soldering iron in behind the MO tube shield. I was going to take out the shield by removing the six screws, washers and nuts but I couldn't budge the screws (and I didn't want to break the screw heads off.) The nuts weren't accessible at all. I went back to trying to remove C117 by breaking one of the leads using a "soldering aid" tool but during the process, the solder lug broke instead. I had to use a different ground by adding a solder lug to a ground screw that was nearby. The parts were installed and they look like an original placement using original parts. I installed one of the 3000kc to 9050kc tuning units in CH D slot and applied power. I had 13mA of PA grid current, 90mA of PA Plate current and 50 watts output power. The +380vdc was +350vdc under load.

Next, I installed the other 3000kc to 9050kc tuning unit and again I had 50 watts output with about 13mA of PA grid current. I installed both 540kc to 1500kc tuning units and then I had what appeared to be a short with no PA grid current. I pulled the BC tuning units and then the ATD went back to having full output. I plugged in one BC tuning unit at a time to determine which one was causing the problem. It was the BC tuning unit I had tuned to 1100kc. The problem is in the MO circuit of the tuning unit and that involves one tapped coil, a five-position switch, three capacitors and some wires. By selecting various switch positions 1 through 5, each of the three capacitors can be eliminated as the problem. That would leave the tapped coil or the wires. Since the rear lever switch shorts out the coil when the tuning unit isn't selected, it would seem that the coil could be eliminated. Since the AM-BC tuning units are virtually worthless and could never be used because their frequency of operation is entirely within the AM-BC band, perhaps the easiest approach is to remove P601 (a banana plug) which is the source of the +380vdc for the tuning unit. This, essentially, would eliminate the problem. The defective BC tuning unit will be replaced as soon as I find a 1500kc to 3000kc tuning unit, which would be usable on the 160M band. Or, another 3000kc-9050kc could provide an additional 80M or even a 40M frequency that could actually be used. NOTE: Jun 3, 2025 - I found two 3000kc to 9050kc tuning modules, serviced them and they are now operational. I now have four 3000kc to 9050kc modules installed that allow four frequencies to be selected in the 80M or 40M bands.

NOTE: When installing the "disconnected" BC tuning unit into the main frame, I noticed a very slight change in the "tune and load." It shows as a 5mA reduction in PA plate current. The load on the +380vdc doesn't change. A slight "touch-up" on the tuning of the two 3000kc to 9050kc modules was required. I think the proximity of any of the tuning units to the RF output buss wiring probably slightly changes the stray capacitance to ground and very slightly changes the "tune and load" settings. Once all tuning units are installed, the final "touch-up" seems to be consistent and remains at those settings.

RF Amp Meter Conversion - 9-14-24 - I picked up the RF Amp meter that came out of a Bendix TA-12 from W7MS today at the Cabela's SNARS Ham Swap Meet in Verdi, Nevada. Mike also included a bag with some extra bezels and an extra spacer. This RF Amp meter, being for a TA-12, had a white scale while the ATD RF Amp meter had a black scale. I had a "dummy" ART-13 RF Amp meter installed in the ATD just to fill the hole.

Luckily, the TA-12 meter is a Weston 507 and the ART-13 meter is a Weston 507, so the scales are interchangeable. The scale swap was easily accomplished although the TA-12 meter needle needed to be white instead of black. I used a Sharpie White Paint Pen which is a very thin paint that covered the needle perfectly in one easily applied thin coat. The needle stops were missing on the TA-12 meter but present on the ART-13 meter. Essentially, between the two meters I made one black scale, 5 Amp RF current meter for the ATD.

The top cover on the ATD had to be dismounted to have access to the RF Amp meter. Next was some work on the spacer which, of course, was for a TA-12. It wouldn't fit into the ATD without some involved modifications. The right side of the spacer had to be cut in order to have clearance for the metal frame's right side shielding of the ATD. This was accomplished carefully with a hacksaw followed by some dressing down with a file. This cut also was necessary for the meter flange. Then the left side had to have .250" of relief depth (removing spacer material) from about 200º up to about 350º for clearance of the ATD metal framework. Finally, about .190" of relief was necessary at the top (almost down to the threaded hole) and additional .250" relief depth from 5º to 20º to clear the upper framework. Once these cuts were made, the spacer fit correctly up flush with the back of the panel. The rebuilt meter was mounted to the spacer and then the entire assembly was mounted to the front panel using the original meter "flange bezel" the rear portion of which fits through the inside of the spacer and hides the spacer when viewed from the front through the bezel glass.

RF Amp meter hook-up required a small section of 12 gauge buss wire because the original meter had been "cut out" leaving the existing buss wire too short to reach the meter. I used a crimp splice (with the plastic removed) and soldered it to make a junction between the original buss wire and the new section going to the meter. The original output buss wire I had connected to the sending relay that then switched to the ANTENNA terminal when transmitting. Now this buss wire had be connected to the meter. Then a new 12 gauge wire had to be run from the other meter terminal to the sending relay. That completed the hook-up. NOTE: Jun 3, 2025 - I found a "RF OUTPUT" curved meter metal tag that is now installed on the RF Amp meter. Original metal tags had "ANTENNA CURRENT" but "RF OUTPUT" is close enough.   NOTE 2: Aug 25, 2025 - I have no idea how the RF Amp meter was originally mounted in the ATD. I finally was able to purchase a complete and original Navy manual for the ATD and there are no photos, no drawings,...not even a description of how the RF Amp meter was physically installed. Lots and lots of mechanical drawings in the manual but absolutely NO PHOTOGRAPHS at all.

Operation - With a 50Z antenna load at a 1:1 SWR and only about 45 watts input, the Antenna Current shows about .4 amps, so the meter needle doesn't move very much at all (mathematically, if 200 watts was the RF output and the 50Z antenna was a perfect resistance only, then 2.0 amps should show on the meter using P/R=√I.) With a very low impedance antenna, such as an electrically short vertical, the current would read much higher. However, this meter scale is for the ART-13 which uses a current transformer in series with the antenna output and the meter I'm using is a thermocouple-type of RF Amp meter so the actual accuracy of this RF Amp meter is definitely compromised (most Antenna Current meters were just used a "peak output" indicators and not for calculating power output since the antenna impedance was probably unknown.) I noticed that with the RF Amp meter installed inline with the RF output now that some slight readjustment of the tuning modules was necessary. Not totally unexpected and the power output is still at 45 watts to 50 watts (measured with an external but inline wattmeter.)


The RF Amp meter is now recessed back about one inch behind the front meter bezel glass. The genuine ATD meter had "RF AMPERES" on the meter scale instead of "ANTENNA CURRENT" that the ART-13 meter scale has. The projection of the scale is also different. With a 50Z antenna and only 45 watts output, the RF current shown on the meter only reads about .4 amps but, with a different scale, it's not an accurate indication anyway. Several months later, I found a curved metal "RF OUTPUT" tag to go on the meter.

Second Net Operation - 9-22-24 - I did a pre-net check out of the ATD loaded to 45 watts output into the Collinear Array at a 1:1 match. I was watching the oscilloscope that monitored the wave envelope sampled at the Johnson Matchbox. This time the waveform looked symmetrical with equal positive and negative modulation. The percentage of modulation still didn't appear to achieve cut-off or 100% negative but that's probably due to the +380vdc running at +350vdc and dropping to +330vdc during voice peaks. This is actually much better than the first "on the air" attempt. I monitored the signal quality by listening to the HRO-7T without an antenna connected and the RF gain reduced as needed. The audio quality sounded very good. So, I was ready for the net. W7MS was also going to run his ATD transmitter. Mike was running his ATD on the dynamotor and was routing the RF out to a Collins 30S-1 linear that boosted his RF level to 300 watts of carrier,...quite a signal. I was relying on the gain of the Collinear Array to help my 45 watts of carrier. I was using the USN/Hammarlund RBG-2 as the receiver. All stations were Q5. Although my signal reports for the ATD weren't as strong as normal, I did get between S7 and S9 from most stations, which is okay for 45 watts. The ATD was much more reliable this time. No problems. The RF Amp meter barely moves during transmission but that's expected with the 50Z low-reactance load of the Collinear Array. Later, I listened to Wavelength Radio's online recording of the net specifically to see the panadaptor display of the ATD signal. It still appeared to favor the upper sideband in the display but, interestingly, Mike's ATD also showed similar characteristics on the panadaptor. More investigation required after I have the ability to power up the ATD with the dynamotor so I have sufficient current available for the +380vdc.
ART/PS +400vdc Redesign - Re-thinking this Project - In order to increase the current availability for the +380vdc, I'll need to rebuild that section of the PS. I've found a 435-0-435vac transformer that's supposed to be rated at 4 amps but that hardly seems likely since that's about 1700 watts on each side of center tap. The rating is probably .4 amps or 400mA which is spec for the dynamotor current on the +380vdc. I've found two GE chokes that are rated at 435mA. I'll probably only need one but, if I have room, I could make a dual choke-input filter for better filtering. I've also found a 5vac filament transformer rated at 10 amps although I really only need 6 amps. I'm going to have to use two 5U4GB tubes in parallel to get the current available up to 400mA. I'll need to punch another hole for another octal tube socket. I have to remove the existing 400vac transformer, the two chokes in that supply and the 5vac filament 3 amp filament transformer. I'll have to move the +28vdc "ON" relay to a location under the chassis. I'm going to move the +28vdc Lambda PS to the rear of the chassis and orient it lengthwise. That will give me sufficient space for the new components. I have to mount the 435vac CT transformer on its side using a "L" bracket-type mount that has the long-side of the "L" mounted to the chassis and then the transformer mounted to the short-side of the "L." The transformer is 6.25" tall and that's about .25" taller than the space I have above the chassis with the cabinet lid closed. With a side mount using a "L" bracket, the transformer measures 5.625" tall.

THIS SOUNDS LIKE A LOT OF WORK! - I'm having second thoughts about this involved of a modification to the Homage a le Valve ART-13 power supply. So, for now,...further modifications to the Homage a le Valve ART-13 PS has been halted. Now I'm going to take a look at the possibility of dynamotor operation,... 

 

Servicing the ATD Dynamotor
Constructing an "Easy-to-Build" 80 Amp +28vdc Power Supply
Operating the ATD on Dynamotor Power

With all of the modifications required for the "Homage a le Valve" ART/PS, sometimes it seems that it would be easier to just use the dynamotor. The 27 amp +27vdc Lambda power supply doesn't provide enough current to even start the dynamotor. Though the manual indicates that about 20 amps is required for ATD-phone operation, this is "running current" - not "instantaneous starting current" - which obviously exceeds the capabilities of the 27 amp Lambda. However, there's an old trick of connecting similar power supplies in parallel to increase their current capabilities. I connected TWO 13 amp Meanwell +27vdc switchers in parallel with the 27 amp Lambda. I set all three PS outputs for +27vdc +/- 0.2vdc. In theory, I should have 53 Amps available. Will this actually work? Read on,...

Servicing the Dynamotor - Although this dynamotor appears to be NOS (and probably hasn't ever been used,) that doesn't mean it would just power right up with +28vdc applied. The bearings still have 80 year old grease inside and the brushes have probably been setting in the same position on the commutators and in their brush-barrels since the early-1940s. All "as found" dynamotors really need to be taken apart and examined first. If everything looks good, then the commutators can be cleaned, the brush surfaces cleaned, the bearings cleaned and relubed and any other clean-up required performed. The component box should be checked. Only after this servicing is performed can the dynamotor be powered up. 

The end-bells on this dynamotor still have the mounting screws "safety-wired" and it's unfortunate that the wire has to be removed in order to dismount the end-bells to have access to the bearings, brushes and commutators. With the end-bells off, I could use the fan to manually rotate the armature to see if it would move and how much "drag" there was. The armature did move and the drag was considerable.

Component box inspection turned up a leaky oil-filled capacitor. This was physically "leaking oil" not current leakage. Although it looked like a lot of oil it really wasn't. The residue was cleaned up with a Q-tip. This "weeping" of oil is common and usually isn't a problem if it's just the "weeping" type of leakage. The capacitor will have to be monitored from time to time to see how much (if any) oil leaks out. Everything else in the component box looked in good condition. All spare fuses and tools were present.

Pulled all of the brushes. These were placed on a white sheet of paper in the proper orientation and placement for reassembly. Pulled the fan off of the shaft to allow access to the bearings and motor-side commutator. Pulled the bearing covers on both ends. Dried grease was obvious. I used WD-40 as a solvent to loosen and remove the old grease. I worked the WD-40 into the bearings on both sides by rotating the armature. At first, with the dried grease, the shaft was moderately difficult to turn and wouldn't "spin." After the WD-40 soaking, the armature spins easily. I let the WD-40 soak in the bearings all night.


ATD Dynamotor - TYPE DMDFX618 sn: 260453
Mfg. by: Continental Electric Co., Inc.

The next day, I cleaned the commutators with 600 grit Al-Ox paper and then washed the commutators with denatured alcohol. I repacked the bearing using red wheel bearing grease and then reinstalled the bearing caps. I checked the brushes and they looked new but a gave them a gentle cleaning with a brass brush just to get any old dirt off. The brushes were installed exactly in the positions that I had removed them from. I cleaned the area around the commutators with denatured alcohol to remove any old excess grease that had been "pushed out" by the repacking process. This completed the servicing.

A Tiny Power Supply for Dynamotor Operation - To power up the dynamotor requires at least 45 amps of starting current. I tried 40 amps and the dynamotor pulses, the relay chatters and the armature slowly rolls but never increases speed. At 53 amps the dynamotor starts right up and gets to full speed in about a half-second. Actually, these switchers will supply even more current "peak" for about ten seconds which is more than enough time to start up the dynamotor. The "peak" current available is about 65 amps. I used an old ART-13 dynamotor battery cable that I modified by reworking the connector. The ART-13 dynamotor uses pin 2 for +28vdc input while the ATD dynamotor uses pin 3 for the +28vdc input. I had to simulate the ATD PTT that "turns on" the dynamotor. For this I used two test leads,...one for +27vdc going to pin 1 and then ground going to pin 4. This connection actuates the dynamotor relay and starts the unit. I measured the voltage output with no load and had about +1050vdc on pin 8 and about +405vdc on pin 5.

The test setup is shown to the right. I found the two spare Meanwell 24vdc 13A power supplies out in the shop. They're connected in parallel using 16 gauge wire to provide 26 amps. The Lambda PS is to the far right and it's connected in parallel to the Meanwell combo. The Lambda provides 27 amps. All three supplies were set to +27.02vdc (this was the highest voltage that one of the Meanwell would adjust to, so it determined the voltage setting for the other two supplies.) The output of the Lambda has the battery-type cable that has the ATD-Dynamotor connector on the other end (this cable is also shielded so it doesn't radiate any RFI.) Since the output of the +HV is +1000vdc, I couldn't use the Fluke DDM, I had to use the Triplett VOM to measure that voltage. 

I used a rubber mallet and a wooden block to remove a dent in one of the end bells. Then the end bells were touched-up with black acrylic paint and then installed. This completed the dynamotor clean-up and servicing. Note,...this is one LOUD dynamotor (like a big shop-vac) I was worried about the "screaming fan" on the Lambda PS,...I can't even hear it when the dynamotor is running.


Here's the "test set-up" for the ATD Dynamotor. The parallel switcher-type power supplies provide 65 amps peak at +27vdc, more than enough current to handle the "instantaneous starting" current to get the dynamotor spinning quickly (with "no load.")

What!  Another Project? - Since it seemed that 53 amps was probably about just 5 or 6 amps over what was necessary just to get the dynamotor to spin with no load, I wondered how the extra load of supplying the +28vdc necessary to power up the ATD would affect the dynamotor operation. About 5 amps are required for the +28vdc on the ATD alone. Each time the PTT is depressed, the dynamotor would have to "start up" and that brings in the "instantaneous starting" current issue. It would be nice to have a lot more "head room" as far as current availability to smooth out the +28vdc during the PTT actuation. Would it be possible to order two more 24vdc 27amp Lambda power supplies? Yes, there are many listed on eBay. Would it be possible to connect three of these type Lambda Power Supplies in parallel and have the current availability be 81 Amps continuous and 93 Amps peak? It seems so. I've ordered the two additional Lambda power supplies and when they arrive,...more experiments will reveal if operation of the complete ATD-dynamotor system is possible with these tiny (but noisy) power supplies. In actual operation, the high current would only be necessary to start the dynamotor. As soon as the armature begins to rotate, the current demand rapidly drops off. When running at speed, the maximum current draw is spec'd at 19 amps and that's the total current including the 5 amps of current that is necessary for the tubes and relays in the ATD. However, the instantaneous current draw for starting the dynamotor from a "dead stop" can be quite high and has obviously been demonstrated to be between 45 and 50 amps,...and that's with "no load." So high current availability is necessary for quick-starting with the PTT actuation and minimal fluctuation of the +28vdc to the ATD. 

Construction of an 80 Amp +28vdc Power Supply That's Physically the Same Size as the Dynamotor - When the new Lambda power supplies arrived, I discovered that although they had the same specifications, they were longer in length and not as wide as the old 27A Lambda. One letter in the Part Number made the difference. It doesn't matter as long as they all can provide 27A. I tested these new supplies and they work fine. I had planned on using a small BUD Industries cabinet that was NOS condition. It measured 12"W x 8"H x 9"D and it had an unused aluminum chassis that fit inside. However, with the different dimensions of the two new supplies, this BUD cabinet is too small. But, I have another unused cabinet that measures 8"H x 14"W x 9"D and it also includes a chassis. But, thinking about the power supplies and their fans and the cooling,...why use a cabinet? The power supplies can just mount on top of the chassis. I had thought that due to the noise of the dynamotor and the power supplies, I'd place both units under the desk. There's plenty of space and lots of air available. The location would have the direct noise blocked, the cables are long enough, etc., so just a chassis for the PS should be easy and will work fine. I plan on using 14 gauge wire for the AC input connections and 10 gauge for the +27vdc 80A output. As mentioned above, this high-current is only "momentary" but I want to make sure that the IR drop is as low as possible and that's achieved with large gauge wiring. The normal "running" current demand will be about 20A, so I'll have plenty of "head room" and that should allow AM modulation voice peaks to not affect the +380vdc very much.

I made my own terminal blocks out of .250" thick delrin. Delrin is a type of nylon that is very strong and heat resistant. I drilled and tapped holes in one block for 4-40 brass screws. The holes were countersunk to allow the screw heads to be flush with the surface when installed. Then a second delrin block was used for the insulated base mount. This allowed the terminal block to be bolted to the chassis and the brass screw heads were insulated from the chassis by the delrin base. The 14 gauge power cord was routed into the chassis at the side and the Line, Neutral and Ground wires soldered to the ends of the three brass screws of the terminal block (the ends were .750" long.) Then three additional 14 gauge wires were soldered to the Line and the Neutral terminals and these routed up through chassis holes to the AC input to the three power supplies. These chassis holes had rubber grommets installed.


DC Output Terminal Block Inside Chassis - the studs are 1/4x20 brass and the wires are 10ga.


AC Input Terminal Block Inside Chassis

Since the Lambda power supply chassis are all mounted directly to the aluminum chassis and the Ground terminal on each Lambda power supply is connected to the power supply chassis, the individual Ground terminals on each power supply don't have to be routed to the AC ground wire terminal. But, the AC power cord Ground wire is connected directly to chassis in two places one of which is a lug under one of the Lambda power supplies mounting screws. The AC input to each power supply uses a soldered ring lug to connect the 14 gauge wires to Line and Neutral. There isn't a power switch or fuse in the AC input line. The AC line "bounce" is very noticeable when powering up the dynamotor so a switch and fuse would just limit the available AC line current somewhat. The 80A PS would only be plugged in when it's being used.

The DC output of each Lambda was then adjusted to +28.0vdc BEFORE the three were connected in parallel. I tried to get the output voltages exactly the same from each supply for equal distribution of the current carrying ability (within +/- .002vdc.) The DC output terminal block was made from two pieces of .25" thick delrin with 1/4x20 brass studs and brass nuts and washers to act as the high current output connections. The clearance holes through the chassis for the DC terminal block studs were .750" in diameter. The 10 gauge wires were soldered directly to the back end of the DC output studs. The power supply end of the 10 gauge wires had ring lugs soldered on and then these were attached to the PS output terminal strip. The wires were routed through the chassis by way of grommet-lined holes.

With the completion of the wiring, I connected the ATD dynamotor using the shielded ATD dynamotor cable. Pin 1 has to connect to +28vdc and pin 4 is connected to ground to energize the dynamotor relay. Start up was immediate with no hesitation and the armature got up to speed in about 1 second.

Switching noise is always a concern with these types of power supplies. To keep the RFI as low as possible the AC input and the DC output were bypassed with .05uf capacitors. The DC cable to the dynamotor is shielded and the power cable from the dynamotor output to the ATD is also a shielded cable. Each power supply is in a metal shielded case that is mounted directly to the PS chassis and the chassis is connected to the AC house ground. I switched on the Siemens E311 receiver tuned to 3.9mc and using a 50ft wire on the floor as a "noise pick up" antenna. The power supply RFI noise wasn't noticeable (not above the ambient noise level.) The starting of the dynamotor is quite a load, so the receiver reacted to the momentary AC line bounce but when the dynamotor was up to speed no RFI noise was heard.

NOTE: It's certainly noticeable that I didn't use the standard "black-white-green" color code for the Line-Neutral-Ground wires. The only 14 gauge stranded wire I had on-hand was green and red, so that's what I used. The 10 gauge DC wires are red for + and black for -, which is more or less standard. Also, note the labels on the power supplies showing the "CE" certification for RFI noise reduction.


Rear View of 80A-PS showing the wiring to the 3 Lambdas


Front View of the 80A-PS showing the DC Output terminals and Heavy-duty AC power cord (14-3 conductors)

In either case, the RFI noise didn't change the S-meter reading reacting to the ambient noise and was only just barely aurally detectable. The receiving antenna wire was in the same room as the dynamotor and power supply. With an outdoor antenna I don't think any noise will be heard in the receiver. In receive, only the power supply is running. When going to transmit, the receiver is disconnected from the antenna and put in standby so no noise would be heard. Only switching power supply noise is on at the same time as the receiver is in operation.
 

The Ultimate Test for the 80A-PS: Oct 7, 2024 - As an "ultimate test" of the 80A-PS's capability, I carried it out to the shop (with just one hand, of course) and I then connected up the GRC-19 transmitter/receiver to the 80A-PS. The GRC-19 is a T-195A transmitter with a built-in +LV dynamotor, a rotating AC generator and, since it's the A version, it has the DC to DC converter/PS replacement for the +HV dynamotor and, in addition, it has three blowers to keep the three 4X150D tubes cool. The T-195A draws about 30 amps running and the R-392 adds another 3 amps. The starting current momentarily pegs the PP-1104-C amp meter (probably some meter-needle "over-shoot") for an instant but there's no hesitation in power-up. The GRC-19 has the highest DC current requirements of any device that I have, so it's a good test for the 80A-PS. The little 80A-PS powered-up the GRC-19 with no hesitation at all. After a short warm-up, I pressed the TEST KEY. I had 125 watts output to the 130' Inv-Vee antenna (shop antenna.) No voltage bounce, just pure DC when transmitting VOICE or working the TEST KEY. The 80A-PS is equal to the PP-1104-C in providing +28vdc at a high current level. And, it only weighs about 15 pounds compared to the 100 pounds of the PP-1104-C (and the 80A-PS runs great on 120vac but the PP-1104-C runs best on 230vac so a special 230vac outlet and wiring are necessary for its proper operation.)

A Gladding & Keystone contract PP-1104-C is shown to the right. These were originally part of a battery charging system used by the military. The power supply can be set to 12vdc at 100 amps or for 24vdc at 50 amps by how the four small studs (between the two main output studs) are connected together using supplied jumper straps. The adjustment knob in the center of the panel is a stepped-switch and you can't adjust the output voltage with the "1104" turned ON. Set first, power up to check voltage, power down to reset adjustment, if necessary - a real pain. Also, the "1104" will operate much better and can supply its rated current if it's powered by 230vac input. They can operate on 115vac at reduced efficiency (this can cause the dynamotor starting relay to "chatter" since the PP-1104-C voltage output drops too low with the load - this only happens on some dynamotors and when the PP-1104-C is operated on 115vac.)


The 100 pound PP-1104-C

Running the ATD on the Dynamotor - The end of the ATD power cable (that had connected to the ART-13 AC PS) had to be modified to use plug-in wire sockets since I don't have an original connector. I removed the ATD cable from the ART-PS. I then clipped off the ring lugs and cut the six wires all to the same length. I salvaged the receptacles from an old six pin fiber board tube socket to make suitable push-on female pins for the cable. The connections are easy for the dynamotor to ATD cable since both box connectors are identical and have the same connections, that is, pin 1 to pin 1, pin 2 to pin 2, etc., the colors of the wires can be used for reference. Also, in this cable two wires are 14 gauge for the +28vdc in and the Chassis-Gnd. The other wires are 16 gauge for PTT, Dynamotor Relay return, +380vdc and +1000vdc. The wire push-on sockets were covered with rubber tubing for insulation. The shield of the cable is connected to the chassis of the ATD on one end and to the Dynamotor housing on the other end.

Since the ATD Dynamotor sounds like a very large shop-vac when running and the 80 amp power supply has three fans running, both of these units were placed under the operating desk. The cable is long enough and there's plenty of room under the desk. I did this the last time I ran a large dynamotor in the upstairs shack (the RU-16/GF-11 set up) and it really does quiet-down the operation.

As mentioned, I had to modify the ART-13 dynamotor cable to work exclusively with the ATD dynamotor. I have another ART-13 dynamotor cable, so changing this one to an exclusive ATD dynamotor cable wasn't a problem. Incidentally, the two large wires in the ATD Dynamotor Cable are 8 gauge wires. I connected up everything under the desk and applied 120vac to the 80A-PS and it came up fine. I used the Throttle Switch to energize the ATD dynamotor. It takes about 1.5 seconds for the armature to come up to full speed. The Grid Current and Plate Current meters don't just "jump" to the proper current, they "ramp up" as the dynamotor comes up to speed. I had about 48 watts output with a 1:1 SWR match and the PA Plate Current was 75mA.


Showing the Dynamotor and the 80A-PS Location under the Desk's footwell

Setting up the ATD to run on the Dynamotor and 80A-PS - Once I had the ATD and Dynamotor running on the 80A-PS with the Collinear Array connected, I attached the 50 ft wire laying on the floor (for maximum RFI noise pick-up) to the 51J-4 receiver in the shack. With just the 80A-PS running no RFI or noise of any kind was heard in the receiver (other than the usual ambient RF noise.) I tuned about 1mc higher than the 3.974mc that the ATD was set up on. Then, using the Throttle Switch, I put the ATD into transmit. Again, no noise was heard in the receiver except for the initial switch pulse of the dynamotor turning on. Looking at the oscilloscope, the waveform looked stable and only a slight ripple was visible at the top and bottom of the waveform.

IMPORTANT NOTE: So, why don't I have any RFI Noise in the receivers when I'm using THREE switching power supplies simultaneously? The most significant noise reduction comes from having EVERYTHING grounded and shielded. The DC power supply to dynamotor input cable is fully SHIELDED. Also, a significant reduction is from having the Dynamotor to ATD power cable SHIELDED. The two ART-13 AC power supplies I built both use Lambda "switchers" for the +28vdc and, again, chassis and cabinets are grounded and ALL cables are SHIELDED. There's no RFI when using these power supplies either. All of the "switchers" that I use are Lambdas (I have some Meanwells for experimentation and test applications but for "on the air" applications, I use only Lambdas.) These Lambda 27 Amp switchers are new enough that they have the "CE" certification label that generally indicates they meet modern specifications for "low noise output." The "CE" switchers have much better filtering than the older types that were infamous RFI generators (and were in use maybe 20 or 30 years ago.) The 80A-PS has the +28vdc current availability that a PP-1104-C has but instead of weighing 100 pounds and being the size of a large stand-up laundry hamper, the 80A-PS can be picked up with one hand (weight is 15 pounds) and is about the same physical size as the ATD dynamotor. I have three PP-1104-C power supplies out in the shop. They're mounted on small furniture dollies and can be rolled around the shop, but it would be totally impractical to even try to move one of the PP-1104-Cs upstairs. They're great high-current power supplies but they aren't "portable" in any sense of the word. The 80A-PS, on the other hand, is easy to move anywhere for experimentation, troubleshooting or operation. Cost? The two recently purchased 27A Lambdas were $45 each. The older 27A Lambda was purchased last year and was $60. So, they aren't really very expensive for what you get. These are not "new" power supplies but are "pulls from working equipment, tested and working" condition. They're always available on eBay.

Operating the ATD-Dynamotor-80A-PS "On the Air" - Before going "on the air" again, I wanted to switch to the Shure 102C carbon microphone, just to be more authentic sounding. The Ch.D module has to be removed to access the microphone switch and a right-angle blade screwdriver helps due to the limited space. I powered-up the ATD and watched the wave envelope on the 'scope. Lots of symmetrical modulation and almost approaching "cut-off" or 100% negative going on voice peaks. I monitored the audio quality by listening on the HRO-7T receiver and, for a carbon mike, the Shure 102C sounds excellent. Also, the dynamotor noise was minimal and no other "noise" was heard on the signal.

Oct 13, 2024 - Nevada Mil-Rad Net - I ran the ATD-Dynamotor combo today using the Shure 102C carbon mike, USN/Hammarlund RBG-2 receiver and the Collinear Array antenna. Very good conditions and the ATD was Q5 from all participants. Looking at the oscilloscope which uses a pick-up coil located inside the Johnson Matchbox, the ATD modulation looked symmetrical. All signal reports were quite complimentary about the ATD audio, even though I was using a carbon mike. Later, I was able to check the Wavelength Radio (on youtube) recording of the NV Mil-Rad Net. This recording features a panadaptor display that allows seeing what the SDR receiver picks-up. Although the modulation level looked good, it was "lop-sided" and favored the upper sideband. When looking at the spectrum, the favoring of the upper sideband wasn't so apparent and it looked like the band width from the ATD signal was about 8kc wide. None of the other Mil-Rad Net stations mentioned anything about this anomaly and it's very likely that, unless someone actually tuned their receiver from one side band to the other, they probably wouldn't actually hear the sideband difference. Since Mike's ATD also showed this same characteristic of favoring the upper side band, I assume that it's normal for the ATD. It's something that wouldn't have been heard in normal reception and would have required a panadaptor to see. Locally, with the oscilloscope monitoring the wave envelope, the modulation looks symmetrical. Other than what could be seen on the panadaptor on the recording, the actual "real time" operation on the NV Mil-Rad Net was flawless.

Vintage Mil-Rad Station for May 2025 - ATD and the RAF R1155 - I've wanted to use my RAF "Lancaster Bomber" R1155 receiver "on the air" for quite a while. While the combination of British aircraft receiver and USA aircraft transmitter might seem "mis-matched" they do look great together. As for operation, I connected the R1155 antenna input to the "Receiver" terminal on the ATD for isolation but still the volume has to be reduced when transmitting. When using 'phones and transmitting Voice, the signal can be heard over the 'phones and can be used as a monitor of sorts. The combo works very well on 75M. Power to the R1155 is provided by a test bench power supply, the Lambda Model 25 (6.3vac and +220vdc) and a small RCA test bench power supply for the bias voltage (-25vdc.) Both are located on a desk behind the set-up with the cable from the R1155 routed back to the supplies. The ATD still uses the homebrew +28vdc 80amp power supply to operate its original dynamotor (both are still located on the floor under the desk in the foot well.)  The mike is a Shure 102C carbon microphone and the key is a USN flameproof type. Phones on top of the R1155 are Trimm Radio Co. Hi-Z type from the mid-1920s. I've checked into the Nevada Mil-Rad Net on 3.974mc and had no problems in either transmitting or in receiving. Even though conditions were poor the ATD's 40 watts and carbon mike was Q5 to all stations.

NOTE: Although the R1155 works well on 75M, it's not the greatest receiver ever built. Even though it's a ten tube receiver, only six tubes are actually used for the receiver with the other four tubes being for DF functions. A separate power supply is necessary. There's no control for selectivity and most of the controls on the receiver are for DF. For receive the controls are Volume, BFO switch on-off, Band switch, Mode switch and Tuning. It's an impressive-looking receiver with a fabulous (but non-illuminated) dial,...just not too many adjustments for a compulsive "knob twiddler" like me.


USN ATD Transmitter with the RAF R1155 Receiver
2025 photo shows all four Tuning Modules are 3000kc to 9050kc. RF OUTPUT tag is on RF Amp meter.

 

More ATD Work - Once in a While Needed Parts do Turn Up

Two Spare ATD Tuning Modules Turn Up - May, 2025 - I'd been watching eBay for any ATD parts or components to show up but nothing ever turned up so I quit checking after a couple of months. Luckily, I daily check "WWII Radio" and "WWII Aircraft Radio" as searches and now, after seven months, I saw two ATD Tuning Modules show up as new listings. BIN for $35 each and they looked in good, complete condition. The best part was that both modules were the 3000kc-9050kc, meaning that 80M and 40M would be covered. These two modules will replace the two useless 540kc-1500kc modules. When installed in my ATD, I'd have four 3000kc-9050kc modules and four channels could be set up for various frequencies in either the 80M or 40M bands. But, first a thorough cleaning and check out.

The two Tuning Modules were in good physical condition although both were a little dirty and one had a little more surface corrosion than the other one. Internally, both modules were complete and in excellent condition with no modifications. Both of these Tuning Modules are CRR-47210 and therefore were built by Bendix Radio. The other two 3000-9050kc modules I have are CQW-47210 built by Sparks-Withington. Both types of modules are virtually identical (slight differences in the fonts used for some of the nomenclature.) All that was required was to lubricate all of the controls so they would rotate easily. The ANT SW control required just a touch of grease on the detent. Also, DeOxit on the spring-loaded contacts. One module was missing a front panel 4-40 BH screw, washer and nut. I ended up replacing both of the top panel 4-40 BH screws with matching vintage screws. The contacts on the rear of the module were cleaned and given a shot of DeOxit where they pivot. I checked the three large bypass capacitors for shorts and also the DCR to chassis at the B+ input terminal and found no problems.

I installed these new (to me) modules in the Channel A and Channel B positions and powered up the ATD. I set the controls on the new modules close to what Channel C was set to so I'd know about what to expect. I checked to make sure everything was working correctly on Channel C and then switched to Channel B. The module was pretty close, coming up on 4.0mc. I "tweaked and peaked" for 3.9mc and had 15mA of Grid Current which is normal and 40 watts RF output at about 70mA of plate current, which is normal. The waveform looked good on the 'scope. I then switched to Channel A and, although the controls were set close to Channel C, I had a very high SWR indication on the inline RF watt meter and the Grid Current was 21mA. The 'scope pattern was "full screen" as would be expected with mis-tuning or no load. I pulled the Channel A module (switching off of Ch. A first) and examined it. Nothing found but I cleaned the contact terminals and rotated the ANT SW control. I replaced the Channel A module, switched to Channel A and now the output was normal with 15mA of Grid Current and 40 watts of RF output with a normal 'scope pattern. I think there was a minor contact problem in the ANT SW that caused the problem and that switch just needed to be rotated a few times to clear up the problem (it had been DeOxit-treated but this switch still needed to be exercised a bit.)

So, now the ATD has four working channels. I can set any of the four channels to various frequencies in the 80M band or the 40M band. The 3000kc to 9050kc modules are probably the most useful of what modules were available when the ATD was new. The 550kc to 1500kc modules are useless nowadays (except for maybe testing purposes or hardware parts) and, while the 9050kc to 15,800kc might allow 30M or 20M operation, the frequency is pretty high for stable operation and there's very little vintage mil-rad activity on those bands. The 200kc-540kc module would allow 630M operation but there ISN'T ANY CW operation on that band unless a schedule is made. Calling CQ in CW on 630M is a futile effort resulting in nothing. Also, operation using the 200kc to 540kc module would probably require having and using the CRR-47206 LF Antenna Tuner. The 1500kc to 3000kc module would allow 160M operation but that would require some vintage 160M mil-rad activity and, although there are a few 160M mil-rad nets in other areas of the USA, there aren't any here in NW Nevada. Besides, 160M is best during the winter and at night. Also, local RFI on 160M can be brutal (and it's only getting worse as more and more RFI generating devices make it into every household.) The 3000kc to 9050kc modules allow operation on the bands that have most of the mil-rad nets and are at a frequency that allows the ATD to produce a stable signal.  Work completed Jun 3, 2025.
 
R.F. OUTPUT Curved ID Plate for the Antenna Current Meter - I saw a RF Amp meter show up on eBay. It was the typical, standard size bakelite case but the meter was a 9 RF Amps FS unit. All I really wanted was the curved metal ID plate that had "R.F. OUTPUT" on it. There was absolutely zero interest in the meter and the seller must have realized that because every few days he lowered the BIN price by a couple of dollars. It went from at starting price of $20 down to $10 when I finally opt'd to purchase. I had to slightly change the curved metal plate to have a silver border but, otherwise, its installation completed the appearance of the R.F. Ampere Meter (that actually has ANTENNA CURRENT on the scale since it's an ART-13 scale.) The curved metal tag was very slightly under-size and needed a slight change in the arc and in the mounting holes (the holes actually had to be made into "slots" to allow mounting the tag - see photo.) This was accomplished with small jeweler's files. To create the silver boarder I used a small flat file and created a very slight bevel around the edges that removed the black paint evenly and looks absolutely original (the bevel is so slight as to be not noticeable.) The RF Amp meter now looks almost exactly like the original type. Interestingly, the original ATD manual never specifies exactly how the RF Amp meter scale is identified, "RF Amps" or "Antenna Current," "RF Output" or other similar indicators. Also, the metal label on the RF Ampere meter isn't mentioned or shown in the manual. The metal label I used is "R.F. Output" which is okay since that is what the meter indicates. "Antenna Current" shown on the meter scale is acceptable since that's also what is indicated. I saw a couple of ATD photos on the Internet and those transmitters had "AMPERES R.F." on the meter scale and "ANTENNA CURRENT" on the metal tag, so I'd assume that is correct, or original.

Original ATD Manual - Aug 25, 2025 - I thought this complete, original manual would provide a lot of information since it obviously was "loaded with" drawings, charts and schematics. What I had used for the ATD rebuild was the online manual that has the most of the text and one schematic from the manual and nothing else. No complete parts list and no drawings online. I had wondered why the online information was incomplete. Now I know,...the original manual is loaded with repetitious mechanical drawings of each tuning unit, charts for settings and outputs for each individual tuning unit loading into non-ham antennae, a few mechanical drawings for the cables, the Remote, the Indicator, the Dynamotor, one wiring layout chart of the transmitter and one mechanical drawing of the transmitter. There are NO photographs at all. Just mechanical drawings of a lot of seldom used peripheral accessories. Possibly the wiring layout of the transmitter might have been helpful but it wasn't necessary for getting the transmitter operational. So, what is online is really all of the essential information needed to rebuild an ATD. However, my initial disappointment in the manual's lack of photographs was later abated when I found that the original manual did have the complete "parts list" with the correct manufacturer part numbers. This turned out to be quite valuable since I could now find the exact part number for ALL of the connectors required for the cables. The complete parts list was one of the sections NOT provided online, so having the manual then allowed me to easily find ALL of the connectors that I needed for the ATD to connect to the Dynamotor, the Remote Control Unit and the Remote Channel Indicator Unit.
 
Pilot's Remote Control Unit Found - Type CQW-23280 - Sept 4, 2025 - The same seller on eBay that had the two ATD Tuning Modules back in May 2025, put up a good condition ATD Remote Controller. Fortunately, the connectors required for the cable to hook-up the Remote to the ATD are also used on the very popular TCS transmitters making these connectors easy to find and cheap. I referenced the correct part number from the manual, Cannon SK-C16-21C 3/4" and several came up on eBay. One seller had 19 available at about $2.50 each. So, finding these connectors wasn't a problem. Once a shielded cable is built, the Remote can be used (or at least demonstrated) as the original intended operation of the ATD that had the Pilot always in charge of operation of the transmitter through the Pilot's Remote Controller (although I don't think I'll "safety-wire" the ATD's LOCAL-REMOTE toggle switch to REMOTE as was originally done for the two-seater aircraft installation.) The Pilot's Remote is a relatively small unit measuring about 2.25" x 6.0" x 3.0" with a Channel switch, a flat button-type of hand key for CW, a MIC jack, a three-position PHONE/STANDBY-OFF-CODE toggle switch and a TRANS INDICATOR that was a visual "pop-up" indicator so the Pilot would know that the ATD was powered since he usually didn't have direct visibility of the transmitter. Once the ATD transmitter was turned on by the Pilot (with the ATD transmitter in REMOTE) and he selected the mode and channel to be used, then the Operator could fully operate the ATD from his position at rear of the cockpit. SIDETONE was usually routed to the receiver audio (to the headsets used by the pilot and operator.) When in "transmit" the ATD SIDETONE would directly drive the 'phones being used for the receiver so both the Pilot and the Operator could monitor transmissions (the manual shows a set of 'phones connected to the ATD SIDETONE implying that only the Operator could monitor transmissions, which doesn't seem likely.) Not all Remotes have the "OFF" engraved nomenclature. The CRR types built by Bendix have the "OFF" position indicated. The CQW types built by Sparks-Withington don't have the "OFF" nomenclature as can be seen in the photo to the right.

Remote Channel Indicator Unit Type CRR-55079 - Quite a while ago, I found an original ATD Remote Channel Indicator on eBay. The four "pop-up" indicators are labeled A, B, C and D and will protrude up almost half an inch with easy visibility since the cylinder is silver metal and the cap is black plastic. The plastic tag on top was for writing the frequencies for Channels A, B, C and D. Unscrewing the knurled thumbscrews allows removing the bottom for access inside or for mounting the base. This is a very small unit measuring only 2.25" x 3.25" x 2.25" with only five wires required in the connecting cable. This Remote Channel Indicator came with its original hook-up cable but that cable was made up of wires installed inside flexible conduit that had a black rubber jacket on the outside of the flex conduit. One would think that it's an original cable since ALL ATD cables were originally custom-made for the specific installation of the equipment into the airplane. However, flexible conduit isn't what's called out in the manual. The manual indicates rubber-jacketed shielded cable. The use of flexible conduit makes this cable somewhat difficult to straighten, to bend or to do anything for ease of routing. However, it's obviously an original cable so I'll use it, after I repair it. Unfortunately, one connector was apparently removed and, in doing this, the wires were "ripped" from the connector and are back about two inches into the flex-conduit. By removing the opposite side connector from the conduit, I can pull the wires out of the conduit from the other side. This would allow replacing the wires (since the originals are now not long enough) and have the cable usable. The Remote Channel Indicator was supposed to be mounted so the rear seated Operator could see the Remote Indicator buttons for confirmation of which channel the pilot had selected (for the Operator) if the ATD transmitter's channel indicating wheel on the front panel wasn't visible to the Operator (sometimes the ATD was mounted under the Pilot's seat facing aft making it difficult for the Operator to see the front panel.) In the standard set-up for operation, the Pilot was always in control of the transmitter for applying power, selecting either CODE or PHONE and for selecting the channel for transmissions. If the Operator needed to make a transmission on a different channel or use a different mode of transmission he had to ask the Pilot to use the Pilot's Remote Control to make the channel and/or mode change. But, after the power on, mode selection and channel selection were made for him, the Operator then had full control of the ATD.

 

Building and Rebuilding Connection Cables for the ATD Accessories

Rebuilding the Power Cable (Dynamotor to ATD) - This "rebuild" is just a matter of removing the tube socket receptacle pins from the wires on each end of the cable and then installing the Cannon NK-M8-23C 11/16" connectors on each end. Six wires comprise the power cable connections even though the Cannon plug has eight pins. Two wires are 14 gauge for the +28vdc and Chassis-Ground connections and the other four wires are 16 gauge for the +1000vdc, the +380vdc and the PTT to Ground (actuate PTT by grounding this wire) and PTT +28vdc to the Dynamotor Relay connection (this is a separate +28vdc from the main +28vdc 14 gauge wire.) The cable is about six feet long and is completely shielded using the copper braid harvested from a piece of RG-8 coax. Two layers of black electrician's tape make up the insulation over the copper braid. There are two chassis-drain wires attached to the braid at each end. These drain wires are mechanically attached to each of the connector shells. This provides the ground for the shield running from the Dynamotor chassis to the ATD chassis. Since I had to use two right-angle connectors, I decided to use a 3.75" riser-platform for the ATD. That way the cables can drop down several inches before being routed over the desk top and down the desk side to the Dynamotor on the floor (under the desk in the foot well.) Also, the space under the riser can be used for stowage of the telegraph keys and the mike. Rebuilt the Power Cable with NK-M8-23C 11/16" connectors on both ends. Due the fairly large size of the connector pin receptacles, I had to use sleeving made from a rubber-jacketed, shielded audio cable that was the correct size. Tested for continuity and shorts with no problems found. The Power Cable is ready to install and use. Tested ATD/Dynamotor operation together and both performed as expected.  9/9/25

Building the Pilot's Remote Cable - The connectors are SK-C16-21C 3/4" which have 16 pins total, 2 are larger pins and the remaining 14 are smaller pins. This cable I built uses two 12 gauge wires (+28vdc and Chassis-Ground) and 11 smaller gauge wires (18 gauge) and one shielded small diameter audio cable (pins 1 and 4 aren't used.) Within the ATD transmitter circuitry, the wire going to pin 12 of the REMOTE box connector is a shielded cable since this is the "ring" connection (microphone) to the MIC jack. If only a carbon mike was to be used, shielding wouldn't have been necessary (low Z and also a DC bias voltage riding on the wire) but the ATD was designed to also use two types of dynamic mikes, a face mask dynamic Type MC-253 or an unspecified type of dynamic throat mike. Both of these type of dynamic mikes were probably around 250Z and utilized a diaphragm attached to a voice coil suspended in a magnetic field to generate a small voltage with the movement of the diaphragm. Dynamic mikes are usually prone to hum pick-up and noise pick-up unless fully shielded (mike body and cable.) If the cable drawing in the manual is studied carefully it will be plainly indicated that the original cable contained an individual shielded wire for connection to pin 12 with the shield grounded to connector pin 2 (chassis-gnd in the transmitter.) I had to be sure to include an audio-type shielded wire in the cable bundle.

The copper braided shield was harvested from RG-8U coax. I cut the braid at about 12 feet in length since the cable bundle was slightly larger than RG-8 coax. The braid can be "pushed together" to expand its diameter to allow it to easily "slip over" the wire cable bundle (it helps to tape the bundle at about 1 foot intervals to keep all of the wires together.) Once the braid-shield was entirely over the complete wire bundle it was pulled tight and then trimmed to the proper length (only about 4" was trimmed.) Next, a wrapping of black electrician's tape was installed over the braid for protection and for appearance. Then the next step was to prep the connector receptacles and prep the wire ends. One problem was how to code the wires since I only had blue and green 18 gauge hook-up wire. I used the five blue wires for the five pins that control the channel select function. The six green wires were assigned to the other functions. Although it's time consuming, the positive identification method is to solder the wires to the first connector as planned. Then, before soldering the second connector, an Ohm meter is used to verify each of the blue and green wires and then "tag" each wire with its pin number to allow a correct connection when soldering. I had a booklet of small stick-on numbered wire identification labels that I used. I only had to ID the blue wires, the green wires and the two 12 gauge wires.

With 14 total wires soldered for each connector, one does have to be careful and use of sleeving certainly helps (sleeving is called out in the manual instructions.) A small "pencil tip" type of soldering iron had to be used and a bench vise to hold the connector while soldering helps a lot. I start at the pins furthest away and work forward. In the particular case of these connectors, it's easiest to have the connector in the bench vise so that the two large 12 gauge wires are installed first. This places pins 13 and 15 furthest away (or at the top) and allows the remaining 18 gauge wires to be easily soldered. But, the pin identification is upside-down, so that has to be taken into account when soldering the wires (I made a drawing of the back of the connector "upside-down" with the pins numbered to help as a guide.) I supported the cable from above so that it "hangs down" to the connector in the vise since this made soldering the individual wires to the specific terminals much easier. With the pre-tinning, only a minimal amount of solder is needed to install each wire. I had to install the sleeving before soldering and then push the sleeve down the wire and over the connector pin. Building a ten foot long, 14 wire, shielding cable is tedious work. It took several hours to complete the cable. Each connector takes about an hour to install.

Once the cable was finished, I double-checked it with an Ohm meter to confirm that all wires were connected correctly and that the shield also was continuous from connector to connector. I then installed the cable to the ATD and to the Remote. I switched the ATD to REMOTE, switched on the +28vdc 80A power supply and used the Pilot's Remote to turn on the ATD. I checked the BATT V to confirm the ATD was powered up. The channel selector had been on "A" on the Remote, so the ATD channel was motor-driven to "A" (the ATD had been on "C.) I selected channel "C" using the Pilot's Remote and the ATD channel drive rotated to channel "C." I had the Shure 102C plugged into the MIC jack on the Remote, I pressed PTT and the Dynamotor turned on and the ATD was ready to transmit (I had the Collinear Array already tuned.) Speaking into the 102C, I saw on the monitor 'scope full modulation was taking place. I tuned in the ATD signal on the Reception Set R.107 with no antenna connected and using 'phones. No distortion was heard. I then plugged the 102C into the ATD MIC jack to listen for any difference,...no difference heard. I switched the 102C back and forth a couple of times to be sure there wasn't any discernable difference when using the Remote MIC input or the ATD MIC input. Also, with the ATD powered-up, the Indicator on the Remote was protruding as it should be. I checked CW operation using the Remote key and operation was as expected. This completed the testing of the Remote. I seems to be functioning as expected.   9/18/2025

I mounted the Remote on the right side of the desk opposite from the side that the ATD is on. This required that the cable be at least a ten feet in length. Exiting the left side of the ATD on the left side of the desk, the cable had to be about 36" long to make it to the floor. Then 60" distance under the desk and another 24" up to the Remote on the right side. Total distance was 120" or ten feet. Also, I had to install a small wooden block to the desk corner post to allow mounting the Remote so that it faces forward on the far right side of the desk next to the upper right desk drawer. The Remote doesn't interfere with opening or closing the desk drawer.

Rebuilding the Remote Indicator Cable - Only five wires are needed for this cable and since only one indicator operates at a time, the current requirements are minimal. Using 18 gauge wire is more than adequate and will allow easy "pulling" of the five wires through the flex-conduit. Shielding is unnecessary for any RFI purposes since only +28vdc is routed to all four indicators and then the selection is made by grounding the desired line to operate the specific indicator depending on the ATD channel selected. However, the metal flex-conduit does provide shielding for the Remote Indicator wiring. The five wires were "pushed" through the flex-conduit after the original wires were removed. The original wiring had been modified to only allow channels B and D to "indicate" and it appeared that this cable had been that way since it was originally built for a specific aircraft installation. The connectors were very dirty and had to be thoroughly cleaned. The original wires unsoldered easily. This cleaning left the two connectors ready to use. These connectors are Cannon WK-6-21C 5/8" for the straight exit and WK-6-23C 5/8" for the right-angle exit.

The flex-conduit is six feet long, so "pushing" modern plastic insulted wire through the conduit is quite easy. I had to make these wires somewhat longer so I could solder the ends to the connectors. The connectors have a special fitting at the exit point to allow a captive nut on the flex-conduit to thread to the rear of the connector fitting. The first connector's five wires were soldered, sleeving installed and then the wires pulled fairly tight from the opposite side and the connector tighten onto the conduit. The second connector had to have only enough wire exiting the conduit to allow soldering the wires to the connector. I used an Ohm meter to identify the green wires going to pins 2, 3 and 4 with dots applied to the insulation with a black Sharpie pen. The green wire going to pin 5 didn't require ID since its identity was obvious. Same with the blue wire going to pin 1. Pin 6 was unused. Once the wires were soldered and the sleeving installed, it was necessary to push the wires slightly into the conduit and the connector assembled and then screwed to the conduit. I used an Ohm meter to confirm all connections were correct.

Although the flex-conduit is rather "stiff" it can be bent to allow proper routing of the cable for installation. Since the flex-conduit is only six feet in length, the Remote Indicator had to be placed fairly close to the ATD, as it was in the rear cockpit of the airplane. The function of the Remote Indicator is very limited and only shows from another physical position what channel has been selected on the ATD. Although I'm sure it couldn't have been heard in the airplane when it was aloft, the indicators do make an audible "click" once the channel selected is finished motor-driving into position. I think the indicators are fairly difficult to see but it probably depended on where in the rear cockpit the Remote Indicator was mounted.

I mounted the Remote Indicator on the wall to the left of the ATD transmitter. It's visible from the operating position but as I mentioned, it's difficult to tell which button has "popped up" unless I get up and look at the Remote Indicator from the side. Where the Remote Indicator was placed was limited because of the lack of flexibility in the conduit and its relatively short length of six feet. It's an interesting piece that really doesn't do much of any importance to the operation of the ATD. In fact, it's sort of Rube Goldberg as far as remote channel indicators go. However, it does operate as intended and is an example of what was used with the ATD during WWII.

Cannon Connector Part Numbers Necessary for the ATD Transmitter

Battery to Dynamotor - GK-C3-21C 5/8" (straight exit) or GK-C3-23C 5/8" (right angle exit) - only one connector required with the right angle type being normally used.

ATD to Dynamotor - NK-M8-21C 11/16" (straight exit) and NK-M8-23C 11/16" (right angle exit) - standard was to use one of each type with the straight connector at the ATD end

ATD to Pilot's Remote - SK-C16-21C 3/4" (straight exit) and SK-C16-23C 3/4" (right angle exit) - standard was to use one of each type with the straight connector at the ATD end

ATD to Remote Channel Indicator - WK-6-21C 5/8" (straight exit) and WK-6-23C 5/8" (right angle exit) - standard was to use one of each type with the straight connector at the ATD end
 

 

Adding a Linear Amplifier Produces 125 Watts Output in AM mode

Sometimes the ATD's 40 watts out doesn't seem to be enough power on 75M,...even using the Collinear Array antenna. I do have the "W6MIT 95" homebrew linear amplifier that could be used to increase the RF power. I have a high-quality 3db attenuator good for about 30 watts. The linear amplifier is limited to about 18 watts input because of the 100% duty cycle of an AM signal and that results in about 130 watts output (25% of maximum output capability.) Since less than 20 watts has to be dissipated, the 30 watt rating of the 3db attenuator should be sufficient. Like most high power attenuators, the 3db attenuator uses male and female N-type connectors. The input to the Linear Amp uses a BNC connector. Connector-adaptors are necessary for the RF output cable from the ATD to the attenuator and from the output of the attenuator to the Linear RF input. The coax from the ATD is about 10 feet long so I'll need about another 10 feet of RG-58 to run to the attenuator/Linear combination. The Linear already has the proper coax to route its output to the Johnson Match Box/Collinear Array antenna system.

The other problem is how to actuate the Control Line of the Linear simultaneously with the ATD PTT. I thought about using the PTT line accessed at pin 11 of the REMOTE where the PTT "floats" at +28vdc and is taken to ground when PTT is activated and using that to actuate a relay to close for the Linear Control Line. However, the acquisition of the ATD Remote Controller has now made accessing pin 11 impossible. The alternative method is to use a "manual control" to actuate the Linear. I'll be using a foot-switch to actuate the Linear in sequence with the PTT from the ATD. Since the ATD would begin its "power up" with the PTT and it takes about 1 second for the dynamotor to achieve full speed, simultaneous actuation with a foot-switch with the ATD PTT should be rather easy.

A ten foot length of RG-58 coax was connected from the ATD to the input side of the 3db attenuator. The output of the attenuator was then routed to the RF Input on the Linear. The RF output of the Linear was connected to the antenna system. The foot switch was connected to the Control Line on the Linear. The foot switch can be actuated just slightly after the ATD PTT is pressed. I had to slightly reduce the ATD power to about 35 watts output for about 17.5 watts (3db drop through the attenuator) going into the Linear. The Linear's RF output was then 125 watts. Due to the slightly reduced power at the ATD, the audio modulation level is quite good and is reproduced well by the Linear. At this power level most stations on the mil-rad net copy the ATD Q5, even in marginal conditions.



The W6MIT 95 Linear Amplifier uses a Eimac Y-826 (HD 3-500Z) as the RF Amplifier
 

 


The ARB with a ZB-3 Homing Adapter mounted on top

The ARB Receiver - The Official ATD "Mate"

For most WWII installations in two-seater aircraft the standard receiver that was paired with the ATD transmitter was the RCA ARB Receiver. The ARB was a six tube superheterodyne receiver that was an updated version of the earlier, mid-thirties RU receiver series. The ARB receiver simplified the hook-up and dramatically improved the overall performance by replacing the RU's TRF-Tracking BFO circuit with a superheterodyne circuit. The ARB receiver tuned from 195kc up to 9.05mc in four bands. Two separate, two-stage IF amplifiers were utilized with 135kc used in the 195kc to 1600kc range and 915kc used from 1.6mc to 9.05mc (a dual frequency BFO was also required.) The lower two bands could be set up for loop operation, specifically for Homing DF purposes.

To simplify the external power hookup, the aircraft +28vdc buss was connected to POWER for tube heaters and inside the ARB was a dynamotor that provided the +230vdc B+. It's unfortunate that the ARB wasn't ever designed for nor was it ever intended to be used as a "ham receiver" (especially in the AM mode.) Its IF passband was purposely designed to be very broad for "Homing" use. Even in "SHARP" the IF passband is at least 10kc wide. The ARB has good sensitivity but a sensitive receiver with no selectivity makes the ARB very difficult to use "on the air" in the AM mode. That's why I use another type of USN WWII receiver, the Hammarlund-built RBG-2.


The "USN" Set-up - While the RBG-2 isn't an airborne receiver, it is Navy. It was used at USN land-based communications and relay stations during WWII. It wouldn't normally have been paired with the ATD. The RBG-2 is a superb mil-rad receiver with excellent performance specs.
BUT,...just look at all of those knobs and switches on the ATD transmitter,...it's hard to believe that the ATD has a total of 50 controls and switches,...10 in each Tuning Module (6 on the panel and 4 inside) for a total of 40 module controls. Then the other panels have 8 switches with 1 mike switch and 1 sidetone volume switch on the inside.

 

 

Return to Home/Index

 

 

Radio Boulevard
Western Historic Radio Museum

 Vintage Radio Communication Equipment Rebuilding & Restoration Articles,

 Vintage Radio History and WHRM Radio Photo Galleries

1909 - 1969

- 60 years of Radio Technology -

 

 

This website created and maintained by: Henry Rogers - Radio Boulevard, Western Historic Radio Museum © 1997/2025