Rebuilding the R-390A Receivers
Brief History - Assessing your Receiver - Disassembly
The Main Frame - The RF Module - The IF Module
The AF Module - The Power Supply Module - The PTO
Restoration - Alignment -
PART 1 - History, Assessing your Receiver, Main Frame, RF Module, IF Module, Audio Module, Power Supply, PTO, Front Panel Restoration, Meters
PART 2 - R-390A Other Details, Contractors List, Receiver Alignment, Expected Performance, R-390A Diversity Operation, Rebuilds and Restorations: Building a R-390A from Spare Parts, (2) 1967 EAC Restorations
PART 3 - Recreations of Famous R-390A Variants: Arvin Industries R-725, Army Security Agency/NSA R-390A, Clark AB/NSA Black Anodized Panel R-390A, USMC Olive Drab panel R-390A
PART 4 - R-389 Restoration - WARNING! This write-up contains "Extreme OCD" detailed information and it's long, Dynamotor R-648/ARR-41, R-392 Info, Other R-390 Variants, Security Dial Cover, CV-979 and CV-979A Cabinets
The anodized Black Panel R-390A receivers were used at Clark AB in the Philippines. It's likely they were also used elsewhere by military security agencies under the NSA. These R-390A receivers also have the Microdial installed on the BFO for easier adjustment for the reception of Dual FSK transmissions. More details in the text below.
R-390A - Recreations of Famous R-390A Variants
Since original examples of these famous variants are very rare and seldom seen, about the only way one can come up with a functional example is to build it. That's what I did with the following four receivers. For me, the the project would begin when the special parts or special modules became available. If I then had enough other parts and equipment in the "R-390A parts collection," that allowed the "recreation" to proceed to its completion. The following are the details on "recreating" four famous R-390A variants.
Creating an Authentic Arvin Industries R-725/URR
I wasn't really looking for another project but when nearly all of the parts turned up in a purchase, well,...I couldn't help myself.
Finding the Parts - I received an e-mail from an audiophile-collector friend of mine asking if I'd be interested in purchasing all of his R-390A parts. There was a main frame with most of the modules, another RF deck, an Audio deck, PS deck, PTO and a front panel, all for $100. It sounded like a good deal so I went over and picked them up. When I got the parts home and closely inspected them I discovered that the main frame was a '67 EAC that had the R-725 mods installed. The main frame still had the Arvin Series 500 IF module installed. The Series 500 modules were built by Arvin specifically for the R-725/URR.
Essentially, the Series 500 IF deck is just like the IF deck used in the R-390. Six stages of IF amplification and no mechanical filters. The original R-390 IF deck used BNC connectors for input and output but the R-390A used BNC Junior connectors. The Series 500 uses BNC Junior connectors to match the R-390A and also the new versions performed any other changes necessary to make the Series 500 just a "drop in" conversion for the R-390A.
Among the other R-390A parts was a Cosmos PTO that had a ferrous metal shield installed around the outer shield-can. There was also a mod to the PTO that had an extra wire exiting from the PTO tube socket area. Another part that was included (but wasn't installed in the main frame) was a small chassis with a 25vac transformer mounted on top and a couple of resistors underneath.
Unfortunately, someone had severely damaged the R-725 main frame. One side looked like it had been hit with an axe. Some of the harnesses had been "chopped" to remove their Amphenol connectors. The front panel was missing. The Veeder-Root counter was missing. Luckily, the special added harness for the addition of the small 25vac transformer chassis was still present although it had been cut for some reason. At least the harness was all there but bifurcated.
I was missing the correct data plate since the original
front panel was missing from the junk R-725 main frame. In early
February 2018, I received a data plate for an Arvin R-725 from Moe
Sellali CN8HD/W9, in Chicago, who is an ardent R-725 enthusiast. Moe
told me that my Series 500 IF module should have a serial number
ink-stamped on the rear of the chassis. According to Moe, when Arvin
completed the R-725 mods to each '67 EAC R-390A, this was the serial
number that was stamped on the front panel data plate. My Series 500 was
stamped "074" so Moe sent me the R-725 data plate with "74" as the
serial number. >>>
photo left: The Arivn R-725/URR built from the 1967 EAC R-390A (originally SN: 974) with the installation of an Arvin Series 500 IF deck, the hum bucker chassis, the special PTO, IF output conx and the Arvin SN: 74 data plate. From the top, the most apparent R-725 addition is the Series 500 IF module. Note how the input and output coaxial cables connect to the mounting bracket for the Meter and IF Gain potentiometers. Also, note that the rear panel IF output requires a special right-angle coaxial fitting with the cable routed to J14 on the rear left corner. Also, the Amphenol power connector is turned 90 degrees from the standard R-390A IF deck.
Purpose of the R-725 Modifications - For Adcock Direction Finders - or - Was that just a Cover Story? - The usual purpose that is given for the R-725 mods was for compatibility with military portable direction finders that used four vertical antennae per installation along with three receivers. The DF system used went back to the Bellini-Tosi type of DF set-up that used two crossed loop antennae with a rotating loop inside to create a radio-goniometer. Bellini and Tosi had discovered that crossed loop antennae would "re-radiate" the signal they were receiving within the small field inside the antenna's space. The "re-radiated" signal retained all of the directional properties of the original signal and could be measured for varying signal intensity dependent on direction. The crossed loop antenna size didn't affect it frequency of operation allowing for reduction in the size of DF loops on LW. Of course, the original Bellini-Tosi system dated from around 1900 and the system was sold to the Marconi Company around 1907. By the early twenties, vacuum tube amplifiers were being added to increase performance capabilities of the DF antennae systems. The most common B-T DF systems used the crossed loops but some larger systems used the four-square vertical antenna system and a rotational loop (the goniometer) within the square. This system was developed by Adcock during WWI and because the connections to and from the four square verticals were underground it didn't respond to skywave propagation and allowed ground wave DFing over long distances. The B-T DF and Adcock systems continued to evolve and improve and the systems were used throughout WWII. During WWII, oscilloscope displays began to be used for direction indications. After WWII, larger DF systems continued to be developed up to the mammoth "elephant cage" antennae ("Wullenweber" was the actual name) that were over a thousand feet in diameter and consisted of several "rings" of circular antennae all working to provide accurate DFing over great distances and wide frequency spans. By the 1990s, most of these large arrays were becoming obsolete and nowadays most have been dismantled.
The mechanical filters used in the R-390A resulted in signal path phase shifts that caused errors to show up in the DFing electronics. When used with the four square antennas, the low frequency modulation added via the radio-goniometer interacted with the mechanical filters creating the error. Early versions of this DF set-up had used R-390 receivers and the radio-goniometer was located quite a distance from the receivers to reduce any interference. In the 1960s, the USAF wanted to reduce the size of the entire DF system so it could be towed around on a trailered hut. This meant the radio-goniometer had to be in the same room as the receivers. This was going to require some protection to certain receiver circuits. The R-390 had been out of production for several years, so the solution was to design the new portable system to use modified R-390A receivers that could be easily purchased. Arvin Industries was the main contractor with Servo also doing some rework. The modified receivers would have the Series 500 IF module, essentially a R-390 IF module that was slightly updated to not require any rework to the R-390A receiver it was installed into. That eliminated the mechanical filter phase shift problem. Additionally, with the close proximity to the radio-goniometer, a 60hz hum appeared on the PTO tube filament and that also interfered with the LF modulation of the DF system. A special "hum bucker" chassis was added to the receiver that essentially operated the VFO tube, the BFO tube and the 3TF7 Ballast tube on +25vdc. Also, a grounded ferrous metal shield was added to the PTO housing to prevent hum "pick up." Arvin bought new R-390A receivers in 1967 from Electronic Assistance Corporation and the modifications were installed and, when complete, the receiver was tagged as "R-725/URR." The tags will generally show Arvin Industries as the contractor but sometimes Servo will be encountered. The quantity of R-725/URR receivers needed by the USAF was fairly small (less than 300, according to Moe) and thus today the R-725 is seldom encountered. Contact number on the R-725/URR was DAAB05-67-C-2338.
However, was there another purpose that was the "real" reason that the R-725 was created? According to an article that appeared in Electric Radio in January 2006 by Chuck Teeters, there was a "top secret" purpose for the R-725 and the receiver "mods" were initially created for that "secret" project. The R-725 was a product resulting from the Cold War jamming that was common between the USA and the USSR. In the mid-to-late 1960s, there was a new system that was being developed called "Tropicom" that was an upgrade to the antennas and transmitters to improve HF communications for the military. The upgrades also included the incorporation of the "F9c" anti-jamming/crypto system. The F9c system used a spread spectrum transmission of digital noise and signal that ran through a digital encrypo-key generator that had 144 stages of looped-feedback that also fed through phase modulators to maintain proper phase relationships of the signal and noise. When used with a R-390A on the receive end, the phase changes in the mechanical filters interfered with the recombination process and the system didn't work. When used with R-390s with a standard IF amplifier circuit, the F9c system worked fine. Since the R-390 dated from the early-1950s, there was only a limited supply of those receivers still available and those that were available needed constant maintenance. The ultimate solution was to have new R-390A receivers built with new-build R-390 IF modules installed.
In order to keep the F9c project "secret," the actual use of the R-725 couldn't be known to those outside the project. Since there really was the Adcock DF system upgrades that really did need a non-mechanical filter type R-390A, the R-725 was directed to be built for the DF purpose only. However, those running the F9c project had the R-725 order quantity doubled and half of the R-725 receivers were procured for F9c use while the other half went to the DF systems. The secret classification stayed on with the F9c system and it was used for quite a long period with many upgrades over the years. So, even though half of the R-725 receivers were used in direct finders, the other half had a "secret life" used in the anti-jamming/crypto communications world of the NSA, USAF and the Signal Corps.
Testing the R-390A with a Series 500 IF Module - With the donation of the Arvin R-725 data plate it looked like I had all of the parts to build-up a R-725 if I could supply a complete 1967 EAC R-390A. According to Moe, when Arvin built-up the R-725 receivers they purchased new '67 EAC R-390As direct from EAC to fulfill the contract, thus all Arvin R-725s are converted '67 EAC R-390A receivers. I decided to use my '67 EAC SN: 974 R-390A because this receiver had recently been partially "cannibalized" to complete another EAC R-390A. I needed to replace a defective RF transformer on the 2-4mc antenna stage and do some minor alignments. Luckily, the "junk" R-725 RF deck supplied a good RF transformer. The first step was to check out and test the Series 500 IF module. One of the IF transformer cans was severely dented and needed "body work" to correct. All of the tubes were missing. I checked over the underneath and all components appeared to be in good shape. I gave the Band Width switch a DeOxit treatment. I needed tubes and tube shields. I found all of the tubes in my tube storage. The shields were "borrowed" from the EAC IF deck as was the 3TF7. The Series 500 is a "tight fit" but it does fit (see above photo.) The chassis is somewhat longer so the captive screws are located on the chassis rather than on the flange. The Band Width and BFO shafts are shorter than on the standard IF deck. The input and output coax connectors are in a different location but the cables reach easily. There is no clearance for the rear IF output cable as it is directly behind one of the 12AU7 tubes. The junk R-725 main frame even had the rear IF output connector totally removed. A special connector is required for the IF output on the R-725 conversion. The Amphenol connector has to be turned 90 degrees but everything lines up and there is ample flexibility to allow for this connection.
With power applied, everything came up as expected. The first thing noticed was that the IF Gain must have been at "maximum" - it was. After some testing and listening, I reduced the IF gain by about 50 percent. This provided ample IF gain and much lower noise levels. Carrier Level was adjusted on 15mc to zero with the antenna disconnected. BFO was zeroed. I didn't do a 455kc IF alignment since this was just a "check out" but the IF deck already seemed to be performing better than expected.
Mounting the Hum Bucker Chassis - Mechanically, the "hum bucker" is mounted in front of the R-390A power supply. This requires a bracket with pem-nut on the PTO side plate and two holes on the outer side panel to mount the "hum bucker" chassis. I removed the PTO side plate from the "junker" R-725 main frame because it had the original bracket already mounted. I removed the original PTO side plate from the R-390A and installed the R-725 side plate in its place. I carefully measured the original "junker" R-725 main frame side panel for the correct location of the two mounting holes. I then marked and drilled the R-390A side panel in the original manner. These modifications allowed the "hum bucker" chassis to mount exactly as it did in the original R-725. Now the "hum bucker" installation was electronically and mechanically complete. On to the PTO next.
Testing and Calibration the PTO - I'm using the original Cosmos PTO from the junk R-725 main frame. This PTO already had the mod installed that lifted pin 3 of the VFO tube from chassis. Then a wire was connected to pin 3 and it was routed back to the PTO connector where it is connected to the unused pin C. Also, a .01uf ceramic disk was installed from tube socket pin 3 to chassis. The PTOs that were used in the R-725 had a ferrous metal shield installed over the can of the PTO and this PTO did have that shield installed. I have a R-390 PTO test fixture that was given to me by W6MIT. The test fixture allows powering the PTO and employs a digital turns-counting dial to accurately set the end-point error to <0.5kc. I had to supply +195vdc B+, Regulated +150vdc, 6.3vac and chassis ground. Output was measured from the coaxial cable of the PTO using a digital frequency counter. I used a Lambda 25 for the B+ and 6.3vac and a regulated +150vdc supply. With the PTO on the fixture and powered up, the first step was to adjust the PTO output to 2455kc, then set the counter to 00.0 and tighten the coupler. The fixture counter works the opposite to how the PTO functions in the receiver. Since it's a mechanical readout on the drive rotation it doesn't really matter and our actual check was to verify that the PTO output changes from 2455kc to 3455kc in exactly ten turns. A quick check revealed that the end-point error was close to 1.0kc. I ran thru each turn to check linearity and this PTO was "right on." If it had been necessary to adjust the PTO end-point I would have followed the procedure as detailed in the PTO section further up this web page. To install the PTO only requires that it be set to 3455kc output with the R-390A having xx.000 on Veeder-Root counter. When the R-390A Veeder-Root counter is set to xx.000 then the Oldham coupler aligns correctly. The power connector is installed and the output coax connected to the RF module. This completed the PTO modifications and, in fact, completed all of the R-725 mods necessary.
Installing the Special Right-angle Coax to BNC fitting for IF Output - If it's attempted to fit the original IF output coax cable onto the original coax box BNC Jr to BNC output fitting, it will become obvious that there isn't enough clearance due to the 12AU7 tube directly in front of the connector. For the R-725, Arvin replaced the rear panel BNC Jr to BNC connector with a special mini coax input at a right angle to BNC output connector. This "low profile" fitting provided enough clearance to then connect the exiting cable to J-14 which is the IF Output on the Series 500 IF module.
To install the right-angle fitting requires a slight enlarging of the mounting hole which Arvin apparently did by filing the hole until the connector fit (I did check the junk main frame and it showed evidence of filing.) The coaxial cable should be installed onto the connector first. The center conductor of the coax is routed through the right-angle tube and the shield is placed over the outside of the tube. The center conductor is soldered to the center pin making sure the teflon spacer is installed afterward. Then the crimping barrel is placed over the shield and right-angle tube and crimped in place. Then the back cover nut is installed. The BNC right-angle fitting with coax attached can then be mounted to the rear panel of the receiver with a locking washer and nut. Then the BNC Jr. end connector can be attached to J-14. This completes the installation. Thanks to Moe CN8HD/W9 for supplying the correct coaxial right-angle fitting (as mentioned, my coaxial fitting was missing from my "junk" R-725 main frame.)
Testing the R-725 - The R-725 mods were for a DF set-up (or for Tropicom) so the changes to the PTO tube, BFO tube and 3TF7 tube filament supply are very subtle and not noticeable by just listening. However, the "big change," that is, adding the Series 500 IF module and thus eliminating the mechanical filters and adding more IF stages,...that is very noticeable. In fact, it's impressive! The gain is amazing. I have the Series 500 IF gain set to 50% and the strong signals will still send the Carrier Level meter to +80 or +90 db. If I tune off of the signal, the meter drops to +10db. The selectivity is still very good. Just about as good as mechanical filters. The R-725 has become my standard "vintage military radio receiver" and it's set-up out in the shop with the T-368 transmitter and using a two half-waves in-phase antenna,...aka, a 275ft center-fed dipole operated on 75M. The R-725 can always be counted on the "pull in" the weak signals and is easily able to cope with any QRM. Audio quality is good and sounds pretty close to a typical R-390 receiver. Probably one could sum up the R-725 as an "easier to work on R-390" with all of the benefits of the R390 without as many headaches.
Wrap-up - Well,...what is it? A restoration or a recreation? I was extremely careful to use authentic R-725 parts harvested from a "destroyed, incomplete" R-725. I was very careful to exactly duplicate how the wiring harness was integrated into the R-390A harness. I even used the original hum bucker harness for authenticity. Original R-725 sheet metal was used where needed. Even the receiver used for the conversion was a 1967 EAC R-390A. The data plate used was an exact copy, etched tag - not a silk screened tag but one made just like the originals. Even the serial number stamped in the tag matches the serial number ink stamped (in 1967) on the back of the Series 500 IF module. And, all of the R-725 parts came from the same "destroyed, incomplete" R-725 which must have been the original SN: 74. So,...when looking over this R-725/URR,...I consider it an authentic restoration of SN:74. Just that the original R-725 SN:74 was modified in 1967 and my modification/restoration was performed in 2018. Close enough,...right?
New for 2022 ! Recreating a U.S. Army Security Agency Motorola R-390A
Recreating a Clark AB "Black Panel" aka NSA R-390A
Refurbishing the Recreated 1961 Capehart USMC OD panel R-390A
General Info about the Receivers - I've always been intrigued by the clandestine, crypto, signal security types of reception use that the military and government agencies like the NSA had for the R-390A. I knew about the Clark Air Base use of "black panel" R-390As and I had seen some of the BFO Microdial installations on other R-390A receivers. I even have a "parts set" Motorola R-390A that has the remains of a Microdial and also U.S. Army Security Agency (ASA) stickers on the rear panel. The opportunity to build up a black panel Microdial R-390A presented itself with obtaining an original "grained and anodized" black R-390A front panel. Where the panel originally came from was an unknown but it appeared to have been stored for quite a period of time. At one point someone had installed some incorrect switches and pots probably to have a place to mount some of the knobs. The panel also came with both of the correct meters. I had several R-390A "parts sets" and a good supply of various modules. However, for the Clark AB Black Panel I decided to use the 1967 EAC receiver that I had gotten from Ham & Hi Fi in 2017 that still had a few problems after its refurbishment (described in Part 2 in "Not Another 1967 EAC Restoration Log.") The receiver was out in the shop installed in a rack with two other R-390A receivers. I had really forgotten exactly what the final reworking had consisted of. I knew I had robbed some parts and exchanged some others. So, first I had to test this R-390A and correct whatever I had done to it. >>>
|>>> July 6, 2022 - In reading what I wrote in 2017, I really didn't know what the status of this receiver was now in 2022. A few years ago I installed this receiver into a CRAIG relay rack that was out in the shop. Then the R-725 project came along and that required the other EAC R-390A receiver I had. The R-725 works great, but I read what I wrote about that project and it doesn't mention what parts I robbed or swapped from this R-390A. I must have left the good PTO in this receiver as mentioned in the prior write-up. I wouldn't have needed it since the R-725 uses a special PTO that I had for that project. Next was the data plate. I mentioned in the prior write-up that I installed a repro data plate but upon inspection this receiver now has an original EAC data plate installed (SN: 974) from the first EAC receiver (that was converted into the R-725.) Since the R-725 has a special data plate, which I had, I must have removed the repro data plate and installed this original one in its place.|
|July 7, 2022 -
Functional Test Results - I performed a functional test which did verify that the
"good" PTO was still installed in this EAC receiver since the tracking was
excellent and the EPE was around 1kc or less.
I performed a reception test on WWV 10mc and got a CL reading of about 50db. The CAL would drive the CL meter to 40-50db. All bands functioned.
The test on 2-4mc found that 2mc CAL signal was about 40db and 3mc was about 20db. This indicates that I really didn't do anything about this problem other than the temporary substitution of the 2-4mc RF transformer to confirm the problem. I must have put the good RF transformer back into the other EAC receiver that became the R-725. So, the low signal performance on 3mc is still a problem that needs to be corrected. I have many spare 2-4mc RF transformers so hopefully I have at least one good one. NOTE: These test results changed when the test was carefully set up. The test results really changed when I went through all of the tubes finding that many were marginal, one was shorted and one was erratic acting. NOS tubes corrected this "problem" along with several other minor issues.
July 8, 2022 - I ordered a Servo Instruments 1309 Counting Dial. This is the type of Microdial that was used on the various security agency receivers and the military crypto, signal security, intercept receivers. Some receivers had a turns-lock that limited the rotation of the BFO shaft. It incorporated the lock into a replacement panel bushing. The 1309 Counting Dial unfortunately is just the Microdial and no turns-lock is included.
July 9, 2022 - Checking What R-390A Parts I Had - I inventoried all of the R-390A parts that I have. I was curious about the 1961 Capehart with the Olive Drab front panel. I had disassembled that receiver over ten years ago and never put it back together because I didn't have a set of meters for it. The spare black panel came with proper meters that can be used to complete the OD Capehart. As it turned out, I found all of the Capehart modules but unfortunately I had robbed Z-503 off of the IF module (as described in the proceeding '67 EAC write-up in Part 2.) I had several IF modules that probably can provide another Z-503 but that will be rework necessary when I get to the OD Capehart part of this project. At least I still have all of the original modules and that receiver did work at the time of disassembly but that's no guarantee on the electrolytic filter capacitors. Also, I seem to remember that the Capehart AF module had a bad transformer on one of the audio outputs.
All total I have eight "parts sets" in
various states of completeness (or lack of completeness.) The best
condition, other than the OD Capehart was an Imperial that looked pretty
good, certainly restorable but was, of course, minus the meters. There was a Motorola with US
Army Security Agency stickers on it and the base of a Microdial on the
BFO but the overall condition was very rough with a "beat up" front
panel and only the RF deck and
AF module present. There were a couple of rough
receivers that were missing most of modules and their data plates. Lots of
"spares" for RF decks, PS modules, Audio modules, PTOs,
and front panels. I had
four IF modules but I don't think any of them are in great shape. One is
a Collins IF module with the fixed-tuned mechanical filters (1956 MWO never
incorporated.) So, lots of parts available and I should be able to do
the Black Panel receiver and the OD Capehart without too much
trouble. It's also a possibility, if I can purchase another 1309
Microdial, I could create a US Army Security Agency receiver from my
fully restored 1956 Motorola R-390A. Also, the ASA stickers would have
to be salvaged or reproduced using the originals on the "parts set" ASA
Motorola. And,...that Imperial is looking better and better every time I
look at it,...It has a full set of knobs, the RF deck is complete, I
have a PS, an IF module and AF module (all Amelco) plus a near-mint
front panel,...perhaps if another set of meters shows up,...another
Recreating an "Army Security Agency" Motorola R-390A
July 11, 2022 - Microdial - Received the Model 1309-100 Microdial today. Original mounting had a turns-locking bushing that only allowed a specific number of rotations, three being standard. I didn't have one of these special bushing so I'd have to use the regular BFO panel bushing to mount the base. The depth of the dial knob mounting on the shaft controls the mesh of the internal gears. Works well. But, I checked the "parts set" Motorola that still had the Microdial base mounted and that base was glued to the panel. With close examination and I could see slight traces of the glue which looked like the old (really old) brown "contact" cement. When I get to the point of actually mounting the Microdial, I'll find out if the panel bushing mounting will work,...it should. But,...if not, I guess that might be why the Motorola Microdial was glued to the panel. At any rate, I ordered another 1309 Microdial today (one for the Motorola ASA R-390A project and one for the Clark AB Black Panel.)
July 12, 2022 - Testing the Motorola R-390A - I did a functional test on the Motorola R-390A today. It had been probably a year or so since it had been powered up but that wasn't a problem. After all, I completely restored this R-390A in 2011 and that included a replacement of all capacitors in all of the modules. In fact, most of the photographs in Part 1 of this R-390A write-up are of the modules from this receiver. Back to testing,... I connected up a 600Z loudspeaker and the two half-waves in phase antenna out in the shop. Lots of signals on all bands. SW BC signals on 12mc pushed the CL meter up to +60db. I tuned in XSQ on 16.8mc (about 1500hrs PDT) and it was quite strong. XSQ is one of the coastal stations in mainland China. They send out a CW ID after a frequency warbling that is sent four times. This is continuously repeated. These coastal beacons are located in many countries but the China and South Korea beacon stations are particularly strong here in the West. The Motorola also received a number of 20M ham signals so I think the Motorola is working quite well.
July 13, 2022 - Removing the ASA MWO Labels - I was able to remove the three U.S. Army Security Agency MWO labels off of the back of the "parts set" Motorola R-390A. One label was in rough condition after the removal. The second label came off fine but was worn through in the center where something had rubbed against it in the past. The third, which had actually been under the first one removed, came off in good condition. I had tried to remove the first label by wetting it but that tended to damage the already fragile paper. The other two labels were removed using a razor blade and this worked quite well with no damage occurring during the removal process. The third label, being that it had been underneath another label, retained its gloss and looked very nice. Also, the MWO number wasn't "scratched out" as the MWO numbers on the other labels had been. This glossy label will be the one I'll install on the new creation of an ASA Motorola R-390A. The label is shown in the scan above right for a close-up and the photo below right showing the label installed. Also shown is the metal military AC plug mentioned a few paragraphs down.
Mounting the Microdial Details - Since changing the Motorola R-390A into an ASA-type isn't very involved, I started that project first. Installing the Model 1309 Microdial requires first removing the IF module. Then slightly enlarging the BFO panel hole by about .060" (out to .50") to allow clearance of the embossed centering ring on the 1309 base. The enlargement of the hole diameter shouldn't be through the panel. Only enlarge the diameter about halfway through the panel (it only needs clearance for the base centering boss.) I didn't have the special turns-lock bushing but I was able to use the standard panel bushing for mounting the base. Best results would be to use an original turns-lock or to use a slightly longer panel bushing, maybe about .125" longer to allow using all of the washers. For now, I just mounted the 1309 base with the bushing, a flat washer and the mounting nut and that seemed to work fine.
Details - The 1309 Microdial has to be taken apart to mount it.
The base is
mounted first. The embossed ring will assure that it's centered. Insert the 1/4" BFO shaft
and clamp from
the rear. Insert the counter knob and then install the outer mount/turns counter. The outer mount is held in place
on the base with three screws
around the perimeter at the rear. I set
the Microdial to 20.0 since the desired BFO adjustment was +/- 13kc and
that happens with about 1.25 turns of the BFO shaft each side of zero
beat. So, with my set up 20.0 is 455kc on the BFO and 32.0 would be
about 468kc and 08.0 would be 442kc. Now
slide the counter knob slightly forward and tighten the two set screws.
Try the dial, it should adjust smoothly (it isn't coupled to the BFO
shaft yet.) Next, loosen the set screws and remove the turn counter then reinstall the IF module
and connect the bandwidth shaft and BFO shaft to the IF module shafts.
Tighten the shaft clamp on the Bandwidth. Remount the turns counter with
the extension shaft already on the BFO shaft but the clamp not
tightened. Now, connect the IF module to the R-390A via
its cables and power up the receiver. With the BFO turned off, tune in a
CAL signal and adjust the frequency dial for exact center frequency.
Turn the BFO on and adjust the BFO shaft from the IF module with your
fingers for exact zero beat and then tighten the shaft clamp and the
turns counter set screws. Try
adjusting the BFO with the Microdial. If the mating of the counter knob
and the turns counter/outer mount is correct, the tuning will be smooth
and have a nice "feel" to the adjustment (it should feel
fairly firm but not binding in any way.) Test by adjusting the receiver
tuning to +13kc above the CAL signal, then adjust the Microdial
and you should hear (barely) the BFO zero beat with the tuned frequency. Repeat
for -13kc from the CAL signal and you should hear the zero beat with the
BFO at about 08.0. Don't expect this to be "dead on accurate" but it should be
IMPORTANT NOTE: Remember that the Microdial is a turns-counter and +/-13.0 on the dial is actually about 1.3 turns of the BFO shaft. It's a coincidence that it also happens to be close to a +/- 13kc excursion either way from the BFO center frequency. I originally set the BFO Microdial zero beat at 13.0 but that ends up confusing the operator (me) into thinking the Microdial is a frequency readout when it's actually shaft turns. In fact, for typical SSB demodulation the Microdial only has to be +/-00.6 of zerobeat.
Mounting the ASA Label -
To complete this transformation, the U.S. Army Security Agency MWO tag was glued in place on the back panel.
Although there was still some of the original adhesive on the back of
the label it certainly wouldn't be sufficient. I used white carpenters
glue applied to the back of the label with a Q-tip. This should be a
thin and even coating. The label was placed on the back panel with the
receiver resting on its grab handles. This allowed using a plastic pad
and a weight to press down on the label to keep it perfectly flat while
the glue dried. After several hours the weight was removed and the label
installation looked quite original.
AC Plug - The receiver had a civilian-type black rubber AC plug installed and that didn't look quite right. I installed a military metal AC power plug to complete the change of this Motorola R-390A into an U.S. Army Security Agency receiver. Except, I did mount a spare security dial cover I had to the receiver. It seemed appropriate.
Operational Details - Tested the operation and the receiver performed well. The adjustment of the BFO is not as "fine" of an adjustment as one would think. Most SSB and CW will demodulate with the BFO set to +/- 00.6 turns from 20.0 on the dial or 19.4 on the dial for USB (I know it's hard to remember that the Microdial is reading BFO shaft "turns" not frequency.) The wide range of adjustment was for RTTY and DFSK signals. It's a nice addition though. Also, of some interest, I checked the serial numbers on this Motorola R-390A and the real ASA "parts set" Motorola R-390A. They are fairly close with the newly transformed ASA receiver being serial number 1873 and the "parts set" receiver being serial number 973 (exactly 900 units apart,) both receivers on the same contract from 1956. So, it's about as close as I can get to an original ASA R-390A receiver.
Recreating a Clark Air Base/NSA "Black Panel" R-390A
Front Panel Condition "as received" - Shown to the right is the black anodized front panel I'm going to use for the Clark AB/NSA R-390A. The panel has many minor blemishes and defects (perhaps the defects will add to the authenticity and character of this recreation.) It appears that the "graining" that was done to the panel before anodizing wasn't carefully done and there are several areas where the "graining" is counter to the horizontal direction it should have. The pin-point "spotting" might be surface problems that weren't addressed before anodizing. In fact, in some places it looks like the anodize and dying process was performed and the panel wasn't entirely stripped or maybe it wasn't cleaned properly before the dye was applied! The most likely reason for the defects would be that this was a somewhat recent anodize job that was done by an incompetent metal treatment shop. Who knows? Maybe it was. The second possibility is that this really is a vintage anodizing of a panel that was worked on to correct some physical damage problems that resulted in some of the sanding and grinding ultimately showing through the finished product with the final disposition being that the panel was rejected. That's why it was never installed on a receiver. The engraving fill paint is mostly present but it's a dark gray color. Some residual red and white paint can been seen in some places in the engraving and that seems weird since part of the anodizing process uses dilute sulfuric acid as an electrolyte. Physically, the panel is in good shape with only two small defects at the bottom right (and that's an easy touch-up) and the top right where the corner isn't square.
So, the first step will be to work with this panel to see if any improvement is possible. The panel was coated with oil or wax (maybe Armor All) that really enhanced the prominence of the condition problems. With a thorough cleaning, touch-ups and with properly filled engraved nomenclature, the panel should look in original "believable" condition and maybe it might even look like a vintage piece (and maybe it is,...it's real history is an unknown.)
|July 14, 2022 - Front Panel Restoration - Looking at the photo of the panel the engraving looks like it's filled with white paint but it's actually dark gray. To have a more "believable" fill paint I used a "manila" color that is a mixture of white, light brown and just a touch of black. Just using white paint or white lacquer sticks will end up with the nomenclature looking way too bright. Manila looks aged but in good condition. Unfortunately, all of the engraving had some issues. Most of the nomenclature had to be "gone over" with a special tool I made to "dig out" the old gray paint. However, on the right side of the panel, some of the engraving had some anodizing and black dye in the engraved areas that had to also be removed. Looking at the "before" photo, CARRIER LEVEL is very obvious that there's a lot of black in the engraving. It's a laborious process to remove but all of the engraving had to be cleaned in this manner so that the new "manila" fill would look correct. Also, on the panel I'm going to do the red fill that I've seen on a couple of black panel receivers. The CAUTION area, DIAL LOCK, ZERO ADJ., STAND BY and CAL will have the red fill while the remaining engraving will be manila color.||July 15, 2022 - Front Panel
Restoration cont. -
Continued on with the "manila" paint fill for the panel nomenclature.
Most of the engraving actually requires two coats of paint for the
proper fill amount. Although I've mentioned it in many other places in
these write-ups here it is again,...I use Artist's Acrylic to mix the
proper color. The paint is applied using a saturated Q-tip just to one
area of nomenclature at a time. Once the paint is applied, I allow about
two minutes for it to set up. Then I use a small piece of paper towel
that's dampened with Glass Plus and folded as flat as possible. I use
these types of paper towels to remove the excess paint. I only use a
paper towel piece one time, then it's discarded and a new one used to complete
the clean up. Sometimes a second coat is needed for a good fill. Once
the paint has dried for about 10 minutes it's okay to go over the area
with a slightly dampened paper towel to further clean up the residual fill paint.
This has to be done for each engraved area. Don't try to do a large area
with several areas of engraving - it won't work. Just do one set of
engraving at a time. It will take about an hour to do all of the panel.
Cut up the paper towel pieces in advance to help speed the process.
The five engraved nomenclatures that are going to be filled with red paint will be next.
|July 16, 2022 - Front
- I mixed red with just a small dab of black to dull down the color a
bit. I had to do this part of the fill carefully to be sure to not get
the residual red paint into the manila nomenclature. Only the STAND BY and CAL
are really close. As expected, the panel's appearance has improved
substantially with cleaning, touch ups and the engraving fill. I'm
pretty sure that once all of the components, knobs, hardware, etc., are
mounted, the panel blemishes should be much less noticeable.
I had to do the slight modification to the BFO PITCH control hole, that is, increasing the diameter of the hole up to 0.5" to allow for the embossed centering ring for the base of the Microdial. I also used a longer panel bushing for this installation and that allowed using all of the mounting hardware. Too bad I don't have the turns-lock assembly but it's not necessary and wasn't installed on all Microdial-equipped receivers.
Further prep on the back side of the anodized panel was required. Aluminum oxide is the ultra-hard coating that the anodizing process creates and it is non-conductive. But, so was the original paint. That's why there were two strips of bare metal where the back of the front panel mounts to the main frame. The bare metal assures that the front panel is grounded so it provides the shielding and ground returns necessary. Additionally, one of the AVC time-constant capacitors is grounded to the back side of the panel by the CL meter. This area also has to be bare metal. I'm not sure if the original anodized black panels were masked or just ground off as needed during installation. In this case, this anodized panel had not been properly prepared on the backside. I had to mask the two long strips for the main frame contact. Once masked, I used a Dremel tool to remove the aluminum oxide in only the areas needed. Same for the small area next to the CL meter.
|July 17, 2022 - Front Panel
Completion - I
had to use the Dremel drum shaped Al-Ox tool about 0.5" diameter. It worked
really well and fast to remove the anodize just in the area that I
masked (and I also "self-masked" with a N95 due to the fine powdery oxides
that became airborne in this removal process.) When the tape was removed the bare metal areas looked almost
I had broken one of the mounting screws that are located behind the data plate (mounts the resistor PC board.) I had to dismount the bracket and then drill out the broken screw. These are "undercut" FL 6-32 screws so I ended up robbing an original off one of parts set to replace the broken screw.
The meters have to be mounted before the panel is mounted to the main frame. Then the panel is placed face down in front of the main frame with two 16ga. wires mechanically supporting the panel running from the top hole in the panel to the top hole in the main frame on each side. This provides a solid support for the front panel as the parts are installed. All of the components were placed in their proper holes and the mounting hardware installed "finger tight" only. The three screws, lock washers and nuts that support the harness clamps to the front panel have to be installed. Now is the time to solder the wires that connect to the CL meter and the LINE LEVEL meter. At this point the panel can be put in place but I had to be sure to install the KC and MC inner shaft bushings and the bushings for Bandwidth and ANT TRIM. With the panel in place now all of the screws can be tightened, all of the control mounting hardware can be tightened and the knobs installed. I had to play around with the MC and KC bushing positions to achieve a decent feel to the tuning. At this point the dial bezel wires can be soldered and the dial bezel mounted to the panel. Finally, the data plate can be installed.
I installed the Microdial on the BFO. Then I installed the PS module and the AF module. I connected a 600Z loudspeaker and an antenna then AC power was applied. The receiver worked like it had at the original testing so there's a bit of work left to do for top performance. I did receive XSG 16.898mc about 569 at about 1500hrs. XSG is a coastal beacon in China.
|July 18, 2022 - Testing
- I calibrated the CL meter and then used the CAL oscillator as a signal
source. When checking 2.000mc the CL read 60db. When checking 3.000mc
the CL meter read 55db. When checking 4.000mc the CL meter read 55db.
This looked pretty normal. As a further test I checked 7.000mc and the
CL meter read 65db. Again, pretty normal. I tuned in WWV 15.0mc with a
tuned antenna and it read 65db on the CL meter. I tuned in a couple of
really strong utility signals in the 16.0mc range and most read between
50db and 70db on the CL meter. All of this looked pretty normal for a
typical R-390A that had been aligned a few years ago.
July 19, 2022 - Performance Analysis - I've been listening for a couple of days now and it seems this R-390A is working pretty well. I've copied most of the Chinese Coastal Beacons on 16.8mc. Copied KPH on 16.625mc on CW on a Saturday early afternoon. 20M hams and SW BC up in the 15mc region are present. There are a couple of weird minor issues though. First, when listening to SW-BC stations the audio seems low compared to the carrier. Second is the limited range of the RF Gain control. All of the RF Gain action is between 8 and 10. At 10, all signals seem normal on the CL but disappear if the RF Gain is reduced below 8.*
I wondered why the dial lamps were so dim. Well, #345 lamps is why. I thought that I had replaced these lamps but obviously I didn't,...weird. The #345s might be 6 volt lamps but they draw a lot less current than the correct #328 lamps. The high filament resistance of the #345 bulbs in combination with the series 2.7 ohms resistor results in the dim illumination. A pair of new #328 lamps installed and the dial lamps were up to the normal illumination.
Wrap-up - As I had hoped, once all of the components were assembled onto the front panel, along with good condition nomenclature, the blemishes are hardly noticeable. I guess if it looked perfect everyone would know it's a recreation. With the minor blemishes, it could almost pass for an original,...well,...maybe. Still,...it's very cool.
|*Follow-up - July 25, 2022 -
Performance Problem Solved
- I got to thinking back about when I went through this particular
R-390A in 2017 and I suspected that I either hadn't tested the tubes or
maybe I used a different tube tester (I had just gotten a really
TV-2 about that time.) I read though what I wrote in 2017
and apparently I did test the tubes and there were a lot of bad ones
that I replaced at that time, apparently from my stock of "used tested
good" tubes. I must have had a very low criteria for acceptance back then
(anything slightly above "minimum acceptable" was okay, I guess.)
The new test (five years later and virtually no hours on the
receiver) used the TV-7B that I've owned, calibrated and used for years.
The testing involved all 26 tubes. I found five tubes that were at or below minimum acceptable, one tube
that acted erratically and one tube
that had high leakage (shows as a short.) The test results were as
follows,...all tubes in the IF module
were in "tested as new" condition. The PTO tube tested "as new." The PS
rectifiers tested good (just an emissions test and above 40 on the TV-7
is okay.) The AF module had both
6AK6 tubes testing at minimum acceptable, the two 5814A tubes tested "as
new." The RF deck had two weak (below minimum acceptable) 6C4 tubes (the R-390A circuit pushes
tubes and they always seem to be weak in receivers that are in "as
found" condition.) Also,
the 5654 (6AK5) for the crystal oscillator mixer was the tube that was showing
a short. The RF amplifier 6DC6 tested erratically and was marginal when
stable. It was replaced with a NOS 6DC6 tube. The OA2 and the 3TF7
tested good. I was surprised at these
test results because I thought that I had been very thorough back in
2017 - but, obviously NOT! All of the suspect tubes were replaced with
either NOS tubes or "tests as new" tubes. With ALL good
tubes, the R-390A pushes the CL meter up to +75db when tuned to WWV
15mc. RF gain range is now 5 to 10 using the CAL+BFO on 6.000mc. The
Chinese Coastal Beacons on 16.8mc can be received with the RF gain at 6
and are still easily Q5 (AF gain at 6.) When listening to SW-BC stations
now, the signals have full modulation and the audio quality is very
good. With all "new" tubes I was able to reduce the IF Gain control to
about 75% advanced and WWV 15mc still reads +65db on the CL meter (no
signal CL is about +5db.)
IMPORTANT NOTE: This "problem" was a very good illustration as to why ALL of the tubes in the R-390A need to be in NOS or in "tests as new" condition. Although any R-390A will seem to function adequately with some marginal tubes installed, top performance requires that ALL tubes are in excellent condition and "test as new."
After the "bad" tube discoveries in this receiver, I thought I better retest the performance of 3.800mc compared to any frequency above 4mc to see if there actually is a loss in the 2-4mc range. Using the CAL and the CL meter I tested several frequencies and found that the CAL would push the CL meter up to +40db at the lowest and up to +60db at the highest. Listening to various bands and frequencies there didn't seem to be any obvious reduction in sensitivity. I'm going to assume that the original observation of low levels on the 2-4mc range was probably due to all of the marginal and defective tubes. Performance of the receiver is now just as expected for a recently aligned and serviced R-390A.
Additional Note: I've set up the Black Panel R-390A to function as the station receiver with the W6MIT 1625 rig as the station transmitter. The 1625 rig uses a T-368 exciter driving a pair of 1625s as the PA modulated by a P-P pair of 1625s as modulators. The transmitter allows using Break-in, so the receiver is protected during transmit. Antenna is a 135' CF Inv Vee with 99' of open feed line to a 1KW Matchbox. The Black Panel operates very well on 75M and also is used often as a SWL receiver on many other higher frequencies.
USMC 1961 Capehart R-390A with Olive Drab Panel
The R-725 and the two proceeding 2022 "recreations" used already complete and
refurbished R-390A receivers so those recreations were fairly simple and
only involved a few parts. This next receiver, the USMC OD Panel R-390A,
was entirely disassembled and the IF module had been scavenged for
parts. This "recreation" ended up being more like a refurbishment of a typically-found,
incomplete and "untested" condition R-390A. I used as many of the
original parts for the receiver reassembly as I could find in my "parts
collection." I had
recreated this USMC version R-390A receiver in 2010 but
I disassembled it in 2012 probably because of the move to Dayton, Nevada
in 2012. After the move, the OD Capehart went out to shop with the other "parts sets." Some
of the original parts were lost over the next ten years and other parts were
installed in other receivers. So, ultimately this project ended up using a lot of
"spares" from the R-390A "junk box"
and other parts coming from the collection of "parts sets." This
"re-recreation" ended up being a whole lot more involved than I originally thought it
OD Origin - Sometime around 2010, my old friend Mike W7MS told me about a Capehart R-390A he had seen a few years earlier that had an olive drab front panel. Mike was told then that the Marine Corps was responsible for the repaint. A short time later, I was refurbishing a 1961 Capehart and another friend, KØDWC, was doing some OD painting over at his QTH. He had automotive quality OD paint that had been custom mixed and he was spraying some of his Army gear using a compressor and siphon spray gun. On a lark, I brought over the "stripped" Capehart front panel and sprayed it OD. It turned out too dark for my taste, so I rubbed and scrubbed it with various abrasives (the OD paint had "hardeners" added) with the intention of just ultimately stripping it but that aggressive treatment actually lightened the OD color into something I thought looked nice and vintage. Additionally, the various nicks and scratches seemed to impart an original appearance to the panel. At that time, W7MS had just recently told me about the USMC OD panel Capehart R-390A he had seen, so with that information in mind, I decided to refurbish my Capehart into a recreation of a USMC version of the receiver. Quite recently, July 2022, I reconfirmed with Mike that the receiver he had seen years earlier had been a Capehart with an OD front panel. The double-check with Mike was to justify what I was about to do next to this old recreated OD USMC R-390A that was now just a couple of steps away from the junk pile.
Just before the closing of the Western Historic Radio Museum in 2012, I had disassembled the Capehart. All of the modules were taken out of the main frame. I think I was planning to rebuild the receiver with new capacitors or something (not necessary since most of the caps are original disks and polyfilms.) The receiver was working but not to the capabilities of a typical R-390A. Through the move to Dayton, I kept the Capehart parts together. Sometime later, I acquired a large collection of R-390A "stuff" that included several "part sets" R-390A receivers and many extra modules of all types. I stored all R-390A parts on the shelves of one large 48" wide, 24" deep by 72" tall shelving unit. For some reason, while rebuilding one of the EAC receivers, I robbed Z503 off of the Capehart IF deck. I also robbed the CL and Line Level meters for one of the other R-390A projects several years ago. Actually, for the past ten years just the Main Frame and Front Panel were bolted together. All of the other modules and parts were scattered in with all of the R-390A parts. When I purchased the black anodized front panel for the NSA recreation it came with both the CL meter and the Line Level meter. Those two parts had been missing from the Capehart for quite a while. So, here it is July 2022, and now I'm thinking about bringing this OD panel recreation of a USMC R-390A back to life.
Locating Modules - I've gotten most of the correct modules together. I didn't remember which PTO was originally in the Capehart. Years ago, I was going to install a Raytheon rebuilt PTO into this receiver but I couldn't find any of the PTOs that had Raytheon labels. I did find two really nice Cosmos PTOs, so I'll start with the best one and see how it goes (if it's a dud, I have lots of other 70H-12 PTOs in the parts collection.) I did find the Raytheon PTO. It was inside the "Blue Striper" main frame. It's pretty rough with some surface corrosion. I'll stick with one of the Cosmos PTOs - and that turned out to be a very good decision.
I'll have to repair the Capehart IF module which hopefully won't be too "hacked" (I don't really remember how careful I was in removing Z503, hopefully I didn't do too much damage.) I have a couple of good "donor" IF modules to provide a Z503. This Capehart IF module was in very good cosmetic condition and did function, otherwise I'd replace it with the fairly nice Amelco IF module I have. A bit later I came to the realization that this Capehart module was actually a "parts" IF module and had never been installed in this receiver.
As I remember, the original Capehart AF module had a bad transformer on one of the audio lines. I tested the AF transformers and T601 (LOCAL AUDIO) has an open primary. I have several AF modules so replacing the bad transformer shouldn't be too much of a problem. Unfortunately, Capehart put their name on the two audio transformers so a "non-Capehart" transformer will be noticeable - but only to OCD restoration enthusiasts (like me.)
I got down the Capehart RF deck and spent about ten minutes on it (straightening two bent brackets) before I noticed that the cams weren't sync'd at +7.000mc and the Crystal Oscillator module was gone. This wasn't even the right RF deck! Looking back on the top shelf there was another Capehart RF deck but this one didn't have any RF transformers and the slug racks were just setting on top of it but it had the Crystal Oscillator module and the cams were sync'd correctly. I didn't remember pulling all of the RF transformers but I do have a large plastic bag full of RF transformers so I guess that's where they are. The other "bent" Capehart RF deck did have all of its RF and Vari-IF transformers installed along with all of the slugs and racks so I could rob those if I need to (probably an easier solution in the long run.)
I found two Capehart PS modules in the pile and I'm pretty sure about which one is original to this receiver since the other one has bashed-in tube socket collars.
All of the knobs had been removed (why'd I do that? No doubt to replace beat-up knobs on one of the other projects.) I may end up having to restore a set since most of the spare knob sets I have are in rough condition (I went through all the knobs and did find a decent set that didn't require restoration - lucky.)
All of the tubes are missing in the IF, AF and RF modules. I checked through the tube storage and I should have all types needed - almost all.
|Several years ago, I mounted a Security Dial Cover on the Capehart. It looks
pretty cool so it will be staying with the receiver. The
Security Dial Cover is shown in the photo to the right. Note in
this photo which was taken 8/2022 that the color shading of the
olive drab is quite different than in the old 2010 photo. I took the
photograph in 2010 using an old 3.1mp Olympus digital camera. I used a
flash and I believe that I also had some tungsten lamps turned on. The
tungsten lamps certainly "yellowed"
the color somewhat and the flash also had a tendency to skew
colors. The photo editor I used then couldn't correct color problems. The new 2022 photos were taken with a 12mp Galaxy S10 with no
flash. The result is a more accurate depiction of the "flat finish" olive drab color on the Capehart
panel. The "finished" photo at the bottom also shows the OD more
As mentioned above, the "aging" that I did to the front panel paint imparted a vintage and original appearance to the panel. When reassembling the OD Capehart, I'm not going to restore the knobs since I want them to look "used." Also, the top and bottom covers are semi-rough originals. The Utah plate was severely bent and had to go through some body work just to get it to lay flat. So, this rebuild is going to end up with the Capehart looking like a pretty good condition original but inside will be all Capehart modules. Performance will be as expected from a rebuilt R-390A.
Rebuilding Some of the Modules - I definitely have to work on the RF deck, the AF module and the IF module so those three units were brought in from the shop first. The RF deck is the most complicated and is missing the most components so it will be first on the bench. The RF deck reconditioning will use the "junk" Capehart RF deck for parts since what is missing on the "rebuildable" deck is present the "junk" deck. The IF module is also fairly difficult due to the component density.
Aug 1, 2022 - RF Deck "Brush and Flush" - I took a look at the gearbox and I must have never given the Capehart RF module the "WD-40 Gearbox Flush" treatment. It's described in the Rebuilding the RF Module section in Part 1. It takes almost an entire can of WD-40 and a long handle paint brush to really flush-out all of the excess grease. I don't know why so many of the earlier R-390A gearboxes seem to have been "greased." The manual specifically indicates NOT to ever use heavy grease (but who reads the manuals?) Most of the Army "car mechanics" that somehow ended up servicing a R-390A gearbox figured "if it's gears, it must need grease" - at least, it seems that way sometimes. Anyway, the last two R-390As I refurbished were 1967 EACs and in both receivers the gearboxes were very clean and didn't need the flush (recheck - I did do a gearbox "flush and brush" on both EACs.) However, this Capehart was full of dirt, dust, dirty dried-up grease and even some spider webs. I did the flush outside and then brought the RF module into the shop and put it on some rags on the floor to drain for awhile.
I also brought into the upstairs rebuilding shop the Main Frame, the AF module, the PS module and the PTO. I already had the spare IF modules stored upstairs. Since the RF module was still out in the shop drying out from the WD-40 flush, I cleaned the Main Frame thoroughly along with the front panel using Glass Plus. Once the RF module gearbox is clean I can put it on the bench and begin reassembly.
Aug 2, 2022 - Greasy Job - An inspection of the RF module gearbox showed that there was still more grease residue. I installed knobs on the KC and MC shafts to allow rotating the gears for a better "brush and flush." Some gears had so much grease residue I had to use a small brass brush to remove it. This "brush and flush" used about half of a brand new can of WD-40 but it's a good solvent and can be purchased anywhere. The RF module was put back on the shop concrete floor with rags under the gearbox to catch the drainage. After a day of setting, the RF module was ready to be detailed to remove the WD-40 residue.
Aug 3-4, 2022 - More Grease - Continued detailing the RF Module. I was very dirty and looks like I never really cleaned it back in 2010 when I built-up this OD panel Capehart. Anyway, there was some minor corrosion on some of the steel parts but nothing that didn't clean off easily with WD-40 and a small brass brush. All tube sockets appear to be in good clean condition and the receptacles for the RF transformers and the Vari-IF transformers look clean. All cams were cleaned of lumpy dried green Army grease. NOTE: If grease is applied to the surface of the cams then the roller bearings on the slug racks won't roll but instead will skid along in the grease. I know the Army manual indicates a light grease on the cams but that defeats the purpose and function of the slug rack roller bearings (that typically don't roll anyway because they are full of dried grease.)
Aug 6, 2022 - Veeder-Root and Vari-IF - Installed a better condition Veeder-Root counter. Started installing the RF/Vari-IF transformers/coils and slug racks. Parts are coming from the other Capehart RF module that was missing a lot of parts that I didn't need. Each RF/Vari-IF transformer/coil has to be taken apart for inspection of the trimmer capacitor (to make sure it isn't stuck) and to clean the sand out of the inside of the can. The contact pins also have to be inspected and cleaned (brass brush lightly then clean with DeOxit) before installation. The slugs had their share of dirt and sand deposits that needed to be cleaned. The slug barrel also needs to be cleaned so the slug can slide up and down with minimal friction. Cleaned and lubricated the slug rack bearings so they will "roll" on the cams correctly. Installed the slug racks and the return springs to complete the Variable-IF section.
|Aug 7, 2022 - RF Transformers and Slug Racks - All of the RF/Vari-IF transformers/coils have to be taken apart for inspection as described earlier. Receptacles were cleaned with DeOxit and a toothpick. Slug racks need to be cleaned and the rollers cleaned and lubricated. Slugs need to be cleaned of sand and dirt for low friction inside the barrel. Installed the 16mc-32mc RF transformers, slug rack and return springs. Interestingly, one slug was missing from the rack. The pile of slug racks that had been installed in this Capehart receiver in 2010 (and the particular slug I needed) was from a R-390 receiver. I just checked one of the other slugs to make sure it had the same colored-dot code and installed it into the new slug rack. RF slugs for the R-390A receivers will have a spline tool adjustment and will be gray in color with one or two colored dots on top. Verified correct operation of the 16mc-32mc rack. Installed the 8mc-16mc RF transformers, slug rack and return springs. After a couple of very difficult to install 4-40 screws to mount the transformers I took a closer look only to discover that sometime in the past someone had used green Loctite when installing the transformers - unbelievable. I had to "chase" each of the pem-nuts with a 4-40 tap to clean out the Loctite. Verified operation of 8mc-16mc rack. Installed the 4mc-8mc RF transformers. One slug missing from the slug rack. Correct replacement supplied by the slug rack parts. Installed return springs, lubed rack bearings and verified correct operation. Saw that the slug rack bearing was hitting one of the shaft clamps in the gearbox. Repositioned the clamp to not interfere with the slug rack.||Aug 8, 2022 -
RF Module Completed - Installed the 2mc-4mc RF transformers. In "chasing" the mounting holes
because of the green Loctite used, I broke the 4-40 tap. Luckily, the
tap must have been defective because the broken tip just "unthreaded"
out of the pem-nut,...that was lucky! Also, the "parts RF deck" had a
bend in the front that made removal of the slug rack I intended to use
difficult. I bent the front a little to get the slug rack out only to
discover that all of the slugs were missing. The adjusting screws were
there but no slugs. I had the same type of slug rack in the slug rack
pile and that one had all of its correct-type slugs. Installed the return springs and
verified correct operation. Installed 1mc-2mc RF transformers with no
problems. Installed .5mc-1mc RF transformers with no problems. Verified
complete operation making sure none of the slug racks were "hanging up."
Installed the 17 crystals required for the Crystal Oscillator. Cleaned receptacles with DeOxit and a toothpick. Cleaned crystal pins with brass brush and DeOxit. Installed the oven (plugs in and covers the crystals) and secured with the three screws. Installed the plug-in dual crystal unit (17mc and 100kc CAL and oven.) Need to find the spring retainer for this crystal although that piece is usually missing. Cleaned underneath and inspected the switches and other components and wiring. Some sand but not too bad. The RF module is ready to install into the Main Frame.
Aug 9, 2022 - No Time - Installed the RF module into the Main Frame but didn't have time to install all of the mounting screws. So,...
Aug 10, 2022 - RF Module into Main Frame and Front Panel Mounted - Installed all of the green head mounting screws for the RF Module to Main Frame mounting and tightened all of the captive screws. Installed both meters to the front panel and then soldered the wires to them from the harness. Mounted front panel to Main Frame. Adjusted MC and KC tuning shaft bearings. Adjusted DIAL LOCK and tightened mounting nut. Applied DeOxit to all of the potentiometers. Installed the 6-32 FH panel screws with conical lock washers. Tightened the four 8-32 FH panel mounting screws. Sorted through the knob bags and found enough good condition knobs to make up a decent-looking complete knob set that I then installed. I looked through the "Utah plates" that I had and all of them were bent up and scratched. I chose the best one, beat it straight (body work techniques so as not to worsen the condition) and installed it with just a couple of screws since I still have to install the tubes. This has completed the Main Frame and RF Module refurbishing. Next is the IF module.
|Aug 11, 2022 - Capehart IF Module Questions - Besides the Z503 being missing, I noticed that the entire CL Meter adjustment pot and wires were gone. I sure don't remember removing the CL pot. A closer inspection of the Capehart IF module was revealing. In checking the wiring for the missing CL pot I discovered that the wire that goes to pin 14 (CL meter terminal) had been moved and was now connected to the 680 ohm cathode resistor on the 4th IF amplifier tube. This, in essence, disconnects the cathode from ground (going though the meter and adjustment network) and routes it to the CL meter and then through the 27 ohm cathode resistor on the AGC TC tube to ground. My guess would be that the CL meter was missing and the wire moved to complete the circuit. However, I didn't do this modification and the 2010 version of the Capehart OD panel receiver had the CL meter and it did work (and it had a CL adjustment pot, too.) So, this means that this IF module was never in the Capehart receiver and it must be one that came with a lot of other "parts" modules. In thinking back, I seem to remember that the IF module that was originally in the Capehart didn't match but I don't know where that one eventually ended up. It might be the Amelco IF module. >>>||>>> Since this Capehart IF
module has the CL pot missing and the Z503 missing, I'll be checking
everything on this module. I inspected the nice Amelco IF module but
it's the older style with lots of paper caps. The other two IF modules are
Collins modules which are very early versions and neither have had the
trimmer caps MWO installed on the mechanical filters. So, I'm pretty
much stuck with this Capehart IF module. Replacements for Z503 and CL pot are going to be salvaged
from a 1960s "spares contract" EAC IF module that was installed in a "Blue Striper" R-390A -
that is, if they check out okay to use. The EAC IF module has a bad 4kc
mechanical filter and a shorted B+ blocking capacitor on the mechanical
filter inputs besides being in kinda rough condition from setting
outside at the St. Julian's Creek Annex.
Removed and installed CL pot and wired. Tested okay. Tested the IF Gain pot and it was open. Removed IF Gain pot from parts module and installed in Capehart IF module. Tested okay. Crystal filter housing has a substantial bent corner. Removed cover and straightened. Also cleaned inside the Crystal Filter. Cleaned the outside of the module. Checked all components and connections to Z503. Missing the 390K resistor and the 2.2K resistor but these can be scavenged from the "parts" module that's donating Z503.
Aug 12, 2022 - Extraction - Since Z503 was coming out of a "Blue Striper" IF module I didn't have to be extremely careful being it's just a "parts donor." But, I did have to be careful of the two standoffs, the 390K and 2.2K resistors and the mounting hardware since these pieces had to be reused. I had to use lots of Solder Wick but everything was removed without damage. I cleaned and tested Z503. It was then mounted onto the Capehart IF module. The standoffs were installed. It's now ready for rewiring. As shown elsewhere in these write-ups there are multiple wires and components soldered to the standoffs that complicate the very easy connections to the two terminals of Z503.
|Aug 13, 2022 - UNBELIEVABLE! - All wires and components connected and soldered to Z503 and the two standoffs with no problems. I then checked over the rest of the module and everything looked okay. But, I thought that I better check the mechanical filter transducers. UNBELIEVABLE! The 8kc and the 16kc MFs had open output transducers. Usually, bad MFs have an open input transducer because of a failure of C553, the plate B+ blocking capacitor. These two MFs have open output transducers which is just a connection to the grid of the 2nd IF amplifier tube. I've run into one other bad MF before on the "Blue Striper" IF module and it has the typical open input xducer and bad C553. Luckily, the "Blue Striper" has a bad 4kc MF but the 8kc and the 16kc MFs are okay. So far, I've been able to rob all of the parts I've needed for the Capehart IF module from the "Blue Striper" IF module. I can't believe what I've found on the Capehart so far,...bad IF gain pot, missing CL adj. pot, I had removed Z503 so that doesn't count and now TWO bad mechanical filters. I can't remember specifically where this Capehart IF module came from but obviously it should have been considered a "junker." I'm beginning to wonder if the two MFs are going to be the last problems,...probably not,...there's always the BFO PTO or the IF transformers.||Aug 14, 2022 - MF Installation
- Access to the underneath of the module area where the filters are is
better if the Bandwidth Switch shaft is removed. It's easy and pulls out
through a hole in the rear apron of the chassis. The connections under
the chassis are the most difficult to do and the 8kc is the most
difficult filter to access. The mounting screws, lock washers and nuts
are particularly difficult to remove or install. The nuts require a
3/16" nut driver if accessible, otherwise I had to use a Philip's Head
screwdriver and hold the nuts with needle nose pliers. All a PITA but
doable. The 16kc filter is much easier to access so it's removal and
installation is not as time consuming. I had to use lots of solder wick
since I didn't want to replace any of the wiring. It's all time
consuming but not really too difficult. It took about an hour and a half to do the
8kc MF exchange and about 45 minutes to do the 16kc MF exchange.
Just to be sure, I checked continuity on all of the IF transformers including the Crystal Filter assembly. I also checked the BFO-PTO for continuity. One other thing to do is probably not necessary but this IF module has been such a pain I'm going to replace C553, the plate B+ blocking cap ahead of the MF filters. Then,...on to the Audio Module for T601 replacement.
|Aug 15, 2022 - Easy One
- The Audio module only required swapping out T601 from a parts module.
The only slightly difficult thing is to dismount the component board to
allow access to the terminals of T601. After the transformer swap, I
checked the two filter capacitors. I had marked on the 45uf
multi-section that "#3 Bad" which was referring to pin 3 on the unit. A
quick test confirmed that the 45uf at pin 3 was indeed bad. The triple
multi-section tested okay. I had powered this Audio module up about ten
years ago and I probably reformed the electrolyics then. I looked in the
shop and pulled the filter caps from another "parts module" and both
checked okay. If I was going to use this Capehart as a station receiver
I'd probably rebuild these or I'd buy the new plug-in replacements from
KE9PQ but this receiver is just for casual listening. Now, I might end
up having to replace these filters sooner than one would think. You never
know with 50+ year old electrolytic capacitors (but these particular
units were built by
Sprague and Aerovox, the best names in capacitors at the time.)
Tube Hunt - I went through the "R-390A Spare Tubes" box to find all of the tubes I needed. I was able to find NOS 3TF7, 6BA6, 6DC6 and 6AK6. In "used-tested as new" 5814, 6AK6 for the Audio Module. For tubes that are just barely acceptable one 26Z5 and three 6C4 (one okay but two are just at minimum acceptable.) I found four 5854/6AK5 tubes and three were so weak they barely moved the test meter while the other one showed high leakage (short.) I think I'm going to have to order some 5854 and 6C4 tubes as these are always weak in almost any R-390A "as found." Installed the usable tubes and found a few IERC tube shields.
Aug 16, 2022 - Nice Find - Ordered one 26Z5 tube since I couldn't find a good spare one anywhere. For testing I can borrow one from another R-390A but eventually one more 26Z5 was needed. I found two good 6AK5 tubes in a VHF receiver I had in the shop. This same receiver was equipped with "loads" of IERC tube shields. In fact, all of the IERC tube shields I'd need for the OD R-390A. I installed all of the tubes and the IERC tube shields with the exception of the one 26Z5 missing. I installed all of the remaining modules so the OD R-390A is ready for synchronizing the PTO which is necessary before any testing and alignments. Before I install the PTO, I'm going to use the PTO bench tester unit I got from W6MIT. This allows doing all of the PTO set ups before installing the unit into the receiver.
photo left: It took two RF modules to build up one complete working unit and two IF modules to build up one complete working unit
|Aug 17, 2022 - A Rebuilt PTO?
- I didn't really think too much about the metalized stick-on tag that
was on the PTO saying that it was rebuilt at the "Tobyhanna Army Depot"
and that it was "set for 2.455mc at +.000," that is, until I powered it
up on the test fixture. Without changing the position of the shaft, the DFC read 2.455mc,...that was something. The test fixture has a turns
counter and a precise zero index indicator for accuracy. I ran the PTO
down ten turns and the DFC read 3.454mc, which is a 1kc end point error
(and that was just a "quickie test.") I checked the linearity and it was
never more than .2kc off. This PTO had been rebuilt and never
used,...well, until now. With careful measurement, the end-point error
was about 0.3kc, impressive for a PTO found in the "spare parts
box." Tobyhanna Army Depot is in Pennsylvania
I installed the PTO with it synchronized to +.000 and 2.455mc output (I marked an index line on the PTO shaft for reference.) Both the gearbox shaft and the PTO shaft oldham coupler pieces have to be loose on their shafts so everything can be aligned mechanically, then the set screws can be tightened. I used a spring from an old VCR replacement kit for the anti-backlash. Checked the rotation and mechanical alignment to assure there wasn't any binding. I really screwed up this synchronization and had to redo it later.
I "borrowed" a 26Z5 tube from another R-390A so I could power-up this receiver. As the receiver "warmed up" the typical operation noise came over the loudspeaker. I turned on the BFO and, as expected, it was way off. Quickie adjustment had the BFO beating with the noise within the 8kc passband. I tried all Bandwidth positions and everything sounded normal. Turned on the CAL and there "weakly" was the calibration oscillator signal. Now I didn't expect the receiver to do very much. After all, it had been completely disassembled, I used all different RF/Vari-IF transformers and some different slug racks, the IF module had lots of rework done to it and it wasn't the original IF module that had been in the receiver either. The crystals for the Crystal Oscillator were a "set" that I had harvested quite a while ago from another R-390A parts set. So,...just about everything in this receiver should need alignment and that's to be expected considering the rework and parts used in this project. A complete alignment will be next. I'll probably find some other problems but "basically" the receiver seems to be operational, just completely out of alignment.
|Aug 18, 2022 - IF Probs
and a PTO faux pas - I went through the IF module
alignment. The 8kc and 16kc mechanical filter trimmers needed some
adjusting which wasn't unexpected. The fixed IF was pretty close but I
tweaked it slightly. The problems started with Z503 (naturally!) The
slug seemed to adjust at first with a slight peaking then I couldn't
turn the slug at all. I tried a little (very little) oil down the barrel
but that didn't help. After a few minutes of this - then I didn't have
any AGC voltage. I took the cover off of Z503 and apparently the coil
itself was turning rather than the slug. This had broken the two wires to
the coil at the two vertical buss wire connections. Luckily, the breaks were
at the buss which is an easy repair. Had the break been at the coil
itself the repair would have been much more involved. But, with the coil
connections repaired I put the Z503 cover back
on. Since I had ample AGC voltage, I decided to leave the slug "as is."
I "zero'd" the BFO.
The Variable IF alignment seemed to be off by 100kc. The only thing that would do that is the PTO. I checked using a DFC on the PTO output with the frequency dial set to +.000 and I had 2.555mc - oops! Some how, I had de-sync'd the PTO when installing it. I had the oldham coupler clamps at 180º opposite of each other and that prevents an accurate mechanical alignment. This time I rotated both of the oldham coupler clamps to be accessible from the same position. In this manner, one tightened clamp will hold the PTO at 2.455mc while the tuning dial can be accurately set to +.000 and then that clamp tightened. A double-check confirmed that the PTO was set right this time. With the PTO connected back to the RF module, I now had lots of Diode Load voltage and went ahead and aligned the Variable IF at the required frequencies. As a test, I connected a ten foot long wire to the antenna input and tuned in 10mc WWV with it having a pretty strong signal. The CL meter showed +30db - not too bad with an incomplete alignment and a ten foot long antenna. I tuned in a local AM-BC station (well, Reno 35 miles away) and the signal had good audio quality especially in the 16kc bandwidth. I'll complete the alignment with the RF tracking and Crystal Oscillator peaking tomorrow.
|Aug 19, 2022 - RF Tracking
- The first step was to adjust the "balanced input" trimmers. A pair of
68 ohm resistors are used to create a network to simulate a "balance"
and then the balance trimmers adjusted for a null at the proper frequency. Next,
was the actual tracking adjustment of the slugs and trimmers in the 16 RF
Transformers. The slugs are adjusted using a spline wrench tool and the
trimmers are adjusted with an insulated blade screwdriver. The important
part of this alignment is to first set the receiver frequency then set
the RF generator frequency and then "rock" the RF generator frequency
for the peak voltage indication on the Diode Load before adjusting the
proper slug or trimmer. The accuracy of the dial readout is dependent on
the PTO and the Crystal Oscillator crystals not the slug position or
trimmer adjustments. The proper position of the slugs and trimmers is to
have the selected tuned circuits resonant with the PTO/Crystal Oscillator
at the same time to provide maximum gain. As the tracking alignment
continued on I noticed that most of the adjustments were very close.
Only where I'd changed a slug or used a different slug rack did the
adjustments require multiple excursions back and forth to get tracking
correct. After the tracking was completed then I peaked the Crystal
Oscillator trimmers although all of them were very close.
A quick test with the 135' Inv Vee and Viking Antenna Coupler connected to the Balanced Input and everything seemed okay. There were the usual ham nets on 40M and several 20M hams from the Midwest to the East Coast. WWV 15mc was checked and the CL meter was at about +30db. I installed the bottom cover and the Utah plate. I listened a little later in the afternoon and tuned in XSG on 16.898mc, a Chinese Coastal Beacon, that was easy copy. Lots of 20M activity. A couple of things noted,...this receiver had to end up using two of the "minimum acceptable" 6C4 tubes. The RF gain is limited in range with about 8 being where things "drop out." This was exactly the same problem I had with the NSA Black Panel R-390A (and the actual same tubes involved - no doubt, I need to order some NOS 6C4 tubes.) The other minor problem is about 200hz of backlash. I think that VCR spring I used on the Oldham coupler was a little on the lightweight side. I'll replace it with a heavy-duty type, as original. Overall, performance is as expected and certainly better than this receiver performed back in 2010 the first time around. Additionally, this time the receiver is equipped with all Capehart modules.
|Update and Follow Up - Aug 26, 2022
- I installed NOS 6C4 tubes for all three oscillators. This fixed the
limited RF Gain range and now moderately strong signals can be copied
easily with the RF Gain at
5 which is typical of a R-390A in good condition. Overall electronic
functionality is very good with 15mc WWV showing +45db on the CL meter
with the meter at 0db when no signal is present but with received
"noise" only. Usually, I set the CL meter to indicate some noise level
like maybe +10db but this CL ADJ is very "touchy" and after many
readjustments it seems to have settled down and is now stable. Also, the 0db
background noise indication is only for the higher frequencies. If I
tuned around the 75M band, the background noise level would indicate
about +10db. Normal for the CL meter circuit and the AGC.
I examined the backlash problem and it's due to excessive wear in the gear train. With close examination I can see that some of the smaller straight-cut gears have somewhat "pointed" gear-teeth which would be indicative of substantial wear. This is probably why the gearbox was loaded-up with grease which can act as a damping agent to somewhat reduce clearances. Special greases, such as Nyo-gel, are specifically designed for damping ability. I have some very high viscosity Nyo-gel that I use for binocular eyepiece focusing threads that would probably work in this gearbox to reduce the backlash. Of course, I'd only use the Nyo-gel where needed since only a few of the small straight-cut gears are worn. Update will follow,...
R-390A PART 1 R-390A PART 2 Return to Home Index
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 © 1997/2023