Radio Boulevard
Western Historic Radio Museum
 

RACAL
British High-Performance Receivers

RA-17 Series, RA-117 Series, RA-6117 built in USA,
RA-6217 Solid-State Receiver, RA-237-B LF Converter

RACAL Company History, RA-17 Circuit Design and the Wadley Loop
Details on the RA-17 Mark I, RA-17 Mark II and the RA-17 Mark III
 Details on the RA-17C-12 US-NATO Version, List of All RA-17 Versions
 Details on the RA-117 UK version, RA-6117 USA-built version, List of RA-117 Versions
RACAL RA-17 Receiver Set-up and Operation
RA-17 Receiver Alignment Overview, Equipment Needed
Procedures for RA-17 Models, Details for a "First Alignment" 
Other RA-17 Refurbishment Information, Performance Expectations
 RA-6217 Solid-State Receiver - Details and Refurbishment
RA-237-B Low Frequency Converter - Refurbishment and Operation

by: Henry Rogers WA7YBS - Radio Boulevard

RA-17L Mark III - B&W from the manual

RACAL

Raymond F. Brown

In 1950, both Brown and Cunningham had quit their jobs at Plessey Company, a large British electronics firm, to start their own electronics business. Both men put up about £ 50 each to get their business started. Initially located in Isleworth, West London, within a year the company became Racal Engineering Ltd.

When Racal business was slow, as it often was at first, they would build things other than electronics, such as golf clubs or shelving units.

In 1954, Racal moved to their commonly known location, Bracknell, Berkshire, which is west of London.

George "Jock" Calder Cunningham

RACAL, The Royal Navy and the Collins 51J Receiver

Things changed for Racal in 1953 when the British Royal Navy became interested in having Racal build a couple hundred Collins 51J receivers for Royal Navy use. The 51J was being successfully used for RTTY communications in the USA military. It used a combination of selected crystal oscillators, a variable dual-IF system and a permeability tuned VFO (Collins' PTO) to provide frequency-solid stability along with a dial readout that was "to the kilocycle" accurate. The 51J was the first receiver used by the US military that provided reliable RTTY communications without modifications. Collins followed it with several other similar receivers based on the same design. The US military R-388/URR was the most successful 51J version and was produced in the greatest quantity (with a civilian designation of 51J-3.) Certainly the Royal Navy was aware of the success the R-388/URR/51J-3 afforded US military RTTY users and they needed the same type of reliability for their RATT comms (British for RTTY.) Photo right shows the Collins R-388/URR version from 1952 housed in a CY-1260/G military cabinet. 

If Racal was going to build copies of Collins' R-388/51J, they would have to be licensed by Collins to do so. Naturally, Racal wanted to use mostly British parts (probably so did the Royal Navy) but Collins Radio Company insisted that only parts from the USA could be used and only receiver assembly could be performed by Racal. Ultimately, the proposal ended up with a group from Collins doing an inspection of the (then) small Racal manufacturing facility. Apparently, Collins wasn't impressed and simply refused to license the manufacture of the R-388/51J by Racal. The only option left was for Racal to design their own receiver that would meet the requirements for the Royal Navy contract.

Racal contacted Dr. Trevor Wadley to help with the project.

Collins R-388/URR Receiver was an upgraded military version of the 51J receiver and probably what that the Royal Navy wanted.

RA-17 Development and RACAL Evolution

The Wadley Loop allowed the new Racal receiver perform like a Collins 51J, with superb frequency stability and 1kc dial readout accuracy, but doing so with an entirely original design. Other unique features included "no band switches." The receiver was tuned entirely by using two VFOs with their associated dial readouts along with specific mixers and bandpass filters. A tuned and amplified pre-selector with front panel tuning control. Six IF band width settings with front panel selector. Five separate audio output lines available on a rear chassis terminal strip, one with an independent AF gain control. 100kc IF dual output on rear chassis. Calibration scale on the BFO control. RF or AF meter scaling switch. Built-in loudspeaker that can be switched off if desired. The receiver was going to be semi-modular in construction and the base chassis was a cast-aluminum, heavy-duty with cast-in bottom sections for the necessary shielded compartments.
 

Dr. Trevor Wadley test-listening to a RACAL RA-17 receiver.
 photo from: www.samilitaryhistory.org

The Wadley Loop - The Wadley Loop was the name given to the RACAL RA-17 front end design that virtually eliminated frequency drift by using a combination of one fixed 1mc crystal oscillator, one VHF tuned MC/VFO and three mixer circuits. The "drift cancelling" happens because "Mixer 1 + the MC/VFO" and "the MC/VFO + Mixer 4" are the inputs to Mixer 2. If there's a frequency drift of, for example, 2kc in the MC/VFO, that frequency drift is seen in both Mixer 1 and Mixer 4 outputs (+2kc.) The Mixer 2 output is the difference of the two inputs (both +2kc) that cancel because of the phase difference in the Mixer 2 function so Mixer 2 output still results in the 2-3mc IF. Since the drift is added simultaneously to both inputs of Mixer 2, the resulting output frequency doesn't change since the difference remains the same, thus the drift is cancelled. Or,... (Mixer1f+2kc drift) minus (Mixer4f+2kc drift) equals Mixer2f+0kc drift.

"Let's assume we want to receive a signal of 12Mhz.

This 12Mhz goes through the two low-pass filters and the RF amplifier and arrives at Mixer 1. The Mhz VFO (VFO-1) is tuned to 12Mhz thus resulting in an output of 52.5Mhz.

The output to the 1st IF is 52.5 - 12Mhz = 40.5Mhz; this signal can pass through the 40Mhz filter which is, as previously said, 650khz wide (this is actually referenced as +/- 650kc, so that's 1.3mc wide total and is shown as 40.65mc to 39.35mc 1st IF.)

First we take a look at what happens at the 37.5Mhz side:

The Mhz VFO signal of 52.5Mhz arrives at Mixer 4 together with the 15th harmonic of the 1Mhz oscillator. These signals are mixed and we get 52.5 - 15 = 37.5Mhz.  (15th harmonic because HAR GEN out is 3.0mc at 0 on MC dial, add 3.0mc so at 12mc on dial HAR GEN out = 15.0mc)

The received 12Mhz signal (which has been translated to 40.5Mhz in the meantime) is mixed with the 37.5Mhz signal at Mixer 2 and becomes 3Mhz. The khz VFO (VFO-2) is tuned to 0, giving an output of 3.1Mhz. These two signals are mixed in Mixer 3, resulting in the final IF of 100kHz.

Signal Path Block Diagram from the RA17 Manual

The signal at the aerial is still 12Mhz. The Mhz VFO has drifted up in frequency by 100khz (0.1Mhz) and now gives an output of 52.6Mhz. The resultant frequency from Mixer 1 is 52.6 - 12 = 40.6Mhz.

Now the 37.5Mhz side again:

Mixer 4 is presented with the 15th harmonic of the 1Mhz crystal oscillator, resulting in an output of 52.6 - 15 = 37.6Mhz.

This signal is mixed in Mixer 2 and results in (yes!) 40.6 - 37.6 = 3Mhz!

This technique ensures that any drift in the Mhz VFO is cancelled out, due to the way in which its signal has been fed into the system twice."

RA-17 Production and RACAL History

Although the Wadley Loop might be the "heart" of the Racal receiver, by this time designated as the RA-17, it was the elaborate Antenna-RF preselector circuit that allowed the receiver's performance to be maximized for any frequency or antenna used. The preselector had a very Hi-Q and had six tuning ranges plus a variable tuning control. This allowed exact tuning of the RF amplifier grid input which resulted in the absolute best transfer of antenna energy to the receiver's front end. A four-step attenuator could be selected if unusually strong signals are encountered. The receiver was triple conversion with a tunable 39.35-40.65mc first IF, then a 2-3mc second IF and finally a fixed 100kc third IF. Six IF bandwidths from 8kc down to 100hz were available on the early versions (13kc to 100hz on later versions.) The BFO was calibrated and had a range of +/-3kc on early versions, +/-8kc on later receivers. The meter could measure RF level and AF level although the meter scaling was 0-200uA (later, a S-meter function was added to North American versions.) Four low power 600Z outputs were available at the rear of the chassis to drive external devices and the built-in 3.0Z speaker was provided for convenient monitoring. An additional 3.0Z 50mW output was available at the rear terminals. The tuning dial was a 35mm film strip that moved behind a fixed (but adjustable) reference index. The film strip had dual 1000kc scales graduated in 1kc increments. The black scale was for HF and the inverse red scale was for LF. A 100kc calibration oscillator could be used for adjusting the position of the moveable index to allow accurate frequency indication. Perhaps as homage to the Collins 51J receivers, the dial escutcheon is quite similar in shape but the kilocycles are within the slide rule portion and the megacycles are in the lower, arced portion - opposite of the Collins receivers. Most RA-17 receivers used 23 tubes but most of the later ones built specifically for North American used 22 tubes since the British GZ-34 rectifier tube was replaced with a solid-state rectifier.

Racal's new receiver, the RA-17, was the first successful receiver to employ the "Wadley Loop" system for oscillator and conversion stability. The first British Royal Navy RA-17 receivers supposedly cost £ 300 each, an equivalent cost then of about $1500. The RA-17 prototypes were built in 1956 with production starting around 1957 and running up to around 1973. The RA-17 design effort wasn't cheap however. Before the receiver was in production, Racal had over-spent by £ 40,000 on the design and had almost bankrupt the company. A contract with the British Army helped Racal recoup their losses and assured their continued future. The RA-17 was upgraded several times but also many special versions were produced (over 40 versions.) The RA-17 "C series" was built for North American or NATO use and has mostly US-compatible parts. 

As the Racal receivers became more and more popular with world-wide commercial and military users, a manufacturing facility was opened in the USA (one of many around the world.) The USA manufacturing facility was called Racal Communications, Inc. and it was located in Silver Spring, Maryland (moved to Rockville, MD in 1972.) All of the receivers produced in the USA (or produced in the UK for North America-NATO) will have either an "A," a "C" or "6" (or combinations) added to the standard model number thus the RA-6117 was the "made in the USA" version of the RA-117 receiver (the RA-117 was a different version of the RA-17 that had an additional IF conversion with a slightly different KC VFO for easier compatibility with other RACAL equipment. For the most part, the RA-117 was produced concurrently with the various types of RA-17 receivers.)   >>>

The Many Names of RACAL

1950-1960s Names: Racal Ltd., Racal Engineering Ltd. 

1960s to 1990 Names: Racal Electronics Ltd (UK), Racal Communications Ltd (UK), Racal Communications, Inc. (USA)

post-1990 Names: Racal Radio Ltd , Racal - The Electronics Group, Racal Radio Group Ltd.

RA-17 Mark I and RA-17 Mark II

Photo of Anthony Howard's RA-17 Mark I

Initial Artwork for the RA-17 Mark I

RA-17 Mark II  - These receivers had several upgrades. Most Mark II versions will have the single pentode RF amplifier and the RF coils in the preselector are an "auto transformer" in construction with no separate primary winding. This style of coil was later changed to a true RF transformer type of construction using a primary and secondary winding. The Mark II BFO NOTE KC/S control only spans +/- 3kc. This function was later increased to +/-8kc on Mark III receivers. The IF BANDWIDTH selections are 8kc, 3kc, 1.2kc, 750c/s, 350c/s and 100c/s while on the later Mark III, the selections are 13kc, 6kc, 3kc, 1.2kc, 300c/s and 100c/s. The receiver's front-end gain control was only marked "IF GAIN" not "RF/IF GAIN" as found on the later receivers.

The meters may be Ernest Turner types or some receivers have a meter with "TRA" on the scale. Many of the serial number tags are marked "RECEIVER MARK II" but some aren't. Knob styles were changed but probably during the Mark I production. Probably the same for the square speaker bezel and the square flange meter. Note the photograph of Dr. Wadley tuning a RA-17 receiver shown further up this page. Note that these receivers have odd tuning knobs (with skirts) installed but otherwise have the appearance of the later RA-17 (that is, not the Mark I appearance.) These receivers could be very early Mark II production. 

RA-17 Mark II - UK Version - Below are two excellent photographs of the standard RA-17 Mark II. The photos are from "Radiosurplus Elettronica srl" - a military radio surplus dealer in Italy. This early version of the RA-17 is serial number N1018. The meter has the logo "TRA" on the scale but is a 0-200uA with 10mw level marked in red. The front panel photograph interestingly shows a rather large set of clear Perspex "scuff protectors." The rear chassis photograph shows the AC primary voltage selector switches on the left side of the rear chassis. This photo also shows the British coaxial connectors quite well. The special connector for the AC power cable (which is in the plastic bag on top of the receiver) is a three-pin receptacle with a threaded collar to retain it securely.

RA-17C SN: N1654 from 1959

RA-17C Mark II - North American Version

SN: N1654
 

The use of "C" as a suffix generally indicated the receiver was built in England for North American users, which normally meant it was for the USA (but sometimes NATO.) The "C" suffix numbers start with "no suffix number" and go up to at least suffix number 26. It's probable that from suffix number 12 on, the receivers are Mark III versions. Shown in the photos to the left and below is a very early RA-17C receiver with no suffix number, actually identified as "RECEIVER MK II" on its serial number tag with the serial number of N1654. Date-codes on the receiver components are from June 1959. Obvious are the different style pointer knobs used on this model. These knobs are somewhat similar to the standard RA-17 pointer knobs but without the grip fluting and without the skirt. These are original knobs and are type 4669P made by Eddystone and seem to only have been installed on the early RA-17C receivers. 

Visible is the Ernest Turner W909 meter with the S-meter scale in green and, of course, the three position rotary switch for MONITOR function to allow RF LEVEL (nomenclature worn off,) S-METER and AF LEVEL. The cinch plug under the meter is access to the S-meter adjustment pot and this adjustment only affects the S-meter calibration. Note that the AC ON-OFF toggle switch is marked "POWER" and not "MAINS" - all receivers destined for the North America-NATO market had this nomenclature change.

The Phillip's head screws that mount the front panel seem to be original. They are 10-32 screws and all eight are identical with the same type of age-related surface oxidation. Also, the Limiter switch has been replaced at some time in the past (switch date-coded 1962.)

Top of the Chassis of N1654 showing that it's in remarkably good condition. Note that there are only six adjustments for the Ant Tune since auto-transformer coils are used in the Mark II (above the "A" marked on the RF/VFO 1 module.) Note the 5V4G rectifier tube lower left of the chassis. This tube has a spring-type retention device that wasn't used on the later receivers.

Unknown RA-17C
 

The photo to the right shows another RA17C Mark II that happened to be for sale at Fair Radio Sales, formerly one of the largest surplus electronics dealers in the USA, located in Lima, Ohio, now out of business.

Despite it's rather poor condition, it does show that these early "C" version receivers were supplied with Eddystone 4669P pointer knobs. Note that the knobs on this receiver are exactly the same type as those shown on SN: N1654. Though difficult to see, the meter is an Ernest Turner W909 with the S-meter scale. I don't think even scuff protectors would have helped this receiver.

This photo was on the Fair Radio Sales website (consequently, its poor quality.) This receiver did sell around February 2020. No serial number was supplied in the description.

RA-17L Mark III

The Mark III changed the RF amplifier from a single pentode tube (CV3998) to a cascade connected dual triode tube (CV5531.) Also upgraded was the A-E preselector coils changing from resonated single coil assemblies to tuned RF transformers. This improved the Q of the preselector for better selectivity. The IF bandwidths were changed as follows; 8kc increased to 13kc, 3kc increased to 6kc, 1.2kc increased to 3kc, 750c/s increased to 1.2kc, 350c/s changed to 300c/s and 100c/s remained 100c/s. These bandwidths provided a wider response for voice transmissions such as shortwave broadcasts while the CW and data mode bandwidths remained essentially the same. The BFO range was increased from +/- 3kc to +/- 8kc and a vernier dial added to the BFO control to ease tuning due to its wide range of adjustability. V-24 was a wire-end vacuum tube diode that was only used in the earliest versions of the RA-17 and was replaced with a solid-state diode in the Mark II and Mark III.

There are five audio outputs available on the rear chassis terminal strip. Three are 600Z ohm windings from the audio output transformer. The panel mounted monitor speaker runs off of the 3Z ohm winding of this same transformer. When the monitor speaker is switched off a 10 ohm load resistor is connected to the winding. The 3Z ohm 50mw winding also has a output connection to the rear terminal strip. The audio level of these four outputs is controlled by the AF GAIN. The remaining 600Z ohm output is controlled by the AF LEVEL control on the front panel and this output has its own separate audio transformer. This same terminal strip also has an AVC connection for running multiple receivers in diversity. Also accessed at the rear chassis is the RF and 1MC coaxial connectors and the HT1, HT2, AVC terminal connections for the LF Converter. If the LF Converter isn't used, then the HT1 and HT2 terminals must have a jumper installed to complete the +250vdc circuit in the receiver. Other rear connections are the dual 100kc IF outputs, the AE (antenna-earth) coaxial connector and the AC power cable connector. The primary taps on the power transformer are brought out to two switches that allow setting the proper AC input voltage level for the receiver depending on what is the local line voltage. Other minor Mark III changes were to the Stand By function with a 9K 10W load was added to the B+ when the receiver was placed in Stand By (also on RA-17C versions except value changed to 10K 10W.)

RA-17L Mark III from 1964 - installed in its original RACAL table cabinet. The data plate is missing from this receiver so the serial number is not known. However the date was determined using the date code on the Gresham power transformer. It's fairly common to find the data plate missing on receivers that were used by the military and then later decommissioned.

The RA-17L chassis is shown above. At the left side starting at the bottom is the BFO, then the two 100kc IF transformers, next up is the 100kc LC filter and the larger shield-box is the Crystal Filter. To the right of the 100kc IF transformers is the KC VFO and plugged into the side of the KC VFO is the 100kc Calibrator. The large module is the MC VFO and the RF/ANT Preselector circuitry. To the right rear is the audio section and the power supply.

The kilocycle tuning dial is a filmstrip type that is six feet long and spans 1000kc with a resolution of 1kc. When tuning the receiver, the dial index remains stationary while the numerical dial scale moves behind the index in a linear fashion. The megacycle dial is circular and is read thru the lower window of the dial escutcheon. The black kilocycle scale is used while tuning from 1-30mc while the red kilocycle scale is used when operating with the optional MW-LF-VLF converter (10kc to 980kc.) The tuning system is super-smooth and very light feeling. The film strip is easy to read to at least 1kc accuracy.

The RA-17L uses triple-conversion and the drift reduction Wadley Loop provided solid stability. The normal Mark III bandwidth was selectable from 13kc down to 100hz in six positions although there were some versions the changed some of the selected bandwidths depending on the end user requirements. The RACAL RATT demodulators used the last IF as an input source. The last IF operates at 100kc and dual outputs were provided on the rear chassis. As with the Mark II, selectable fast and slow AVC and a Noise Limiter were provided on the Mark III. A standard diode detector was used and on very early Mark I receivers the BFO was lightly coupled (10pf capacitor) but by the Mark II the BFO coupling capacitor was increased to 47pf which provides sufficient injection to allow easy demodulation of SSB signals. The Mark III provided a toggle switch for the meter function with RF Level and AF Level as functions. The 10mw level is marked on the meter. This is measuring the 600Z audio output that has a separate audio output transformer and the front panel Audio Level control adjusts this output as needed. By the time the Mark III was introduced, Ernest Turner was supplying the meters and they were identified as Type W909.

RA-17L Mark III receivers were generally equipped with military/commercial designated tubes. Ten CV138 tubes, one CV5531 tube, one CV2209 tube, two CV4012 tubes, two CV3998 tubes, four CV4009 tubes, two CV140 tubes and one CV1377 tube. There are some of these tubes that had RACAL approved substitutes. Also, the Pro-Electron (European designation) equivalents were usually listed for most tubes and most of the tubes could be substituted with US-type tubes. One exception is the CV2209/CV4064 or S6F33,...some lists show the 6AS6 as a substitute but this tube's screen grid and suppressor grid pin outs are reversed when compared to the correct tube. Also, "S" in S6F33 indicates the tube was a "special" 6F33. Whether a standard 6F33 will function correctly in the circuit is unknown (by me, anyway.) The RA-17L Mark III used 23 tubes but if one is using an earlier manual, V24 maybe be shown (CV469.) It was a wire pin diode tube used in the Mark I and possibly Mark II receivers. It was replaced with a solid-state diode probably by the Mark II but definitely by the Mark III production.

The panels on the RA-17 receiver were originally Light Admiralty Gray (British Standard Specification 381C colour 697) which really isn't gray at all but is sort of a light greenish-slightly bluish-grayish color that was supposed to have been carefully designed to reduce operator fatigue. Many of the panel colors that we encounter today have faded somewhat and are a lighter shade than the original color was. It depends on the environment that the receiver operated in and, to a great extent, its physical condition.

Many RA-17 receivers panels were equipped with panel "scuff protectors" that mounted against the front panel and behind the Kilocycle and Megacycle tuning knobs. These large, circular protectors were normally made from clear Perspex (a British company that produces acrylic plastic sheets) and shielded the panel from "fat finger" syndrome. "Fat finger" wear was caused by the receiver operator rubbing their finger tips against the front panel by "over-gripping" the tuning knobs. The result, overtime, usually left the front panel paint worn off in a "ring" surrounding the knobs. Most of the "scuff protectors" were mounted to the panel by way of a metal "mounting ring" that was placed through the panel holes for the KC and MC tuning. The "scuff protector" hole was then placed over the "mounting ring" and then the tabs were bent over both on the back of the panel and on the front over the "scuff protector" hole. This method allowed some adjustment of the "scuff protector" position and allowed removal, if desired (much better than gluing.)

The MC and KC tuning knobs along with the four or five "non-winged" smaller knobs have a collet-type of grip on the shaft. Knob removal requires using a hex wrench to loosen the chrome nut at the center of each knob. These collets must not be over-tightened when reinstalling these knobs.  

Most of the RA-17 side panels have a slope beginning about halfway back and the top cover also slopes at the rear half of the receiver (see photo above and photo in RA-17C-12 section showing the top cover.) The front panel is steel and is the standard 10.5" x 19.0" but the receiver depth is 20.5" which is considerably deeper than any USA-made receivers. RACAL did supply a table cabinet for the receiver depending on the end-user's requirements. The top cover could be removed if the receiver was installed in a cabinet and was to be used in a warm environment.

The standard RACAL RA-17 cabinet was initially black wrinkle finish. Some later receivers, depending on the particular end-user, were supplied with cabinets painted other colors. The RACAL RA-17 cabinet shown in the photos is an original version in silver hammertone finish. This paint-finish cabinet was sometimes used commercially (see photo further up of the BBC-Crowsley Park RACAL installation with the two desk-mounted RA-17 receivers in silver cabinets.) The RACAL table cabinets are steel construction and each of the sides have a recessed area where the louvers are (see photo right.) There are rear mounted engagement pins that mate with holes in the lower part of the rear side panels of the receiver. This provides rear support for the receiver chassis. The cabinet uses a single D-zus screw to secure the lid. Cabinets will generally have a RACAL metal label mounted inside rear bottom. Four large rubber feet are bottom mounted although some military receivers had shock mounts and skids (usually required if the receiver was installed as equipment in a towable or mobile communications hut.) Normally, the installation of the receiver into the cabinet included two chrome-plated strips that mounted over the rack mounting notches and protected the front panel paint from "rack rash." The Racal cabinet weight is 30 pounds which when added to the 67 pounds that the RA-17 weighs pushes the total weight of a cabinet-mounted receiver to nearly 100 pounds. 

Showing the left side recessed louver panel and close-up of the chrome strip rack screw panel protector. The silver hammertone paint on this cabinet is original. These types of cabinets were generally used for commercial installations, such as the BBC's Crowsley Park.

RA-17C-12  Mark III

Other minor C-12 Series changes were to the 3.0Z ohm 50mw audio output that was increased to 1 watt. Also, a 10K 10W load was added to the B+ when the receiver was placed in Stand By (9K 10W on RA-17L versions) and standard UHF and BNC connectors were used for rear-chassis connection auxiliary inputs or outputs for accessory devices. The Antenna-RF preselector (panel ID: ANT TUNE on the C-series, AE TUNE on UK versions) is the same elaborate antenna tuning network that essentially provides a precisely tuned, Hi-Q output that feeds the RF amplifier grid. The preselector has six tuning ranges and a variable frequency "tuner." This tuned RF input can be bypassed in the "wideband" position which is then an untuned "wideband" fixed 75Z ohm input. The Antenna Tuning provides an increase in sensitivity by way of much better signal transfer to the RF amplifier grid and also the Hi-Q allows for better selectivity for the incoming signal which helps with reducing interference from strong, near-frequency, near-field signals. The four-step attenuator remained unchanged and can be inserted into the signal path to reduce the input level to the RF amplifier if intensely strong signals are encountered. The RF amplifier was a standard pentode tube in early RA-17 receivers but the C-12 Series used a cascaded dual-triode tube 6ES8 (the RA-17L Mark III also had this change although the tube used was an ECC189 or CV5531.)

To readout the tuned frequency one has to add the megacycle dial setting to the kilocycle dial reading. If the megacycle dial is set to 14 and the kilocycle dial reads 200 then the tuned frequency is 14.200mc (as shown in the RA-17C-12 photo above.) A 100kc Calibration oscillator is provided to assure frequency readout accuracy.

RA-17C-12    SN: N4144  from 1960
The "C" versions were built in England for North America and NATO so some of the panel nomenclature was changed. "Mains" was changed to "Power." "AE Tune" was changed to "ANT Tune." "AE Range" and "AE Attenuator" were also changed to have "ANT" in place of "AE." All hardware is UNF-USA threads. All tubes used are USA types. The Carrier Level meter switch (MONITOR) and circuitry was changed to have a S-meter function and scale on the meter. The "black scuff protectors" are unique to this receiver and are not standard accessory equipment. Mostly like these are opaque black to hide panel damage around the KC and MC knobs.

Top view of the RA-17C-12 chassis
Note the Hammond power transformer with the "NATO" part number.

The use of triple-conversion and the drift reduction Wadley Loop provided solid stability. The normal C-12 Series bandwidth was selectable from 13kc down to 100hz in six positions but earlier versions will normally have 8kc as the widest bandwidth available. Last IF operates at 100kc with dual rear chassis outputs available. Selectable fast and slow AVC and a Noise Limiter were provided. A standard diode detector was used but (like the Mark III) the BFO is coupled via a 47pf capacitor which provides sufficient injection to allow easy demodulation of SSB signals. BFO is +/-8kc on the C-12 versions and usually +/-3kc on most earlier versions. The "C" versions had a three-position switch allowing RF level, S-meter and AF level functions. C-12 meters were generally supplied by Stark Electronic Instruments Ltd of Ajax, Canada but earlier C-versions use Ernest Turner meters. Several types of audio outputs are available at the rear connector terminals. C-12 series uses BNC connectors to provide 1MC output and RF input (for the LF converter and other compatible accessories) but the earlier versions use a UK-type connector. 100kc IF out is provided with SO-239 connectors. On all of the C-series, the AC power cable is permanently attached to the receiver chassis unlike the UK receivers that had an AC power cable that was removable and attached to receiver using a military-style connector.

The GZ-34 rectifier tube was replaced with a 5Z4G tube in the earliest C-series receivers but by the C-3 version solid-state rectifiers were used reducing the tube count to 22 tubes. The RA-17C Series tunes from 1.0mc up to 30mc in 29, one megacycle wide tuning ranges. Tuning from .5mc to 1.0mc is possible but with noticeably diminishing performance as the frequency is tuned lower within that range. 

The RA-17C-12 SN:N4144 shown in the photos has opaque black "scuff protectors" installed. This type could be installed if the panel already had "wear-rings" since their installation hid any damage and prevented any further damage. Since these type of protectors aren't transparent the panel nomenclature that's covered is engraved into the black-on-white plastic material. It's quite likely that these opaque black "scuff protectors" were custom made for this particular receiver and really weren't "standard" accessory equipment since I've never seen any other "scuff protectors" like them.

Top of the Chassis RA-17C-12 - In the photo view, top left are the Selectivity filters and in front of those two modules is the IF section of the receiver with the BFO in the very front (under the speaker.) Top center is the Harmonic Generator and the 1mc oscillator (note the 1mc crystal on chassis.) Below the 1mc crystal, the cluster of three tubes comprise the Harmonic Mixer and the 37.5mc amplifiers. The left module with the yellow tape is the KC-VFO and conversion mixers and the module just to the right is the calibration module and then the MC-VFO and conversion mixers with the RF amplifier tube to the far right. Front center is the film strip dial drive, speaker to the left and meter to the right. The power transformer and power supply components are located upper right in the photo. Audio transformer is in front of the power transformer. The RA-17L uses Gresham transformers while the RA-17C-12 uses Hammond transformers.

RA-17 Versions

RA-17 - Mark II - Improvements to 100kc filtering, also minor improvements to appearance. Different knobs, different meter housing, different speaker bezel when compared to Mark I

RA-17L - Mark III - New selectivity positions, double tuned antenna coils, dual triode RF amplifier tube, usually also referred to as the Mark III version. There are at least twelve different types of "L" versions.

RA-17C Series - Primarily for North American market with US-type tubes, hardware, BNC connectors, many other minor differences. Over 20 numbered C-versions. Alternately identified as AN/URR-501A by the US military.

The following is a detailed list of the many versions of the RA-17 that were produced. This list is from: "Keith's Vintage RACAL Enthusiast's Site" www.recelectronics.co.uk

Notes: RA-78 was a frequency measuring receiver with nixie tube digital frequency readout and other auxiliary gear - all equipment mounted in relay rack
           RA-66 was a panadaptor that needed the MA251 buffer module to be installed into the RA-17 for interfacing between the two units
           MA-143 was a six-channel crystal-controlled oscillator, worked with RA-117 but could be used with RA-17N - provided six crystal-controlled frequency selections
           MA-257 was a MC stepper unit for selecting ten scanning MC ranges, the ten ranges ran in sequence.
           MA-1350 was a Decade Frequency Oscillator synthesized frequency lock for RA-17 tuning - worked with the RA-17 tuned frequency and MA-1350 switch settings

Other data from Rinus Jansen's "The Racal Handbook"

* = corrections supplied by Neil Clyne

Using the RA-17L and the RA-17C-12 "On the Air"
Two Active Ham Stations using RACAL Receivers at WA7YBS

The station on the left side of this 2025 photograph is the Collins 32V-3 transmitter paired with the RACAL RA-17L. I've been using this set-up for several years and have found the combination works really well. Of course, the RA-17L hears everything on the Nevada Mil-Rad Net since that Net operates on 75M which isn't much of a challenge for just about any receiver and the antenna used is a Collinear Array that provides about +1.9db of gain on 75M. Although copy is almost always Q-5, I've found with very weak AM signals, the RA-17L works quite well in the exalted-carrier mode, that is, AM with the BFO on. In this mode, there is one very narrow "zero-beat" spot that, as the signal is tuned through, on that "spot" the signal will "jump up" about 3 or 4 S-units and become fairly easy to copy. This only works on AM signals that are right at the noise level. Selectivity is excellent and QRM is never a problem. I do most of the listening in the 6kc bandwidth. The Collins 32V-3 is a first-rate, medium power transmitter. Runs 110 watts RF carrier output power (on 75M.) The Industrial Arts desk lamp on top of the RA-17L (in silver hammertone finish that matches the RACAL cabinet) was made by DAZOR Mfg. Corp.

The station on the right side of the photograph is the W6MIT-Svoboda "1625 Rig" running into the W6MIT-Svoboda Linear Amplifier that's setting on top of the "1625 Rig" (the linear amp's power supply is on the floor.) This transmitter is paired with the RACAL RA-17C-12. It was the second RACAL receiver that I found locally and was able to purchase here in NW Nevada. It was housed in an industrial steel cabinet that was absolutely enormous and weighed more than the receiver. I removed the receiver from that monstrosity and, consequently, the RA-17C-12 has been "sans cabinet" for quite a long time now. This NATO/North America version of the RA-17 works very well and has all of the same great performance features and characteristics found in all of the "Mark III" versions. There are a few differences in the "C" versions - the increased 1 watt audio output from the 3Z audio output that's used to drive the Collins 270G-3 loudspeaker on top of the RA-17L (the RA-17L's loudspeaker is a wall-mounted 12" that's out of the picture.) The three function Carrier Level meter is convenient as are the solid-state rectifiers in the power supply along with the all-USA UNF hardware and all-USA-type tubes used in the "C" versions. The W6MIT-Svoboda equipment is "homebrew" from the 1990s. The transmitter plus linear will run 140 watts RF carrier output power in the AM mode.

The RA-17L and the RA-17C-12 integrated into two active ham stations

  RA-117 - Built in England    RA-6117 - Built in the USA

The RA-117 was introduced somewhat later than the RA-17 but for the most part was built along side that receiver. It's a slightly different receiver than the RA-17 and was developed to allow more of a variety of auxiliary equipment to be interconnected and operate with the receiver (such as the MA-79G exciter.) The RA-6117 is a version of RA-117 built in Silver Spring, Maryland, USA by Racal Communications, Inc. It's almost identical to the RA-117 but using all standard USA manufactured tubes. All of the hardware is also standard US threads and sizes. USA-made components used. USA-style Raytheon knobs are used on the RA-6117.
 

The RA-117 antenna preselector circuit was slightly changed to provide a "wide band" selection that was an untuned hi-Z input or to "wide band 75" selection that was an untuned 75Z ohms input. The USA-built RA-6117 eliminated the 75Z ohm wide band position. There is a switch position for 75Z WB but it isn't wired and there's no panel nomenclature. The six-step frequency ranges remained unchanged and still allowed tuning for maximum response for the particular frequency and antenna in use. There was also the step-attenuator provided for coping with very strong signals but still being able to retain the tuned selectivity that the R.F. TUNE (was ANT TUNE on the RA-17) provides. The RF amplifier used a cascaded dual triode tube (6ES8.) The major changes in the RA-117 were to the conversion frequencies and building in more capabilities of the receiver to interface with other types of RACAL equipment. 

photo left: RA-6117  SN: 193  - built by RACAL Communications, Inc. in Silver Spring, Maryland, USA, ca: 1966

The KC-VFO/Mixer operated into a fixed 1.6kc IF which was then converted (crystal oscillator/mixer) to a fixed 100kc IF. The 3.6-4.6mc KC-VFO also featured an external input and an external output to allow the RA-117 tuning to control (or be controlled) by other devices. The RACAL RA-79G exciter when used with the RA-117 could be set up to allow the RA-117 KC-VFO to control the exciter's tuned frequency, allowing a "transceive" operation. VFO selector switch is on the RA-117 front panel next to the Kilocycle tuning knob.

IF bandwidths from 13kc to 100hz, Fast and Slow AVC, three position switch allows the meter to act as an RF signal level, AF level or S-meter, standard envelop detector - no product detector are the features provided and all were basically the same as the RA-17C. Several BNC inputs/outputs are provided for External VFO, various Oscillator Outputs, 100kc IF Output along with terminals for several 600Z ohm outputs and a 1W 3.0Z ohm output. There were some versions produced that had modifications to the AVC usually consisting of longer delay times to allow better copy of RTTY signals (British - RATT signals.)

A separate audio output (600Z ohm) with an Audio Level control on the front panel that has its own output transformer and operates separate from the standard receiver audio output. This output was for driving a data device, an audio input RTTY converter, for example. The small built-in speaker can be switched off if a larger speaker is desired (connecting to the 1W 3.0Z ohm output works best.) RA-117 receivers use 25 tubes. RA-117 Series tunes 1-30mc in 30 bands. Also, .5-1mc tuning with noticeably diminished performance as one tunes lower in frequency within this range.

The UK-built RA-117 front panel was the standard Light Admiralty Grey. Apparently all USA-built versions were cream color with USA-style knobs. Scuff protectors aren't usually found on RA-117 receivers because the EXT/INT VFO front panel switch location interferes with their proper positioning. Although the RA-17 and RA-117 panels are standard 10.5" x 19.0" the receiver's chassis is quite deep at 20.5" which, of course, won't fit into any standard American cabinets. RACAL offered table cabinets that were specifically for the RA-17/RA-117 and these types of cabinets only increase the overall receiver foot-print an inch or two. Even at that, the RA-17 and RA-117 are huge receivers that demand a lot of bench space. Other after-market general purpose cabinets that have the necessary depth are usually very large, very heavy and really increase the overall receiver size and weight substantially. However, most receivers usually will have their dust covers and, like the R-390A, look quite nice setting on a table "sans cabinet" and in that configuration don't seem to take up all that much bench space.
 

Photo Right - RACAL RA-6117 paired with the W6MIT-Svoboda "Big Rig" at WA7YBS - The RACAL RA-6117 was the very first RACAL receiver that I ever found locally and was able to purchase. That was over 10 years ago. Since July of 2023, the RA-6117 has been paired with the W6MIT-Svoboda homebrew transmitter, the Model 45098, aka "The Big Rig" - a 400 watt input AM transmitter that looks like a small AM-BC transmitter. W6MIT, John Svoboda (SK,) built this transmitter in 1998. It has a 4-250A PA modulated by a pair of 572B tubes. The RF exciter is out of a T-368 military transmitter. The external Audio Driver is a Bogen CT-100C PA amplifier. 300 watts RF carrier output power. The "Big Rig" is located out in the shop with the other really heavy equipment - a T-368/R725 combo and a GRC-19. The turquoise "missile room" chair is from Burpelson AFB. For more photos, click on this link "W6MIT-Svoboda Transmitters - Professionally Engineered Homebrew Rigs"   Also more circuit and construction details on the "Big Rig," the "1625 Rig" and the W6MIT linear amplifier, all great examples of the "homebrew art."

RA-6117 Top of Chassis - Much of the component layout is very similar to the RA-17 receiver. Note the extra module located directly behind the KC VFO. This is the Crystal Oscillator and Mixer module for the 1.6mc to 100kc conversion used only in the RA-117 or RA-6117 receivers. Note that above the MC-VFO, the Calibrator module now has the two tubes oriented to a horizontal position. Many more coaxial cables in the RA-117 to allow for the INT-EXT VFO connections and output connections for the dual 100kc IF outputs. At the rear-center of the chassis, note the small Potter-Brumfield relay. I added this relay to provide the remote standby function that wasn't originally installed on this particular version of the RA-6117.  NOTE: I later removed this relay mod because of its non-originality to this particular receiver. In original installations in the "Built in England" RA-117 receivers, a square hole was located under the relay for the wiring. In my installation, no holes were drilled or cut and no components removed or changed for this non-invasive change. However, I found its function to be unnecessary. The same remote standby function can be accomplished using the B+ terminals (as recommended in the RACAL RA-17L manual.) Also, note how the side panels and the rear panel comprise an almost complete enclosure. The top cover is flat with small perforations for venting heat (see photo directly above to see what the receiver looks like with the top cover installed.) This is typical of the USA-built receivers. UK-built receivers used the standard RA-17 type of side and sloping top. The components in this RA-6117A are 1966 date-coded parts.

Additional NOTE: This receiver is actually the RA-6117A-1 version but the 100kc to 25kc converter that mounted externally to the rear panel was removed sometime in the past. The mounting holes for the converter can be see in the rear panel (shown in the photo left.) This absence of the converter doesn't have any effect on the receiver's performance since it was externally connected to the 100kc output BNC connector for its input and then the 25kc output was on the converter itself.

RA-117 Versions

Notes: MA282 was a frequency changing coupler so the RA66 panadaptor would work with the RA-117
           RA220B was a Marine Receiving Terminal
           RA133A Dual Diversity Receiver

         
From "Keith's Vintage RACAL Enthusiast's Site"  www.recelectronics.co.uk   -  other data from Rinus Jansen's "The RACAL Handbook"  -   RA-6117A and RA-6117-A1 data from the RA-6117 manual

RACAL's Ham Receiver?

So,...why don't we see very many RA-71 receivers come up for sale nowadays? The original advertised selling price was $1200!
 

photo right: RACAL RA-71

Details on RA-17 and RA-117 Setup and Operation

The early RA-17C receivers used a Gresham T-1078 power transformer (UK RA-17L also used the T-1078 but provided primary voltage selector switches on the rear chassis apron.) The pin outs are printed on the label on the side of the transformer. The label is shown below because the Gresham pin-outs are somewhat different than later transformers. Later receivers that used solid state rectifiers have the Hammond power transformers that didn't have a 5vac winding or the HV CT. There isn't any terminal to winding information printed on these transformers. Only the terminal numbers are shown on the transformer. The terminal to winding cross-reference is not shown anywhere in the RA-17 UK manual and not even in the RA-17C-12 UK-USA manual. The probable reason is that the standard RA-17 British receivers had two voltage selector switches on the rear chassis for setting the primary voltage, essentially what's shown on the schematics. But, USA version receivers will have soldered wire connections to the power transformer terminals for setting primary voltage requiring access underneath the chassis. As mentioned, the terminals are numbered on the transformer but those numbers and their relationship to the primary winding aren't shown on the schematic. I'm not sure how the writers of the RA-17C Series manuals expected the USA users to set the primary voltage other than measuring the DC resistance of the windings and figuring out what went where.

Fortunately, the RA-6117 manual has a power transformer drawing and the numbers are shown on the RA-6117 schematic. The later C-Series used a Hammond transformer. The USA-built RA-6117 receivers did have a slightly different power transformer (BTC Co.) but luckily both transformers are shown in the drawings. The terminal drawing, schematic and listing are shown below. The actual relationship of the terminal numbers to the primary winding is the same for both transformer types. Only the physical location of the terminals is different. Also note that in reference to the Gresham T-1078, if the 5vac winding terminals and the HV CT were eliminated, then the remaining terminals to windings layout is the same as the Hammond transformer.

My line voltage runs 122vac so I've set my receivers and accessories to the 120vac setting. This requires soldering one wire to 125vac and soldering the other wire to -5vac resulting in 120vac input. With the terminals provided it's possible to set the primary voltage from a low of 100vac (110vac minus 10vac) up to 125vac (125vac minus 0vac) or 200vac (210 - 10) to 250vac (250 - 0.)

Terminals Connections to Windings Cross-Reference
For Hammond or BTC Transformers

Terminal 14 - HV
Terminal 15 - HV
 

Terminal 20 - 0vac Primary Adjust - subtracts 0 from other Pri Term selected
Terminal 21 - minus 5vac Primary Adjust - subtracts 5vac from other PT selected
Terminal 22 - minus 10vac Primary Adjust - subtracts 10vac from other PT selected
 

Terminal 23 - 6.3vac - connected to T1 shield (Terminal 31) and to chassis
Terminal 24 - 6.3vac
 

Terminal 26 - 110vac Primary
Terminal 27 - 125vac Primary
Terminal 28 - 210vac Primary
Terminal 29 - 230vac Primary
Terminal 30 - 250vac Primary
Terminal 31 - Shield-Chassis

Terminal connections are the same for either TF4TX03LA or TF1RX03YY, only the physical location of the terminals on either transformer are different.

Gresham T-1078 Transformer Label

RA-17C-12 uses a Hammond power transformer identified as NATO 5950-21-862-1369/TF4TX03LA with terminals as shown in the drawing on the left.

RA-6117 uses a BTC CO. power transformer identified as TF1RX03YY/BX-3944 with terminals as shown in the drawing on the right.

The only schematic showing the terminal numbers as referenced to the actual windings of the transformer is in the RA-6117 manual. Shown above is a portion of the schematic showing the terminal numbers. The terminal number to transformer winding connections are the same for both types of transformers.

Caveat on Date Codes - On the Plessey capacitor and the Gresham transformer,...these components were dated at their manufacture. Then they were shipped to RACAL, went into RACAL stock, then they had to be drawn out of stock to assemble a receiver. It would be expected that any date code will be a date somewhat earlier than the actual receiver built date.

As an example, the RA-17C Mark II receiver has the Gresham power transformer that's dated "QF" or June 1959. However, the Plessey filter cap has the date code of "MAY 59" - at least a month earlier. Probably the earliest this receiver would have been built would be in August of 1959 (and that's just a guess.) Still, date codes can get you fairly close to a build date estimate.

If the S-meter position is used it will show a relative indication of signal strength but the readings will depend on several factors including antenna type, the received frequency, the setting of the RF Attenuator or the tuning of the preselector. The adjustment pot behind the cinch-plug (just below the meter) is for the S-meter function only. Note that the scales are slightly different between the Turner meter and the Stark meter. The Turner shows a little over S-7 at 100uA while the Stark shows a little less than S-5 at 100uA. The Turner meter shown is on a RA-17C Mark II while the Stark meter shown is on a RA-17C-12 Mark III. Stark meters were also used on the UK-built RA-117 receivers.

The AF meter indication works only on the 600Z ohm 3mw line which has its own individual output transformer and is controlled by the Audio Level control located next to the AF Gain control on the front panel. If you have the Audio Level control set to minimum you will not see any indication on the meter. If you want to see an audio level then you have to install a load resistor of approximately 680 ohm 1/2W load resistor to the audio line terminals on the rear of the chassis. The other unused audio outputs should also probably have similar load resistors installed. 

With the meter switch in the RF Level position, if the BFO is turned on, it will show a constant level on the meter. RF Level only shows received signal carrier levels in the AM mode. In CW or SSB, the meter will show the BFO level plus the level that the signal is above the BFO level. Since CW and SSB are not 100% duty cycle modes the meter reading will follow voice or CW peaks that are above the BFO level.

The meter used in the RA-6117 is a Honeywell Model HS2Z. It has the same type of scale as the Stark Electronics Instruments 605HSR meter but the housing is somewhat different having a convex glass cover.

Four examples of meters used are shown below

Ernest Turner - Model W909

Ernest Turner - Model W909 w/S-meter scale

Stark Electronic Instruments - Model 605HSR

Honeywell - Model HS2Z

Reworking the RA-17 Receiver

The RA-17 or RA-117 are not easy receivers to work on. The semi-modular construction allows removing individual circuits from the chassis but almost all modules are a combination of hard-wiring interconnections to other modules or to component boards along with some coaxial cable-BNC (or British coax connectors) interconnections (the Calibration module does plug-in and has only one coaxial cable connection.) Almost all module removal will always require partial receiver disassembly and almost always some wire unsoldering. With good documentation, the process isn't too difficult but be prepared - none of the major modules just "unplug" or "plug-in." Luckily, most manuals are available online. The RA-17 and RA-117 are comparable to the R-390A or the 51J Series for level of circuit complexity. Comparing ease of disassembly for component replacement, the R-390A is physically much easier to work on because the modules are "plug-in" and easy to remove (well,...except the RF module.) The other comparison, the 51J/R-388, is not modular and some areas under the chassis are physically difficult get into for component testing or removal (the crystal oscillator is almost impossible to get into.) However, if you've successfully gone through R-390A or 51J receivers, the RACAL RA-17 will require about the same level of rework ability.

RA-17 Receiver Alignment Overview

General Info - Alignments should follow the basic idea of the instructions in the manual. The 40mc BP filter and the 37.5mc BP filter have several adjustments and the 40mc BP filter requires a spectrum analyzer with sweep generator (laboratory equipment) for accurate alignment. Originally, where there was access to the 40.0mc BP filter adjustments, metal labels were installed over access holes that warned "Do Not Touch - Prealigned." It's very common that these 40.0mc BP metal labels will be found missing for obvious reasons. Thoroughly read and reread the instructions before "diving into" an alignment. You should thoroughly understand how the RA-17 works before attempting an alignment. Adjustments are located almost everywhere,...top, bottom and side of the chassis. Expect to have to move the receiver on its side or rotate it around the bench for much of the alignment. Probably placing the receiver on its right side would provide access to all inputs and alignment adjustments with as little moving as possible. Since a R-load needs to be connected into the IF circuitry during alignment, the larger bottom cover needs to be dismounted. The R-load is necessary because during the alignment without the load the IF output presents a "broad double-peak" and if adjusted to one of the peaks would have the IF slightly misadjusted and trying to find the exact center of the "dip" (the center between the two peaks) is difficult. The 4.7K eliminates the "double-peak" for an accurate adjustment. Naturally, the top cover also needs to be off for the alignment but all tubes and circuits are shielded even with the top cover off. The left-side panel will have to be dismounted for access to the 100kc LC filter and access to C-136 (VFO-2 trimmer.) The manual's alignment procedure does mention that the equipment used should be of laboratory-quality and that a digital frequency counter will be needed for accurately aligning the Crystal Filter. Using a modern synthesizer-type RF signal generator would eliminate the need for a separate DFC (for checking the signal generator frequency.) The manual also indicates that using the Carrier Level meter as an output indicator is specified in most of the alignment measurements (using the uA scale set to 100 - half scale on the meter) but a VTVM can also be used depending on what exactly is being measured (the output of the 37.5mc BP filter requires a VTVM with RF probe.) Also, a heterodyne frequency meter is specified for alignment of the 40.5mc to 69.5mc VFO 1 but, nowadays, one might think a digital frequency counter could be used instead but there are difficulties involved when trying to measure VHF sine waves with regard to triggering points on the sine wave causing frequency readout errors. A modern oscilloscope that can sweep at 0.1u sec per division (10mHz per division) would show slightly over 4 sine waves per division at 40.5mc and just under seven sine waves per division at 69.5mc,...some modern 'scopes will also measure frequency in a digital display somewhere on the screen. Even a modern VHF receiver with synthesized digital readout of the received frequency could be used for the VFO 2 tracking alignment.

RACAL's Alignment Procedure Caveats - Be prepared,...the alignment procedure is fairly complicated and it isn't exactly a "step-by-step" procedure. The writing is mildly convoluted making it somewhat difficult to follow. First, don't be in a hurry to complete the alignment. Read through the alignment procedure several times as this will help you to understand what is trying to be accomplished (this is assuming that this is going to be the first RA-17 you've aligned.) Like many 60+ year old instructions for electronic devices that require periodic alignment, the RA-17 manual was written at a time when most test equipment used vacuum tube circuitry and analog dials. Consequently, many of the steps described in the alignment procedure are due to the limitations of the test equipment of the time. The heterodyne frequency meter for VFO 1 adjustment is a good example of a type of test equipment that just isn't used anymore (although this might just be the best way to align the VFO 1.) The procedure has some expected measurements and specifications added into the instructions that are helpful but do lengthen the instructions quite a bit.

Extra stuff needed,...a 4.7K resistor is required for loading the first IF transformer while it's being aligned. Also, a 1K resistor load is required for the Ant Tune (Aerial Tune) adjustments and a 0.1uf coupling capacitor and a 12pf shunt capacitor are specified. The 12pf shunt is for an external multimeter of a specific type called out in the procedure. The shunt might not be required depending on what equipment is actually used for the measurement. I used a military HP VTVM using a RF probe measuring RMS (Type ME-26D) and the shunt wasn't needed. 

Weird stuff,...the 100kc LC filter has two unconnected wires inside the shield-box that must be soldered to terminals inside the filter box. These wires connect the necessary built-in damping resistors for alignment (the shield-box must be remounted for alignment) and the wires are disconnected after the 100kc LC filter alignment is completed. The shield-box is on and off four times during the alignment! And, unexpectedly, to be able to dismount the 100kc LC filter shield-box, the left side panel of the receiver has to be dismounted! Two access slots are provided on the left side panel but why they are present when the side panel has to be removed is just one of those "weird" things. The Crystal Filter can be aligned through a single access slot provided on the left-side panel but dismounting the left-side panel has many advantages during the alignment and is highly recommended. The Crystal Filter alignment will require either a digital frequency counter to set an analog signal generator accurately (actually mentioned in the later RA-17C-12 manual) or using a modern synthesizer-type of RF generator. Using a synthesizer makes setting up the Crystal Filter very easy since the frequencies are just "punched in" and the accuracy is unparalleled.

Pains and Oddities - To further complicate alignment, ALL of the adjustments are referenced by using the component ID, e.g., adjust L22 or C127, etc. and then there's no "specifically for alignment" locator drawing pointing out just where the particular adjustments are. You have to use several of the component identifier photos/drawings to locate the actual adjustment specified. Even then not all of the identifiers are shown. For example, the 4.7K R-load is referenced to the inductors of the IF transformers but the actual connections to the IF transformer terminals aren't mentioned and the terminals aren't identified on the schematic either. You have to trace the wiring (on your first alignment anyway) to actually know which terminals to connect the 4.7K R-load to. Another problem is C-195B C-trimmer on the second IF transformer,...it's not even shown on the Mark II schematic but it is present in many Mark II receivers. Additionally, C-195B isn't shown on the component identifying drawings. One has to just assume that the unidentified C-trimmer on the second IF transformer must be C-195B. While carefully doing the read-through of the procedure (several times,) locate ALL of the adjustments and mark them in pencil on the component photos/drawings and add pencil notes to the procedure in the manual (again, assuming this is your first RA-17 alignment.) That way you'll know where all of the adjustments involved are located before you start the alignment. This is just one of several tedious pre-alignment steps that are required on just about any complex receiver on the "first" alignment. Some receivers will have identifying markings on the adjustments on the chassis from technicians that performed alignments in the past. They are helpful most of the time.

Since the receiver is on its side for most of the alignment, clip leads can be used for connections under the chassis (ball-clips work best but alligator clips also work.) For the 4.7K R-load needed, making up the resistor with wire clip leads in advance will help (although it's only needed twice.) Using tube extension sockets will make signal input connections much easier. Most of the tools needed are standard alignment tools. Non-conductive types might be needed in the front end of the receiver. If the receiver has been totally disassembled for a complete rebuild, the manual's alignment procedure does provide instructions for setting up the mechanical alignments necessary.

Other Complaints - Due to the VHF conversion that has VFO 1 operating from 40.5mc to 69.5mc and the 40mc BP filter along with the 37.5mc BP filter, one might find that a lot of their vintage test equipment just can't operate at those frequencies. Although I found that my General Radio DFC worked great for measuring the 1.000mc crystal oscillator, it was totally confusing to use for measuring the output f of VFO 1 (just adjusting the trigger would then result in a different frequency display.) The VTVM ME-26D worked fine for HF RMS measurements when using its RF probe. For all frequency requirements below 20mc, the HP synthesizer worked great. For higher frequencies I used a FNIRSI Oscilloscope with 50mhz synthesizer signal generator combo and that worked fine. What would be nice additions (for me) to the test gear for aligning RA-17 receivers would be a good spectrum analyzer with tracking sweep generator that can operate at MW up to VHF, a separate modern sweep generator that operates up to 100mc and a VHF receiver with synthesizer digital frequency readout capable of covering 30mc up to 100mc. 

Alignment Procedures Should be "Marked Up" - The RACAL manual's procedure doesn't "read" like the typical U.S. Army Technical Manual (TM) instructions that were written for Army radio technicians to easily follow in a "step-by-step" method and accomplish the alignment in a "quick and timely" manner. RACAL's alignment instructions aren't as difficult to follow as, for example, the Hammarlund SP-600VLF receiver's "impossible-to-implement" instructions but the RACAL instructions definitely are not easy to casually read-through unless you understand exactly how the receiver is designed to work. The instructions are not written to allow a "quick and easy" method of alignment either. But, the more you read the procedure, the easier understanding it becomes. Don't be in a hurry, re-read the procedure several times and go ahead and make pencil notes in the manual. It all helps on the "first" alignment.

Modern Test Equipment - Is it possible to go through the RACAL procedure and eliminate the 60 year old test equipment references and substitute modern equipment? Using modern test equipment will usually be easier to adjust, much more accurate in the set up and give better results for the alignment. The Internet has quite a few examples of "updated" procedures with hints on using modern equipment for aligning the RA-17. Some of these instructions are selected portions of the complete procedure and most of those are about the 40mc BP filter, the 37.5mc BP filter or tracking VFO 1 with suggestions on aligning these difficult to adjust circuits. In fact, the RACAL manual's procedure says that the 40mc BP filter shouldn't be tampered with unless you have a sweep generator and spectrum analyzer which, 60 years ago, nobody had except for commercial outfits and the military. Nowadays, good quality spectrum analyzers with built-in tracking generators are very accurate, very small but also pretty expensive. The are quite a few low priced DSA - Digital Signal Analyzer - devices that might work in combination with other equipment. These devices are made in China and are basically a two channel oscilloscope with a built-in RF synthesizer signal generator. There's one Chinese tablet-sized 2 channel oscilloscope with RF synthesizer generator that claims to also be a spectrum analyzer, all of about $295 (the DPOS350P.) However, the RF synthesizer generator in this particular device unfortunately doesn't have a sweep function that works with the spectrum analyzer so sync'ing becomes impossible. This type of spectrum analyzer is referred to as FFT or Fast Fournier Transform which transforms the amplitude vs. time of the 'scope to amplitude vs. a frequency spectrum. In other words, FFT only shows basic signal strength over a wide frequency range which is good for harmonic content types of observations. If you have (or have access to) a spectrum analyzer with tracking generator then alignment of the 40mc BP Filter becomes possible (although check performance first using the spectrum analyzer, adjustment may not even be necessary.) If you don't have the proper test equipment then don't attempt the 40mc BP filter adjustments. It's possible to check the 40mc BP filter response with a signal generator but it's not a very accurate procedure and certainly shouldn't be used for adjustment. The use of a synthesized signal generator also makes the entire alignment much easier. When the RA-17 was built, RACAL had test jigs with oscilloscope displays and built-in sweep generators with markers for adjusting the filters and it was still a fairly time-consuming process. Modern test equipment is probably ten times more accurate than what RACAL used 60+ years ago. If you can afford it, modern test equipment will ease all of the adjustments quite a bit provided you know how the RA-17 is supposed to work and you know how to use the modern test equipment and you know what you're looking at. Check www.radiomuseum.co.uk as Allan has multiple pages about working on various types of problems with the RA-17 and details on aligning the RA-17 BP filters and VFO 1 with a modern equipment. Lots of photos too.

Summary - Like any complex receiver, the "first one" worked on is your "learning experience," any that follow become more familiar and are almost always much easier to get through. If you've successfully reworked R-390A or 51J receivers, then you'll sort of know what to expect in a RA-17 or RA-117 receiver.

UPDATE NOTE: Most of what is written above is based on the RACAL manuals for the RA-17 Mark II and the RA-17L and the RA-17C-12, both Mark III receivers. I've found a British Military Technical Handbook for Field and Base Repairs for the RA-17 Mark II and for the RA-17L Mark III (hereafter referred to the Handbook.) Since this handbook was written for technicians (I think) it seems to be very direct in describing each of the steps necessary for alignment which helps to clarify the tasks. It's basically the same information that's in the RACAL manuals but presented in a somewhat easier-to-follow format. However, there are absolutely no component location drawings provided so the user has to have component location drawings/photos from another source (the RACAL manual, probably.) The Tech Handbook is from 1967. I found this manual/handbook on the Internet while searching for alternate manuals for the RA-17 Mark II receiver (free download from VMARSmanuals.co.uk.) As mentioned, the handbook covers both the Mark II version and the L Mark III version. This Handbook should be used along with the RACAL manual - it can't be used exclusively because of the lack of component location information.

RA-17C Mark II - Details for a "First" Alignment

Performing the "first" alignment on a "new-to-you" piece of equipment is always time consuming. If the piece of equipment is complex then even more time is needed. Before the alignment can be performed you have to become familiar with the design. That requires thoroughly reading and understanding the circuit descriptions in the manual. Next, the alignment procedure has to be thoroughly read (and reread) to understand exactly what needs to be accomplished. Sometimes procedures are easy to understand, even for complex equipment. Other times, a poorly written procedure complicates what could otherwise be an easy task. Going though the "first alignment" will usually reveal if different types of test equipment (perhaps other than what you have) are going to be required. But, ALWAYS,...the "first alignment" is a tough one that takes a lot more time to accomplish because so much has to be "learned" before you can easily perform and successfully complete the tasks. The "second" alignment is normally much easier and the time involved much less than that "first" one. I've detailed each section of the alignment procedure with notes and observations along with a photo of the alignment adjustment location. The following section isn't an alternative to the RACAL alignment procedure,... it's just additional information that might be of some help as you work your way through that "first alignment."

Set-up - The top and bottom covers need to be removed from the chassis. The left side panel must be dismounted to access several adjustments during the alignment. The receiver is placed on its right side but will often need to be turned over for top side tube removals, signal inputs or adjustments. For the initial alignment, the 100kc IF, V12 has to be removed from its socket. The signal generator input is through a 0.1uf capacitor to the grid of V14 (pin 1, I used a tube socket extender for an easy connection.) Signal level is about 300mv RMS for half-scale on the CL meter in RF Level. IF Gain at max, NL off, MAN selected and the speaker can be turned off, if desired. Not all changes in the set-up are mentioned in the following sections. I'm calling attention to points of confusion or unusual steps that are required by the alignment procedure. At all times through the alignment, follow the procedure instructions exactly.

Showing the IF transformer base numbers

The signal generator is connected through a 0.1uf cap to pin 5 of V12 but the tube is not installed in its socket. Receiver bandwidth is set to 1.2kc and the signal generator is set to 100.000kc exactly (I used the HP 3325A synthesizer generator so it was just a matter of "punching in" the exact frequency.) The upper three C-trimmers are adjusted for maximum indication on the CL meter (in RF Level) and then the lower left C-trimmer is adjusted for peak response. This order is repeated a few times to be sure the settings are accurate. The lower right C-trimmer isn't adjusted at this time but is adjusted during the Crystal Filter alignment. Next, the shield-box is again dismounted and the two wire jumpers are unsoldered from their respective terminals and tucked into the space between the trimmer and the fixed capacitors. Then the shield-box is remounted.

The six bandwidths are then tested for their proper response by varying the input frequency (the synthesizer can be "step adjusted" by just holding the "modify" button in) and watching the CL meter for the indicated -6db points and -66db points. The 0-200uA scale on the CL meter can be used to estimate the -6db points.

The adjustments on this 100kc LC Filter were "right on" except for C-148 (lower left C-trimmer) was just slightly off. Set-up for this adjustment takes longer than the actual setting of the trimmers because of the need to remove the shield-box, solder the jumpers, install shield-box, adjust, remove shield-box, unsolder jumpers, reinstall shield-box. Total time,...about 30 minutes.

Shield-box off showing the wire jumpers (not connected yet)

The Handbook procedure is actually much easier to use for the Crystal Filter adjustment since it has a table that shows the input frequency, which Phasing trimmer to adjust and the expected frequencies to expect for a bandwidth of -6db (-66db is also shown but this is impractical to measure accurately using the CL meter.) The RA-17C-12 procedure has all of the same data but written out in sentences within one paragraph which is a little more difficult to follow. The RACAL Mark II manual is like the RA-17C-12 procedure.

I think the Crystal Filter section is burdened by listing too much data for testing specifications. The table used in the Handbook just "lays it out" making it very easy to follow. I just measured the -6db points of the selected bandwidths to see if I was close with the adjustments I made. The chart shows the expected frequency input that results in the -6db roll-off for each bandwidth. Very easy. As mentioned in the procedure, extreme accuracy of the input frequency is necessary to measure the bandwidth. The manual specifies using a digital frequency counter with the signal generator,...and this will work fine, if you have to use that method. I was using a HP 3325A synthesizer signal generator, so all I had to do was "punch in" the frequency desired and I was ready to go. For varying the frequency for bandwidth measurement, the synthesizer can easily "step" up or down in frequency by any increment that is programmed in. Very quick and easy.

The shield-box is dismounted to show that the actual crystals are in glass tube envelopes and have tube shields installed.

This VFO-2 tracking alignment is pretty easy. I corrected the 8kc error I had in about 15 minutes of adjusting. Just as a note,...since you are handling the film strip with your fingers be sure your hands are clean and dry. Also, be very careful and gentle with the film strip as you create the "loop" to derail the sprocket holes from the drive sprocket teeth. It all went okay on this receiver but if, for some reason, the film strip is dried out and brittle problems could easily happen.

This alignment resulted in the 0 to 1000kc range tracking perfectly. But, in checking the exact linearity, I found that about half of the 100kc test points had errors from 2kc up to as much as 4kc. This requires gaining access into VFO-2 and bending plates on the air variable capacitor to correct the linearity problems. It's really not bad enough to go through that effort since the dial index can easily correct for any calibration needs.
 

Located between the two IF transformers and behind V16, C-136 is difficult to access. And, the left side panel must be dismounted.

Many times, these access holes will be covered with yellow tape. Sometimes holes are punched through the tape or the tape is removed entirely. The tape can be carefully removed for the alignment and then reapplied when it's finished. The tape can be secured with a small "dot" of Duco Cement, if necessary.

This particular adjustment is very easy and doesn't take long to accomplish.

NOTE: Date lag to the next step was due to waiting for delivery of the FNIRSI DPOS350P.

2mc-3mc Mixer adjustments are usually covered with yellow tape, the C-trimmers lower (larger) holes, L-adj upper (smaller) holes

With a very accurate signal generator input on TP-1 (I used the FNIRSI synthesizer since the HP's max f is 20mc) and a HP VTVM (ME-26D) with RF probe monitoring TP-3 (measuring RMS,) exactly 37.500mc is used as the frequency input and the appropriate trimmers are adjusted in order for a peak output as shown on the VTVM. I went though the adjustments twice just be sure they were at maximum (all of the trimmers were very close and only required a slight adjustment.) In the RACAL procedure, a specific multimeter is called out and this multimeter might need to have a 12pf capacitor shunt. The procedure indicates that this capacitor shunt might not be needed. I wasn't using the specified multimeter (a British make) but I used a Military HP VTVM ME-26D with RF probe instead (without the shunt.)

The 37.5mc BP Filter is tested by accurately measuring at what frequency the -6db points are by tuning the signal generator both above CF and below CF, noting the frequency where the -6db points are located. The mean difference should be 37.500mc +/- 20kc and the -6db bandwidth should be within 229kc to 300kc. Total time was actually more for the set-up than the adjustments,...about 20 minutes total. 

Photo Right - Notes: L2, C2 (marked "XTL-C2") access holes are shown for 1.000mc Crystal Osc. adjustments and C7 (marked "X") is part of the VFO 1 adjustment.

The 37.5mc BP Filter adjustments are under the holes in the left vertical shield and in the lower horizontal shield. Rear of the receiver chassis is to the left. Large shield access holes are marked L2, C2 and the "X" marks C7.

Warning Tags on the 40mc Band Pass Filter. The front of the receiver chassis is to the left.

Usually, the 1mc oscillator doesn't need to be adjusted and shouldn't be adjusted unless an accurate DFC is available. My General Radio 1192-B DFC (the greatest resolution DFC I have) can read 1.0mc out to seven digits or 1,000,000hz. Okay for a check but not to do any adjustments unless, like me, you find the crystal oscillator is running at 1,000,043hz. C2 could only "pull" the frequency down to 1,000,023hz. Oh well, that will have to be close enough (~0.002% error.)

The 1mc crystal and V12 have to be removed. Then a pick-up loop for the DFC input is wrapped around the base of V7 (shield removed, pick-up loop is just a few turns of insulated wire connected to the measuring device input coax cable center conductor.) The two adjustments, L36 and C77 are both hidden by the Calibration Module. Luckily, the Cal Module is very easy to unplug and place over to the side, out of the way. Loosen the knurled nut and remove the thumb screw then the Cal Module can be unplugged. With this module out of the way, now L36 and C77 are easily visible. L36 has a lock-nut on the adjustment. L36 is adjusted for 40.50mc with the MC dial centered on "0" and then C77 is adjusted for 69.50mc with the MC dial centered on "29." Repeat the adjustments to confirm that tracking is correct. Next, check each of the MC numerals 0 through 29 to see that the correct frequency (a 1.00mc change with each numeral) happens at about the center of the index line for the numeral.  >>>
 

With the Calibration Oscillator unplugged and set to the left, VFO 1 adjustments L36 (left) and C77 (right) are located just in front of the tubes V5 and V7

A check of the VFO 1 output will require removal of V7 and checking TP-2 with a VTVM with RF probe (or an oscilloscope.) Verify that the output is about 1.5vrms on all 29 positions of the MC dial. C7 can be adjusted (under the chassis in central large shield area with access hole) for equal response between "28" and "29." Replace V7, V12 and the 1mc crystal. A quick test of the calibration,...with an antenna connected and the KC dial set to "0" I set the MC dial to 10 and when the receiver came up, WWV 10mc was right on. I went to 15 on the MC dial and WWV 15mc was right on. I went to 20 on the MC dial and WWV 20mc was weak but was "on frequency." I checked 20M and 40M. I heard a CW signal on 20M and a weak SSB station on 40M. I was using an indoor loop antenna so that's why the "lack luster" performance. Everything was where it should be so the calibration is much better than it was at the start of this alignment.

This has been the most time consuming part of the alignment. This was mainly because of using an inadequate device for the first attempt. Using the capabilities of the DPOS350P allowed for an accurate set-up that took about 30 minutes to accomplish.

Mark III Only - The Mark III actually uses RF transformers with primary windings and has both L and C trimmer adjustments to accomplish the correct span for each band's tuning range. The procedure for alignment is much more involved than the Mark II adjustments. Removal of vacuum tubes, 1K load resistor and several adjustment steps that are described in the manual. For the most part these adjustments are to confirm that the "AE Tuning" works correctly and has sufficient adjustability to span the tuning range that's selected.

AE Range adjustments are slug-tuned inductors on the RA-17C Mark II

* "Shotgunning" is a technician term for wholesale component replacement to correct a problem rather that troubleshooting to the component level. However, when you're dealing with components that are 60+ years old, sometimes there are so many defective parts, replacement of all paper dielectric capacitors and checking all resistor values is faster and does get the device working, which is the object.

The Calibration Oscillator can be unplugged from the side of the VFO 2 if repairs are required

A couple of reception tests showed that even though most of the "as found" adjustments had seemed close, they must not have been that close in reality. Immediate sensitivity improvement was very apparent. Once I had the receiver connected to the Collinear Array a multitude of 20M ham stations were received including several extremely strong SSB signals. 40M was also alive with signals. Even very weak 40M signals could be intelligibly received (in other words,...not lost in the noise.) The receiver seemed to have much better sensitivity and a lower noise floor than before. The calibration on the MC dial was a vast improvement with the index never being outside the width of any of the dial MC indicator lines. The KC dial tracking was exact for 0 to 1000kc alignment but each 100kc had some linearity errors. Moving the KC index easily could allow accurate calibration. Selectivity seemed better than before with 8kc bandwidth and 3kc bandwidth performing as would be expected. The 1.2kc, 750hz, 300hz and 100hz bandwidths were much better than before so now 1.2kc can be used for SSB in heavy QRM situations and CW is easy to copy even in the 100hz position. The AE Tune also was able to "peak" any received signals. I found that the SSB signal levels on 40M were so strong that it was easier to just switch in the Attenuator on strong signals and switch it out for normal signals. For receiving SSB, the IF GAIN must be reduced to clear up the moderate distortion on strong SSB signals. I had the AVC turned off and the AF GAIN advanced. This resulted in the best sounded SSB signals. Usually, one station would be intensely strong and the other station weak. This was where the Attenuator was handy to use. I just switched in enough attenuation to clear up the distortion on the strong station and then switched to no attenuation for the weak station. Heard 8P6JJ on Barbados on 20M. Heard the Chinese Mainland Marine CW Marker Beacon XSQ on 16.852mc at about 1600hrs PDT. This is one of the "test signals" I use since it's not a very strong signal and is a fairly high frequency. The RA-17C had no trouble receiving XSQ. The other "test signal" is Trenton Military VOLMET (Trenton, Ontario) on 15.035mc USB, this signal provides voice weather reports for Canadian pilots. In early August, conditions must have been very poor as I couldn't hear Trenton at all. By August 19, 2025, Trenton Mil was coming in very well, indicating that the RA-17C is receiving whatever is actually propagated to this QTH. Comparing the RA-17C Mark II to the RA-17L Mark III, the two receivers are very close in performance capabilities now.

I'd have to conclude that even though the alignment adjustments needed seemed slight, the performance level was greatly improved with this alignment. The RA-17C is now performing very close to how the RA-17C-12 performs. The main difference between the two receivers is now the +/- 3kc BFO (with a non-vernier adjustment) and the 8kc maximum bandwidth (no 6kc or 13kc bandwidths like the Mark III versions have.) I can't wait to put the RA-17C-12 on the bench and see how that alignment goes,...a lot faster I'd imagine.

Other RA-17 Refurbishing Info

Solder Joints - There is a popular belief that the "notorious Racal solder joints" are more common than in other electronic assemblies. However, one does find these types of "soldering problems" in all types of electronic devices regardless of manufacturer. There are "cold solder joints" that just need to have the solder "re-flowed" to correct. My RA-6117 had a cold solder joint in the 1.6mc to 100kc conversion module that caused intermittent operation. The repair required removal of the converter module (not as easy as it sounds) and going over all of the solder joints. This corrected the problem for several years. A few years ago the problem did reoccur and seemed to be caused by the Oscillator tube socket. Spreading the tube pins for better contact in the socket seems to have corrected the reoccurrence of the problem for now. There are also "unsoldered joints" that are more common that one would think. Recently, my RA-17C-12 had no audio output when switched on. Interestingly, just two tubes, the filaments of the 12AT7 and the 6AQ5 weren't illuminated. I checked the tubes and they were good. The problem was that just these two tubes have their tube heater voltage supplied by just one wire that's routed back to the power transformer. On the transformer turret-slot for the 6.3vac there were two "stacked" wires, a 16ga wire (main source for the tube filaments) on top and one 20ga wire at the bottom of the turret that supplied the heater voltage to the 12AT7 and the 6AQ5. To my complete surprise, the bottom wire had never been soldered. The solder wasn't flowed properly and only the 16ga wire had solder. Just "a little tug" on the 20ga wire and it came out of the turret-slot. How did the RA-17C-12 work all these years? The small gauge wire is hard-drawn copper which is very stiff and resistant to bending. The "spring" nature of the wire must have put a slight pressure on the unsoldered joint that apparently was enough contact for the two tube's heaters to function for 60+ years. Dirt and oxidation must have worked into the joint causing it to become non-conductive. The 16ga wire had to be desoldered to allow repositioning it to the bottom of the slot and then resoldering the 20ga wire at the top of the slot. Proper solder flowing of the joint corrected the problem. As with any "half century old" electronics gear, you will find problems, even on functional units. It's not uncommon for some types of soldering or lack of soldering problems to take decades to surface. 

Incomplete Shielding and Over-heated Parts - A source of instabilities and birdies can be many different things for a receiver that has been tampered with and needs realignment or the receiver may have a multitude of defective capacitors if it was stored in a poor environment. Many of the problem areas are listed in the manual with specific components involved and how to test what exactly is "leaking" into the circuit and causing the instability. An easy fix sometimes is just poor contact on the shielding under the chassis. There are seven smaller shields that fit into place under the chassis. Six are providing complete shielding of the receiver front end circuitry and one shield is over the RF attenuator. These must be present and must be making good contact with the chassis to prevent feedback and oscillations from occurring. Other instabilities can be caused by the components in both of the VFOs where the underside is completely sealed with no way to dissipate heat build-up. The heat, especially in military 24-7 operations, seems to have sometimes caused component drift in receivers with a military history of ownership. Unfortunately, getting into the underside of either VFO requires some serious disassembly of the receiver. A poor-condition receiver is going to need total disassembly anyway for a complete examination and probable refurbishment.

Paint Touch-up - Matching front panel paint color - I had some minor chips and scratches on the RA-17L that would look better if touched-up. I used Testor's Model Lacquer Paint. It's the paint found in hobby stores in the very, very small glass jars. I used Flat Green - it's a light olive green (~1 part) mixed with Gloss White (~10 parts) and a just very, very small "touch" of Gloss Dark Blue and an even smaller "touch" of Flat Black. The gloss mixed with flat will dull down the gloss to match the panel better. Do your matching in as much natural light as possible. I mix by eye to get the general tint and then get a close match by actually touching up an edge of the panel. Wipe off your mixes until you get it close. Remember, the paint will dry slightly darker so have your "wet" match just a slight shade lighter. Wait for about ten minutes for the paint to dry to judge how close of a match you have. If it's good, then proceed with touching up the panel. This procedure is just for small nicks and chips, not to cover up major problem areas.

Cleaning Knobs with Collett Grips - To remove the smaller knobs that use the collet grips, loosen the hex-socket screw about one-half to three-quarters of a turn. Using a small plastic-head hammer, just tap the front of the hex-socket screw lightly one tap. This should loosen the collet grip for knob removal. The large tuning knobs usually are easy to remove by just loosening the hex-socket screw. If you plan on doing a soak cleaning of these types of knobs be sure to disassemble the collett grip from each knob. There is a spacer in the smaller knobs that is made of garolite which is a fiber board type of material that might be damaged by soaking. The larger tuning knobs have a plastic spacer. The assembly consists of a brass collett, then a spacer and then the tightening screw. When the knobs are apart then the collett can be cleaned as they usually are somewhat oxidized even though they are brass. Use a brass "tooth brush" to clean off the oxidation. After the knob body is clean and dry (after its soak) then reassemble the collett grip in each knob. Be very careful to not over-tighten the hex-socket screw when reinstalling the knob or it's possible to crack or break entirely the front part of the bakelite knob. This screw should only be tightened enough to prevent the knob from slipping on the control shaft. If the knob is broken, it is possible to use epoxy and carefully repair the broken piece. The "epoxy-glued" knob would then have to be used on a control that doesn't require a "firm grip" - like the A.F. Gain control.

RF Amplifier Tube CV5531/ECC189/6ES8 - This is the dual triode tube used in the Mark III versions of the RA-17. More than once I've heard about the failure to this tube happening more often than one would expect. The symptoms of the typical failure are weak signals. You can hear the stations a little but the CL meter doesn't move in RF mode of measurement and the Antenna Tune/Preselector won't peak the signal. I my case, the tube tested as triode g-7,p-6,c-8 showed a hard short. The second triode was okay. The tube was a Raytheon ECC189/6ES8. I replaced it with a NOS GE 6ES8. The new tube corrected the problem. I wouldn't have mentioned it but I've heard of this same problem from other RACAL enthusiasts. NOTE: I've had the 6ES8 fail in my RA17L again. The NOS GE 6ES8 lasted about one year or maybe just a little bit longer. The failure is the apparent hard short mentioned above. My stock of NOS tubes are all GE 6ES8. Maybe a different manufacturer would have better longevity, but I doubt it (I don't know how long the Raytheon ECC189/6ES8 had been in the receiver when it failed.)

Dial Lamp -  If you have to replace the dial lamp, the original was 6-8v 200mA and when illuminated produced a distinct yellowish color to the MC film strip and the KC dial (see photo directly below for the typical dial illumination color.) For a new replacement, the first inclination is to install a #40 lamp with a 6.3v 150mA rating but this lamp is "over-bright" and produces a white KC dial and very bright MC film strip. The closest bulb to the original is a #50 rated at 6-8v 200mA and 1c.p. with a screw base. #50 lamps should be easy to find. The #50 lamp is slightly brighter than the original but not to the excess that a #40 bulb would be.

Performance Opinions

RA-1217, RA-6217 Solid-State Receivers

These all solid-state RACAL receivers from the late-sixties, early-seventies, are interesting because their circuitry is almost exactly like the vacuum-tube RA-17 versions. The RA-1217 and the RA-6217 use the same type of input Antenna preselector, the same separate MC and KC VFOs, the same 40mc and 37.5mc band pass filters, the same Wadley Loop, the same triple conversion,...on and on. The KC VFO and the 3rd IF are a bit different and there's some IF conversions for external devices. The only huge difference is that these receivers are all discrete transistor circuits so the operational voltage is only -16vdc. Small in size and light in weight, the performance is first-rate, making the RA-1217 or the USA-built RA-6217 a very nice addition to the ham station landscape.

By the late-sixties, the RA-17 and RA-117 were competing with many solid-state communication receivers built by other companies. In 1967, the all solid-state RA1217 (UK) and RA-6217 (North America) were introduced. The same concept of using triple conversion with a Wadley Loop became the basis for these new receivers. Mechanical digital readout and a tuned preselector, along with a very small, light-weight package were the main features of the RA-1217 and the RA-6217. These receivers were comprised of several plug-in modules fitted to a main frame chassis. As with earlier receivers, many Racal accessories were available for these receivers and there are many inputs and outputs (and modifications) on the rear panel to accommodate all of the optional equipment that was available. The receiver operated on 115vac to a power supply module that then supplied the -16vdc to operate the circuits. At only three and a half inches in height, several receivers could be operated within a very small rack space. The RA-1217 and RA-6217 were produced up into the late-seventies.

photo left: A US Army Signals Security Agency (SSA) Radio Operator using two RACAL RA-6217E receivers and the matching panadaptor, the RA-6366B - Vietnam War Era. Note that there is another RA-6217E dual receiver and panadaptor setup behind the foremost operator. This US Army photo is from their website, https://www.army.mil   The caption reads "The Army Signals Security Agency provided direct support Signals Intelligence detachments to tactical units throughout South Vietnam."

RA-6217E Receivers in Storage Unit? - I found three RA-6217E receivers at an estate sale in Moundhouse, Nevada in October 2019. To say they were "bargain priced" would be an understatement ($10 each.) All three receivers were jammed into a shelving unit at the rear of a densely "packed" storage unit that had "floor to ceiling" slot machines, slot parts, piles of cardboard and lots of junk test equipment (and just a few radios.) One RA-6217E receiver was almost a complete unit, lacking only the top cover and the shield over the RF input section. The second receiver was about 75% complete having nearly all of its modules but missing the important IF module, the tuning knobs and the top cover. The last receiver was about 40% there, missing several modules, the meter, all the knobs, the front panel and both top cover and small bottom access cover.

Getting one RA-6217E Working - Oct 2019 - On the test bench, I checked over the mostly complete RA-6217E and, other than being filled with sand and tree leaves, it looked like it was in pretty good shape. A careful power-up had the receiver working,...sort of. Contact cleaning and manipulation of the controls got the receiver working fairly well, at least on AM, USB and CW. LSB and the tuned BFO were non-functional. The RF meter was pinned to the negative but worked in the Audio mode (these problems ended up being corrected with adjustments that were on the IF module and the meter driver board.) The audio was taken off of the Phone jack (600Z) and used to operate an eight inch 600Z loudspeaker. Audio was impressive.

Luckily, the 75% complete, "parts set" receiver had the RF module shield so that was removed and installed on the RF module of the operational receiver. The mechanical digital readout on the working receiver was in terrible physical condition but the two "parts sets" had excellent digital readouts. Replacing the mechanical digital readout is no easy task but I removed the readout set from the "75% receiver" to formulate a removal procedure (removal isn't covered in the manual.) It requires moderate disassembly of the receiver's front-end along with maintaining a mechanical alignment during the readout swap. Before starting, I tuned the receiver to 00-000 and slowly tuned further negative in MC and KC until the mechanical stops are contacted. In my case, the MC stopped at -00 and the KC stopped at -938. When removing the readouts, the MC readout is one unit and the KC is another unit. They had to be removed individually. To access the MC/KC readout requires front panel removal. Then I dismounted the meter, the meter switch, both dial lamp mounts, the fine tuning pot, the VFO switch and the harness clamp. The MC rotary switch has to be dismounted and the cover removed off of the VFO-1. Access to and removal of the MC readout was now very easy.  >>>

photo above: The completed RA-6217E SN:1162. Note that these RACAL receivers feature "right-hand" operated kHz tuning. The RA-6217 version "E" was optimized for use with a RA-6366B panadaptor and the RA-6397A external antenna preselector. This particular RA-6217-E was built in Silver Spring, MD (pre-1972.) Note the "vibro-etched" SNs engraved on each side of the RACAL label (Rcvr SN:1162 and Power Unit PU SN:883.)    2025 photo

Despite the quite different appearance and solid-state circuitry, the RA-6217 has the same basic triple-conversion, two VFOs, 40mc and 37.5mc BP filters, Wadley Loop and Antenna Preselector Tuner as its older vacuum tube brother, the RA-17.  2019 photo

Two problems remained,...LSB still didn't function (I'm suspecting the 1.6015mc crystal) but I can use the variable BFO to receive LSB, so that wasn't a crucial problem. Then there was the missing top cover. I made the top cover out of .060 aluminum. Luckily, there's no venting of any type - just a flat piece of metal. Once the aluminum piece was cut the mounting holes were measured, marked and drilled. Then the aluminum was sprayed with Easy Off Oven Cleaner (NaOH Sodium Hydroxide.) NaOH will produce a matte finish on the aluminum surface. Just spray on, let it set for a few minutes and then rinse with cold water. Don't rub, let it air dry or use a heat gun to dry. The finish will be flat or matte aluminum which is like the original. There's no need to coat the finish with any sort of "clear coat spray." The cover can be mounted as soon as it's dry - and it was. I used 4-40BH (binder head) screws for mounting and I should have countersunk the cover holes and used 4-40FHUC (flat-head under-cut) slotted machine screws. I'll change this mounting on the next "bench visit."

Photo left - shows the top of the RA-6217E with the top cover removed. The left-most module is the RF module. Upper left in the chassis and to the right of the RF module is the Power Unit and next to it, to the right, is the Control Line Filter unit (for the RA-6397A.) The long narrow modules are 40mc BP filter and the 37.5mc BP filter. Below them and to the left is the 1st VFO (MC,) to the right-upper is the 1MC Osc and below it is the 2nd VFO (KC.) To the left of VFO-2 is the thirty position switch for the MC position information to operate the RA-6397A. The long module to the right is the IF, Det, AF module. To its left are two Mixer modules. Though it's difficult to see, the chassis is a complete aluminum tub with welded corners and an almost complete bottom. There's a small access cover on the bottom side of the chassis. Most modules plug into sockets though some modules do have other wiring also. The front of the receiver has most of the components mounted to the chassis and the front panel then mounts to the front of the chassis. The meter mounts using its metal bezel through the front panel and secured at the backside of the front of the chassis which somewhat complicates front panel dismounting. Since this is a fairly early RA-6217, the meter bezel is solid on the bottom. Later versions modified the meter bezel to allow light from the two dial lamps to then also illuminate the meter.    

The other BNC connectors are for 2-3mc input and output for LF converter or panadaptor, VFO-2 input and output for diversity operations or panadapter, J102/J103 are 1mc input and output for ext.1mc standard or diversity, J3002 is a conversion 455kc output for ancillary devices, below J3002 is the BNC for 1.145mc output for diversity and below that the BNC is for 1.6mc output for panadaptor. Outputs from TB101 are marked for proper use. TB3001 is for diversity operation hook-ups providing AM DET, AM DET LOAD (tied together for normal ops) and the AGC connection.

The rear panel connections provide for a variety of accessory equipment hook-ups   2025

A new power-up test using the Collinear Array antenna went surprising well because I was able to easily receive Trenton Military 15.035mc USB VOLMET. I had been trying to receive Trenton for the past few weeks with no success using the RA-17C receiver that I had just aligned. The RA-17C was performing quite well but it just seemed that Trenton wasn't transmitting until I tried the RA-6217E. I don't think that anything is malfunctioning in the RA-17C and the inability to pick-up Trenton was probably typical summer propagation conditions (and I did check a couple weeks later and Trenton was coming in just fine on the RA-17C.) I did some more listening on the RA-6217E and on the 20M band I heard a CO (Cuba) station that was very strong (not that Cuba is that great of a DX location.) There were ham stations about every ten KC so lots of 20M activity. I also tried 40M but had to use the tuned BFO since SSB +1.5 (LSB) doesn't work. The testing was intriguing enough to foment a more thorough revisit to the RA-6217E.

1. The problem mentioned above concerns the non-operative LSB +1.5kc. A quick check determined that this is still inoperative. The 1.6015mc crystal is defective. The days of ordering a custom ground crystal are long over - at least from sources in the USA. I usually look at the hundreds of HC-6 crystals listed on eBay and I can usually find something close. In this case, I the closest I found was 1.6424mc. It might work though it's pretty far off frequency,...42kc off or about a 2.5% error! Test upon arrival. As expected, 42kc off frequency is too far off. There really isn't anything available that would function correctly and be original other than another 1.601500Mhz quartz crystal. I'll be looking for a spare RA-6217 IF module to show up sometime - and hope the crystals are still present.

2. Digital readout is about 4kc to 5kc off and can't be corrected with the CALIBRATE/FINE TUNE which is already at one end of the adjustment. This might be the result from changing out the KC (and MC) mechanical-digital readouts. A mechanical solution wasn't possible. The KC VFO needed a tracking alignment and that corrected the error. VFO-2 operates from 3.6mc to 4.6mc but just using the KC digital dial readout provides an for an accurate alignment.

3. Top cover secured with BH screws. Needs to be changed to countersunk holes and 4-40 FH UC screws. UC is "undercut" where the bottom of the FH cone is machined off to allow using the FH screw to secure thin sheet metal. This problem was easy to correct. Proper 4-40FHUC screws were harvested from the "parts set."

4. KC digital readout has some minor binding and is very noisy in operation. I never lubed the mechanism so it may require just a little drop or two of oil. This is a common problem with mechanical digital readouts that require a 90º angle drive using open gears. Tuning slowly is okay but any rapid tuning will cause "chatter" in the gears. The oil seemed to make the chatter worse. I usually don't do this but a small amount of heavy grease on the conical gears seemed to quiet down the tuning action.

5. KC and MC dial locks don't work. These should work on friction but the pair of dial locks on this receiver have always been a problem. I got the MC working somewhat but the KC dial lock doesn't really "lock" the tuning. I'm going to check the dial locks on the "part sets" and see if something can be salvaged for the KC dial lock. I found a plastic bag that I had stowed with the "parts set." It contained all of the knobs including two dial lock assemblies. These appear to be in much better condition than what's installed on the receiver. I don't know why I didn't swap them out when I had the front panel off back in 2019. Since the RACAL manual has no information on disassembly, I'll have to see if I remember how I did it six years ago. It all came back to me and installing the replacement dial locks got both MC and KC dial locks working correctly.

6. Lack of sensitivity from 16mc up to 30mc. I can inject a signal generator output into the receiver and it does respond. After swapping the RF module, I could pick up 20mc WWV, but just barely. It doesn't seem likely that both RF modules would have similar alignment problems so this lack of sensitivity must be due to something else. See "Test" section below for more information.

To remove the IF module, first I had to take off the top cover. Then the right side cover has to be also taken off. The IF module slides out the back of the receiver but there are four screws that have to be removed from the IF module's back panel. Then there's a 4-40 FH screw underneath the receiver that has to be removed. Then the small coax connector is unscrewed to disconnect the coax. Now the IF module can be removed by sliding it out the back of the receiver. Next the top-side cover has to be taken off the IF module to have access to the circuit boards inside. This cover is held in place with ten FH machine screws. Now the IF module is fully accessible, even when it's plugged into the receiver.

The first check was to make sure both +1.5LSB and -1.5USB oscillators had -16vdc routed to the circuit when the DETECTOR MODE switch was placed those respective positions. This test showed that the -16vdc was going to either of oscillator circuits when switched on by the DETECTOR MODE switch. The next test was to see if perhaps the trimmer capacitor was set to a point where the oscillator wouldn't "turn on." This wasn't the case since the trimmer could be rotated through its full range and the +1.5LSB oscillator wouldn't turn on. At this point I noticed that the 1.6015mc (+1.5LSB) crystal seemed loose in its socket. I removed and plugged it in firmly but the oscillator still wouldn't come on. Next, I removed the 1.6015mc crystal and "rang it" using a 'scope and oscillator. Although the crystal did seem to "ring" okay, the sine wave through the crystal was of low amplitude. The final test was to "swap crystal positions" putting the LSB xtal in USB and the USB xtal in LSB. The BFO non-operation then switched to the USB position. This confirmed that the 1.6015mc ("rang" at 1.6035mc) was somehow defective and the cause of the problem. I returned the crystals to their correct positions and the USB -1.5 was again operational and LSB +1.5 was non-operational.

I looked through the crystals available on eBay and found a HC-6 type 1.64346mc available. That was the closest I could find after looking through 14 pages of crystals. If this crystal does oscillate (it's 2.5% high in f.) I might be able to "pull" its frequency slightly closer using the trimmer C (very slightly.) Test when the crystal gets here.

Showing the Crystal Controlled BFO for USB and LSB. The defective crystal is on the right and is marked 1.601500mc. The two trimmer capacitors are "pulling" adjustments for precise setting of the crystal oscillator frequency.     2025

This photo shows all of the adjustments for VFO-2. The lock nut for C4 has red GLPT on it. The drive gear at the rear of the VFO has the bellows-coupler that is loosened to adjust the high end of the range. The blue ten-turn pot in the foreground is the CALIBRATE/FINE TUNE control.  2025

The dial locks mount to the front panel. Rotating the outer wheel raises the front friction plate that contacts the back of the tuning knob and "locks" its movement.      2025

Inside the RF Module showing the Ant Tuner on the left and the RF Amplifier PC board in the center. The 30mc LP Filter is far right. The Attenuator resistors are mounted on the switch. Just behind the air variable C, the antenna fuse is visible (connected to the Antenna Input BNC.) 2025

Performance - Not very much change. I can now pick-up WWV 20mc (though it's extremely weak) in WB position. I can't pick-up 20mc WWV with the Ant Tuner on the 16mc to 30mc position. In fact, nothing above 16mc can be received except for a signal generator input. The next step is an alignment (I suspect that the 40mc BP filter has had its adjustments tampered with.) While an alignment won't necessarily fix the problem, usually when going through an alignment you'll find where the problem circuits are.

Test - Since "on the air" reception conditions aren't very good in mid-August, I thought a better and more accurate test would be to run the FNIRSI DPOS350P synthesizer into the receiver's antenna input and measure the output at different frequencies. I started at 3.975mc and ran a 20mv sine wave as the input (a strong signal level to have a high CL meter reading.) I used the Ant Tune to adjust for maximum CL meter deflection. I tested 3.975mc, 5.560mc, 10.560mc, 15.560mc, 16.560mc, 20.560mc, 25.560mc, 28.560mc and 29.560mc. In each case I was able to tune the Ant Tune (when set to its proper range) for a minimum of 50% deflection of the CL meter (usually +60%.) I tried WB with similar results. I did note that on the 3.975mc test that the MC dial "3" was not centered. On all other frequencies, the MC dial numeral was centered (generally indicates how well VFO-1 is tracking. The procedure indicates the low end adjustment shouldn't be tampered with,...only the high end. However, access is provided for both C3 and L1.) After this test, it seems the receiver is actually functioning on the frequencies above 16mc. In fact, when switching to the Collinear Array antenna, I was able to receive 20mc WWV, although the signal was not very strong. Alignment of the 40mc BP filter and the 37.5mc BP filter, per the procedure, requires a tracking spectrum analyzer (that I don't have.) I can perform the alignment of the IF stages and the RF amplifier section. I've already adjusted the tracking of VFO-2 and that's very good now. VFO-1 tracks very well from about 4mc on up to 29mc. I suspect that since the 40mc BP filter and the 37.5mc BP filter adjustments are very easily accessible, someone might have tweaked one or the other of the filters,...maybe both. The 40mc BP filter is critical to performance because the output of Mixer 1 goes to that filter and the filter's output goes to Mixer 2. These are the critical conversions that involve the incoming signal conversion to VHF (39.35mc to 40.65mc,) then the Wadley Loop and with the 37.5mc input to Mixer 2 for the 2-3mc IF. The 37.5mc BP filter is a product of VFO-1 and the 1mc harmonic generator output going to Mixer 4 and provides the other signal to Mixer 2 to create the 2mc-3mc IF. At any rate, a tracking spectrum analyzer is necessary to check those important circuits and I don't have that type of spectrum analyzer (just the FFT type.)

Until I have gotten a spectrum analyzer with tracking generator (an expensive piece of test equipment) and can actually "see" the passbands of the 40mc BP filter and the 37.5mc BP filter, there's not too much more that can be done to SN:1162. It's usable as it is now but I'm sure it can be improved with an alignment when the necessary equipment becomes available.

I'll certainly revisit this RA-6217E SN:1162 when I have a spectrum analyzer with tracking generator. Until then, the receiver is basically functional. More to come,...

Interesting "Covers Off" Photos of the "parts set" (which explains the dirt)

Left: Inside the KC VFO (VFO-2) showing the brass air variable, the VFO C4 adjustment trimmer (inside the metal can shield) and the vertically mounted circuit board. Externally, note the bellows coupler connecting the air variable to the split-gear worm-drive KC tuning mechanism.   2025 photo

Right: Inside the MC VFO (VFO 1) showing the brass air variable, the circuit board and the adjustment trimmer.  2025

Note the Rockville, MD address for RACAL Communications, Inc. They moved from Silver Spring, MD in 1972 and this manual is dated 1972.  2025 photo.

Using the Racal RA-237-B  L.F. Converter

with the RA-17C-12 Receiver on MW, LF & VLF

IMPORTANT NOTE: June 20, 2025 - Back in 2019, when I performed all of the following VLF, LF and MW tests and wrote up these NDB reception logs, tuning in 20 or 30 NDBs in a 25 minute nighttime listening session was "standard reception" in the 200kc to 450kc part of the spectrum here in the rural area of North-Western Nevada. That was about five or six years ago. One wouldn't think things could change so much in such a short time period. Nowadays, I find that under the best conditions, using the best vintage equipment in "real time" listening, the average number of NDBs tuned in might be four or five, in a 30 minute listening session. NDBs are being decommissioned faster than ever before and the push for their complete elimination is unrelenting. Note the difference between the lengthy 2019 logs and the very short 2022 log (and that was just a three year period.) The fact that no USA pilots are using NDBs anymore is one factor. Airport economics is certainly another. So, when reading the following write-up on the RA-237-B LF Converter, especially the reception logs, keep in mind that this section was written several years ago. Most of the NDBs listed in the logs aren't even "on the air" anymore. It's unfortunate that this very enjoyable part of LF and MW listening has all but disappeared. But then, Long Wave Broadcasting is also "long gone." And, the 630M amateur band is just about useless since any of the minimal activity on the air is using data mode transmissions that require special equipment and computers. Sorry,...no human-sent CW signals there anymore.

The converter has shielding to the various modules but full shielding of the chassis was not provided. Racal indicated that a special cabinet could be provided to house both a RA-17 and a RA-237-B (14" x 19" opening.) 

RA-237-B SN: 371 - Servicing - This RA-237-B SN: 371 was in very good condition but it did need a little adjustment for good performance. Of the six tubes, four had to be replaced. I ended up replacing both dial lamps because one was open. I replaced both lamps with #40 lamps, threaded base 6v 150mA (actually using #50 lamps would probably be more correct since they are 6-8v 200mA.) The ANT. TUNING readout was mechanically off by about 30%. This required loosening the set screws on the fiber tuning condenser drive gear, setting the condenser to full mesh and then adjusting the readout to slightly below the low end of the scale. This really isn't "calibrating" or alignment, it just mechanically has the dial readout agree with the mechanical position of the tuning condenser. I also changed the AC input voltage primary on the power transformer from 115vac to 120vac since our line voltage here runs around 122vac. A few sheet metal dents and bends required minor body work. Minor cleaning, mostly dust. The dial drum was sluggish in changing position with band switching. A small drop of 10W machine oil on the bearing got the dial drum rotating properly.
 

photo left: RA-237-B chassis top.

Operation - With the RA-17 turned on and the RA-237-B power on, select "10kc to 980kc" with the OPERATION toggle switch. Using the Kilocycle tuning and observing the red scale on the RA-17 dial select a frequency, e.g. 350kc. On the RA-237-B, using the ANT RANGE switch, select 210kc to 500kc range. Using the RA-237-B ANT TUNING, adjust the dial to read approximately 350kc. As you near 350kc, the background noise in the RA-17 will increase and will peak somewhere near 350kc as read on the RA-237-B dial. As you search for a NDB station with the RA-17, peak the RA-237-B every 5kc or so. The converter's preselector circuit has a very high Q and the tuning is very sharp. The frequency is read directly on the red scale of the RA-17 and the approximate antenna peaking frequency is read on the RA-237-B dial. When searching for very weak signals, switch off the AVC and control the front end gain manually. This will prevent the noise level from controlling the AVC and reducing the front end gain. Manual control also is necessary during very noisy conditions. The Hi-Q ANT TUNING is narrow enough to reduce a lot of the LF reception noise in most cases and AVC can be used during good conditions and moderate signal levels. If you want to return to HF just switch the OPERATE toggle switch to 980kc to 30mc position and the RA-17 front end takes over and allows HF reception. Below are reception logs for the RA-17/RA-237-B combo from Nov 10 to Dec 12, 2019 (and reception updates out to Feb 20, 2020.)

RA-237-B SN: 371 on top of RA-17C-12 SN: N4144

Reception Log for November 10, 2019  -  2150-2215hrs PST  -  Ant 135' "T"

1. YTL 328kc - Big Trout Lake, ON, CAN
2. VT 332kc - Buffalo Narrows, SK, CAN
3. DB 341kc - Burwash Landing, YK, CAN
4. RYN 338kc - Tuscon, AZ
5. YZH 343kc - Slave Lake, AB, CAN
6. POY 344kc - Powell, WY
7. SBX 347kc - Shelby, MT
8. YXL 346kc - Sioux Lookout, ON, CAN
9. NY 350kc - Enderby, BC, CAN
10. IN 353kc - International Falls, MN
11. YWP 355kc - Webequie, ON, CAN
12. ZF 356kc - Yellowknife, NWT, CAN
13. YQZ 359kc - Quesnel, BC, CAN
14. RPX 362kc - Roundup, MT
15. YMW 366kc - Maniwaki, QC, CAN
16. ZP 368kc - Sandspit, Queen Charlotte Islands, BC, CAN
17. EX 374kc - Kelowna, BC, CAN
18. CNP 383kc - Chappell, NE
19. YWB 389kc - West Bank, BC, CAN
20. PNA 392kc - Pinedale, WY
21. YL 395kc - Lynn Lake, MB, CAN
22. ULS 395kc - Ulysses, KS
23. ZSS 397kc - Yellowhead-Saskatoon, SK, CAN
24. FN 400kc - Ft. Collins, CO
25. QQ 400kc - COMOX Vancouver Is., BC, CAN