Radio Boulevard
Western Historic Radio Museum
 

CANADIAN  MARCONI  COMPANY

CSR-5, CSR-5A Receivers

Receiver No.1 - Type CSR-5 110930-Z - Table Cabinet Version - SN: 394

Receiver No. 2 - Type CSR-5 110480-Z - Rack Mount for CM-11 - SN: 665

CMC History, Receiver Circuit Design and Construction, 6SG7 Upgrade
Paint Variations, Dial Escutcheons, Refurbishing the VP-3 Power Supply,
 Re-restoration of SN: 394, Refurbishing SN: 665,
Performance Comparison and Analysis


1944 Type CSR-5 110480-Z, SN:665, R.C.N. PATT.3AU/17 used in the CM-11 Transmitter-Receiver

The visually stunning Canadian Marconi CSR-5 receiver is just about as good-looking as a WWII receiver can get. If you're impressed with the appearance of the Marconi/RAF R1155 receiver's semi-circular, multi-colored but non-illuminated tuning dial, you'll become instantly infatuated with the CSR-5's spectacularly illuminated version of the semi-circular, multi-color dial. But what about performance? Is the CSR-5 all looks with nothing behind the dial? The Canadian Marconi CSR-5 receiver's performance and ergonomics are simply superior in every comparison to its British airborne cousin, the R1155 (though that's really not much of a comparison.) The CSR-5 is a first-class, WWII communications receiver that can be successfully used in any amateur vintage military radio station and it's an excellent band cruiser with audio that isn't too bad either. Although many of its CSR-5 sisters may now be shelf-queens, it's more than just a pretty face. The CSR-5 can easily become a favorite of just about any fan of WWII radio equipment by providing superior performance and impressive bench-presence combined with the robust construction demanded by the Royal Canadian Navy.

CANADIAN MARCONI COMPANY

A Brief History of the Company - The Canadian Marconi Company started out in 1903 as The Marconi Wireless Telegraph Company of Canada (MWT/C.) At first, the company was part of British Marconi, officially known as Marconi's Wireless Telegraph Company, Ltd., but soon, with Canadian government encouragement, MWT/C became an independent company. By 1919, after the WWI wireless ban was lifted, MWT/C created Scientific Experimenter Ltd, in order to sell wireless ham equipment. By 1922, they were selling broadcast radios. By 1925, MWT/C had changed their name to Canadian Marconi Company (CMC) and entered into a cross-licensing agreement with Canadian General Electric Co., Ltd, Canadian Westinghouse Co., Ltd and Northern Electric Co., Ltd to "protect" their manufacturing patents to the exclusion of other Canadian radio companies. At the same time in the USA the so-called "Radio Group" operated a similar "cross-licensing" arrangement headed by General Electric with Westinghouse, AT&T and RCA as the members (United Fruit Company had been an earlier member of the "Radio Group" but their radio company, Wireless Specialty Apparatus, was purchased by RCA around 1924 and became Radiomarine Corporation of America in 1927.) CMC became involved in radio broadcasting in 1925. During WWII, CMC grew as a company and produced communication equipment for the war effort. In 1953, the English Electric Co., Ltd. purchased a controlling interest in CMC (50.6%.) Sometime later, GEC (General Electric Company - UK) owned controlling interest in CMC. Since that time, CMC has had a multitude of different controlling companies, different owners and many name changes. CMC is still in business making mostly aircraft electronic equipment as "CMC Electronics."

CSR-5, CSR-5A Receivers


1944 Type CSR-5 110930-Z Table Cabinet Version SN:394 (chas.) or SN:682 (data plate)  
 Northern Electric Co. Loudspeaker (looks a lot like an RME housing with a NE property tag installed.)

The CSR-5 Receiver - Brief History - In 1943, Canadian Marconi was contracted by the Royal Canadian Navy (RCN) to build a high quality receiver for their use onboard RCN ships. The RCA AR-88 had been approved for the Canadian military as the general purpose receiver but it wasn't specifically designed for use at sea, although it could easily perform duty at sea. The RCN wanted a "designed for sea use" receiver and wanted to do business with Canadian Marconi even though RCA-Montreal built all of the AR-88LF receivers at that time. The CSR-5 receivers were supposed to be ready mid-1943 but delays pushed initial delivery to either the end of 1943 or early 1944. Around 700 receivers had been ordered and New Zealand wanted another 100 sent to them so these receivers comprised the first of the CSR-5 receivers. Sometime in 1944, the CSR-5A was introduced. This version had several small changes inside and also a few minor changes to the exterior but basically the "A" was very similar to the earlier CSR-5. The most obvious exterior difference is the SUPPLY switch and the SELECTIVITY control locations that are spaced vertically from the top of the front panel down 1.5" for the CSR-5 and down 2" for the CSR-5A. Internally, the CSR-5A replaced two of the 6SK7 tubes in the receiver with higher gain 6SG7 tubes. The table model cabinets were slightly different for the CSR-5A using rear chassis guide posts that the CSR-5 didn't employ along with different venting screens and two lid pull-holes rather than the one lid pull-hole in the CSR-5 cabinet.

The CSR-5A was built up to 1945 with most of production occurring in 1944. The total production quantity of CSR-5 and CSR-5A receivers wasn't particularly a large number,...CSR-5 authority Jerry Proc estimates that possibly 700 CSR-5 receivers were built and perhaps 1000 CSR-5A receivers were built. But he cautions that this is just an estimate based on known serial numbers and that the exact method of CSR-5/A serial number assignment is not really known. 

1944 RCA CR-91 - Camden-built AR-88LF

Shown to the right is the 1944 RCA CR-91, a Camden, NJ-built, USA-version of the AR-88LF. The AR-88LF was exclusively built in Montreal, QC, Canada. During WWII, the CR-91 was produced at the Camden, NJ plant in small quantities to supplement Montreal's AR-88LF production. The AR-88LF, the CR-91 and the CSR-5 all tune the same approximate frequency range of 80kc to 550kc and 1500kc to 30mc.


The WWII RCA CR-91, the Camden-USA version of the AR-88LF
CSR-5 Receiver No.1 is shown in the photo to the above-left. It's a Type CSR-5 110930-Z (table cabinet type) with the serial number "394" (on receiver chassis - photo further down this page) or "682" (on cabinet data plate shown in photo to the right.) It was rebuilt cosmetically and electronically several decades ago with no regard for originality whatsoever. The panel and cabinet color are a much darker color (other than black) than was normally used on CSR-5 receivers and the white silk-screened nomenclature on the panel uses a font that isn't like the original lettering and is too small in size. I went through this receiver electronically about six years ago and now, functionally, this CSR-5 receiver works very well. The CSR-5 data plate shows the 110930-Z identification as a table model receiver. The serial number of 682 doesn't match the serial number on the receiver chassis, 394. There are several possibilities for how this mismatch could have occurred.

CSR-5 Receiver No.2 is a rack mount version shown in the header photo above. Its chassis serial number is "G0 H50" and that's punch-stamped on the rear chassis. The data plate was normally mounted to the rear of the top cover and, since the top cover is now gone, the actual serial number is unknown,...or is it? I found that "665" had been punch-stamped on the inside of the right side-panel. Almost certainly, "665" is the actual serial number. A metal tag indicating "R.C.N.  PATT. 3AU/17" mounted on the left side panel. If the CSR-5 was used as part of the CM-11 receiver-transmitter combination, it would have a smooth semi-gloss light gray panel and the R.C.N. PATT.3AU/17 tag mounted on the side panel (see photographs to the right for serial number and RCN tag.) This receiver is virtually all-original,...well, original if the RCN depot reworking and refurbishing from WWII up to the 1950s or 60s is still considered "original."


This type of metal tag was installed on CSR-5 receivers that were part of the CM-11


 Training Radio Operators at RCN Communication School at HMCS Cornwallis, Nova Scotia in May, 1952. There were 53 radio training stations at the school each with a CSR-5A receiver.   photo from: Jerry Proc's CSR-5A webpage -  www.jproc.ca/marconi/csr5a.html

 Where Used - The CSR-5 was used for several purposes, even though its intended primary use was supposed to be aboard ship. Shipboard use required extensive shielding in the receiver to allow its operation along side other equipment without causing interference or radiating the LO from the antenna and for the receiver's reception to not be interfered with by the operation of other shipboard equipment. The CSR-5 apparently met the <400pW LO to antenna requirement. But, the RCN found other uses for the CSR-5 such as surveillance and enemy signal intercept monitoring during WWII. Canada's DOT had a very small quantity of CSR-5 receivers in use at some airports and for some marine port use. These CDOT CSR-5s were painted black crackle or black wrinkle finish with stenciled nomenclature and were probably delivered directly to the Canada DOT from CMC. Most CSR-5 receivers weren't installed onboard ships until the early 1950s. The later version CSR-5A was mostly built after the initial contract of CSR-5 receivers although it's possible that the production overlapped somewhat (probably mid-1944.) It's also possible that some CSR-5A receivers were built up into 1945 but the majority of receivers were built in 1944.

Design and Circuit Details - The CSR-5 receiver's frequency coverage was in two sections, 75kc to 530kc and 1.5mc to 30mc, in six bands. Each scale on the large illuminated dial was color coded. There are two known variations that swap the colors assigned to Bands B and A and that might coincide with the CSR-5A introduction. Also, on most dials the Mauve color, light purple, usually assigned to Band B, is actually gray. Also, later dial scales, probably "As" only, have double lines with indexes while the earlier dial use a single scale line. The CSR-5A replaced two of the 6SK7 tubes with 6SG7 tubes for the 2nd RF amplifier and 1st IF amplifier as the major change (6SG7 tubes have twice the transconductance when compared to 6SK7 tubes.) There were many other minor changes in the cabinet vents style, power connector orientation, dial scale index style, etc. Many of the earlier CSR-5 receivers were upgraded by the RCN with the 6SG7 tubes although the actual upgrade only consisted of installing the 6SG7 tubes and replacing a few biasing resistors. However, the manufacturing silk-screen was not changed for the chassis, so many CSR-5A receivers will have hand-modified 6SK7 to 6SG7 tube identification. Rubber stamped ink was normally used but sometimes hand-written will be encountered. In performance, the 6SK7 to 6SG7 change benefitted performance by providing better sensitivity on the higher frequencies. If the upgrade to the CSR-5 was thorough and included the bias adjustments, the "upgraded" CSR-5 will also perform  better on the higher frequencies.

The CSR-5 has eleven tubes in the receiver with two RF amplifiers, Mixer (also provides Xtal Osc function,) LO (9002 miniature tube,) two 575kc IF amplifiers (dual wave traps are in-circuit to prevent IF interference,) Crystal Filter (Selectivity positions 3 and 4) AVC/NL, Det/1AF, AF output and Voltage Regulator. The RF Gain control is a stepped attenuator type with 4db steps on the low end and 20db steps on the high end. AF output is 2 watts into a 10K Z load for loudspeaker, both Hi-Z (5000Z+) and Lo-Z 'phones (up to 1000Z) and a 500 ohms Z line audio with a separated winding with link provided for grounding the center-tap. The "Crystal" socket and "Crystal-IN" allowed drift-free operation through a crystal-controlled local oscillator function but the receiver still had to be tuned to the intended receive frequency for the RF and Mixer stages to be "in tune." The CSR-5 Crystal socket only allowed using the large style crystal but the CSR-5A used a different socket that had three receptacles that allowed using either the large crystal types or the standard FT-243 with 0.5" pin spacing. When installed onboard ships, a dual shock mount system attached to each side of the cabinet. The two shock feet on each side were mounted to an angle bracket that was bolted to the lower part of each side of the cabinet. To save bench space sometimes the shock feet were mounted directly under the cabinet.


Chassis top of CSR-5 sn:394


Under the chassis of CSR-5 SN:665 with the two bottom covers removed showing how a virtually all-original receiver should appear.

 Panadapters - There was a Panoramic Adaptor connection provided on the chassis using a screw terminal connection. Since a panoramic adaptor could be used with the receiver, a carrier level meter wasn't required. The panoramic adaptor had to have an input frequency range that would accept 575kc but the manual never specified where to find a manufactured panoramic adaptor with that input frequency capability. Most panoramic adaptors had adjustable inputs that ranged from 400kc up to 500kc since most receivers used the standard 455kc IF. More specifically, the US Navy receivers RBB, RBC used a 400kc IF (required panadapters designed for 400kc input,) the Signal Corps Hammarlund Super Pro used a 465kc IF, the US Navy-National HRO and RBJ used a 456kc IF and the USN-National RAO-6, 7 and 9 receivers used 455kc. Although the CSR-5 has the capability to allow panoramic adaptor use, it's unlikely it was ever historically set-up that way. Jerry Proc's website has a concept method that uses an AM-BC receiver modified to act as a frequency converter. It's an interesting idea if you really want to use a 455kc input panadapter with the CSR-5. Be aware that the panadapter connection on the CSR-5, and on almost all vacuum tube receivers, is a direct connection to the Mixer plate and has B+ present when the receiver is in operation. Most panadaptor input connections used a very high resistance value in series for isolation. Also, be aware that connecting a panadaptor to the Mixer stage will generally have a detrimental effect on the received signal. For that reason, some receivers had a vacuum tube buffer stage between the receiver's Mixer stage and the panadapter input, the USN RBB and RBC, for example, used a panadapter buffer that was an official USN upgrade kit found installed in many RBB and RBC receivers.

Front Panel Color Variations - The CSR-5 and 5A front panels are found in a variety of different paint colors with the most common original color being a wrinkle finish, very light greenish-creamy-grayish color with black nomenclature. Smooth finish semi-gloss light gray is found when the receiver was rack mounted and part of the CM-11 transmitter-receiver rack system. Beige or a light tan wrinkle finish was a common paint color in the stand-alone receivers. Many different shades of gray are often encountered and occasionally black wrinkle finish might be seen. Black Crackle, aka "alligator" finish, along with black wrinkle finish with white stenciled nomenclature was used on the Canadian DOT receivers. These CDOT receivers also had a chrome "Marconi" badge mounted just below the dial escutcheon. There are some examples that were apparently painted dark blue. Along with the different colors, shades and texture variations will be found both black or white nomenclature depending on the panel color used. Most of the unusual color variations might be from post-WWII RNC repaint jobs but, could also be from recent amateur restoration attempts where authenticity wasn't a primary consideration. Restorations nowadays that involve a complete repainting have to deal with the silk-screened nomenclature that will usually significantly complicate any refinishing plans. It's possible that stencils could be used since there are a few apparently original (well,...vintage) examples using this approach. Getting stencils made is certainly possible although somewhat expensive and then doing the actual application and having it look professional would require some experience working with stencils. The same caveat can be applied to rub-on dry-transfers. Unless one has considerable experience in applying this type of lettering, the results will be disappointingly obvious.


CSR-5 Crystals

Crystals - Shown in the photo right are two CSR-5 Crystals with their original boxes and desiccant packet. One crystal frequency is 5540kc and the other crystal frequency is 8710kc. The design of these crystal resonators utilizes an air gap meaning that the crystal holder also has ground glass spacers to maintain the proper air gap and the quartz crystal can move slightly within the holder. National Co. crystals in the 1930s HRO crystal filters also used an air gap type quartz crystal of similar design.

If "CRYSTAL OSC - IN" is selected, then the CRYSTAL UNIT installed in the front panel socket is connected into the Crystal Oscillator circuit that uses part of the 6K8 tube for the oscillator function. Although the crystal installed determines the frequency of the oscillator, the receiver has to be tuned to the desired frequency of operation so that the RF sections and the Mixer section are also tuned to work with the crystal oscillator frequency that will actually be 575kc higher than the operating frequency. When calculating the desired crystal oscillator frequency, the IF has to be added to the desired frequency. With the crystals shown, for example, 8710kc would provide 8135kc reception and 5540kc would provide 4970kc reception. If the crystal is to be used on a harmonic for operation on Band A one has to remember that the oscillator-Mixer circuit on that tuning range has the oscillator as the higher frequency instead of the lower frequency as on the frequency tuning ranges, B through F.

Loudspeakers - Two types of loudspeakers were available for the CSR-5. A table cabinet speaker and a dual rack mount speaker. The table speaker was an eight inch diameter speaker with a 10,000Z ohm to 8Z ohm matching transformer. The dual rack mount speaker is virtually unknown although it is pictured in the manual. It's not specified in the manual if the two speakers had individual inputs (most probable) or if the speakers were connected in parallel with a single input (unlikely, but in the photo on Jerry Proc's website, the CDOT set up shown has one dual speaker panel per receiver implying that perhaps the speakers were connected in parallel.) The photo to the right shows two CSR-5A receivers installed onboard the RCN ship Athabaskan in the early 1950s. Note that it appears the receiver on the left has the "block" lettering on the escutcheon implying it was repainted at a RCN depot. Note the Marconi loudspeakers mounted on the bulkhead above the receivers. These are the standard Marconi table speakers that were available for the CSR-5/5A receivers.

Other RCN Mods - The RCN usually added a SO-239 coaxial cable receptacle to the receiver to the chassis just to the right of the antenna terminal board. Original CSR-5 cabinets had to be modified to have a clearance hole added for access to the SO-239.

Also, note in this installation on the Athabaskan how the shock mount feet are mounted directly to the underneath of the cabinet on the receivers. This certainly simplifies the shock mount installation and significantly reduced the width of the receiver when compared to the width if the standard shock mount brackets were installed (added about 3" to the cabinet's total width.) The receiver in the center of the equipment is a version of the RAK made by RCA-Montreal, the TE236, that was used to monitor 500kc (although it tunes 15kc to 600kc.)

There was an RCN mod that changed the alignment for the lowest tuned frequency on Band F to around 69kc to allow reception of 73.6kc, a RNC Marine Command Broadcast station frequency.


CSR-5A receivers aboard the RCN ship Athabaskan in 1952. 
photo from:   www.jproc.ca


Marconi cursive script on escutcheon

 The Dial Escutcheon - Originally, the CSR-5 had Marconi CANADA CSR-5 silk-screened lettering on the dial escutcheon but, when receivers were refinished by RCN depots in the 1950s and 60s, many escutcheons were repainted and only block-letters were used. Many different fonts and letter sizes were used but, if the silk-screening was done at a depot, whether the receiver was a CSR-5 or a CSR-5A, the escutcheon was (always?) screened as "CSR-5A." The block-lettering shown to the right isn't typical with most examples found with much smaller lettering size. Some escutcheons were repainted by the RCN and no lettering was applied.
 

Both photos are from Jerry Proc's website, www.jproc.ca


Block Lettering from probable RCN repaint
CSR-5 or CSR-5A - Easy Identification - The easiest way to identify whether a receiver is a CSR-5 or a CSR-5A when viewed from the front of the receiver is to observe the position of the POWER switch and the SELECTIVITY control. On a CSR-5, the spacing from the top edge of the panel down to the switch or control is 1.5" but, the CSR-5A receiver spaces these two controls from the top edge of the front panel 2.0" down. Also, the CSR-5 receiver CRYSTAL socket is a two-pin type for large crystals with two holes in the panel for the crystal socket receptacles while the CSR-5A has an oblong panel cutout for the three-pin crystal socket receptacles (for large crystals and FT-243 crystals.) Once you know these differences, the receivers are easy to identify, even from photographs.

Manuals and Discrepancies - The manuals that I've seen have all been for the CSR-5A. All of the schematics I've seen are also exclusively for the CSR-5A. I haven't been able to find a manual or schematic that's specifically for the CSR-5 only. Perhaps there never was any specific CSR-5 documentation because the time period between the initial CSR-5 and the later CSR-5A is only a matter of months. Most of the CSR-5 receivers ended up being partially converted by the RCN to the CSR-5A using 6SG7 tubes for the 2nd RF amplifier and for the 1st IF amplifier. But, there are other differences that will be apparent if the particular receiver being worked on is an "upgraded" CSR-5 and CSR-5A documentation is used. Normally, during the conversion the BFO was not even looked at. The CSR-5A receivers supposedly have an adjustable iron-core inductor in addition to the air-variable capacitor for frequency adjustment. The CSR-5 uses a bank-wound air inductor with no adjustment of the BFO frequency other than the air-variable C. Another conflict is the last IF transformer, T3, is referenced more than once in the CSR-5A manual as having two trimmers, however, the all of the receivers will have T3 with just one trimmer and the CSR-5A schematic shows T3 with just one trimmer (probably an error in alignment instructions.) It's possible that R27, the cathode resistor for V4 (1st IF amplifier) was a 100 ohm resistor in the CSR-5 using a 6SK7 tube. Since most early receivers were converted to 6SG7, R27 will usually be a much higher value, perhaps up to 500 ohms. The more one "digs into" an RCN upgraded CSR-5, the more slight differences will be found.
 

The VP-3 Power Supply


VP-3 Power Supply for the CSR-5A Receiver but it can also be used with the CSR-5.      2010 photo

 The CSR-5 requires a separate power supply. There was a rack mount AC operated power supply, the WE-11 that was specifically for the CSR-5. The other power supply was the VP-3 that was specifically for the CSR-5A but, since the voltage outputs are the same, it can also be used for the CSR-5. The VP-3 provided 12vac 3A Filament for the series/parallel connections for all 6.3vac tubes, except the BFO and LO that use series resistive loads, that were used in the CSR-5 and +250vdc 110mA B+ with an input requirement of either 12vdc for the vibrator PS or 115/230vac. Two 6X5GT rectifiers are used in the VP-3 bringing the receiver's total tube count to 13 tubes. The two terminals with hex nuts installed are for the 12vdc input if the vibrator supply is desired. Internally, to change from AC operation to DC operation, first the AC plug on the AC cable has to be inserted into the dual slotted opening on the chassis next to the rectifier tube. This actuates a switch under the chassis, S-2. Then the six-pin "option plug - PL-1" is removed from the "AC" socket and installed into the "DC" socket on the chassis. Then the toggle switch on top of the chassis is flipped to "DC." This toggle switch usually has a guard that has to be removed to actuate the toggle switch. Next, connect 12vdc to the + and - terminals and make sure a 12 volt four-pin vibrator is in the socket.


Top of the VP-3 Chassis - the two rectifier tubes are Canadian Marconi 6X5G tubes. Chassis ID is for 6X5GT tubes.
When operating on DC, the vibrator creates an interrupted DC that is connected the power transformer primary winding that is specifically for the +12vdc vibrator output. This works with the power transformer secondary winding to provide HV to the 6X5GT rectifiers for B+ and another secondary winding provides the 12vac for tube heaters. Only a pi-network is used for B+ filtering. Originally, a spare vibrator could be stowed in clips located on the underside of the top cover. My VP-3 is set up for AC operation. A separate power transformer primary winding is provided for either 110vac or 220vac input. In AC operation, the VP-3 power supply is a very basic power supply with a pi-network filter. The outputs to the receiver are two cables with Cinch-Jones plug connectors. I've wrapped the two cables so as to handle like a single large cable. At the receiver-end, the cable has a non-original single C-J female plug that mates with the non-original 10-pin C-J male plug on the rear apron of the receiver chassis. This is a post-RNC, amateur modification from the original dual C-J connectors to a single larger 10 pin C-J connector. Originally, the CSR-5 used two C-J receptacles similar to the connectors the VP-3 power supply had except the receiver C-J connectors are male. The two pin C-J connector is routed through the cable and to the receiver and then to the SUPPLY switch on the receiver's front panel allowing the operator to switch "on" the VP-3 from the receiver. The three pin C-J connector has 12vac filaments, chassis ground and +250vdc B+. Unfortunately, the clips that help to retain the two-pin C-J plug were removed, probably to allow using the mil-grade AC plug (that does fit into the original C-J receptacle.) Changing this mil-grade AC plug back to the original type of C-J connector would be an easy task (and I do have the proper C-J plug and I did do the change.) Another non-original problem is the AC power cable that is a three-conductor grounded type with a three-pin AC plug. Of course, this AC plug can't be installed into the chassis AC receptacle to switch the VP-3 to DC operation. Replacing the AC power cable with a two-conductor cable and installing a proper AC plug will be necessary if DC operation is to be used. Of course, a dummy plug could be used to actuate the switch. Inserting the AC plug/cable into the chassis socket was a safety feature for DC operation preventing inadvertent or accidental "plugging in" the AC line during DC operation (if both power sources were available.)


Underside of the VP-3 Before - The top 2mfd tub capacitor is leaking oil. The two 22uf electrolytics aren't soldered on the negative side. Very sloppy workmanship.

 VP-3 Inspection and Repair - Apr 11, 2026 - I've owned this VP-3 since 2010 but I don't think I ever pulled the bottom cover off. If I had, I would have immediately seen the two 2uf oil-filled tub capacitors that were leaking oil. The oil was all over the bottom cover. The terminal seals were the red rubber type that are also found on the tub capacitors used in the WWII AR-88 that also are very prone to leaking oil (these seals are over 80 years old so oil leakage is a common problem.) Additionally, some of the capacitors had been replaced with orange drops and IR electrolytic capacitors. The two 22uf electrolytic capacitors for the B+ filter weren't soldered at the ground connection,...I guess there must have been enough contact to allow filtering since I never experienced hum in the CSR-5. The capacitor replacement workmanship was the worst I've seen in a while and looked like everything was done in a "rush" with components installed with hook-splices and glop-soldering. All of the replacement parts needed to be removed, the mounting terminals cleaned of the excess solder, then remounting the component using insulating sleeving on the leads and a proper soldering technique employed for the reinstallation.

 
VP3 Underside after removing the oil-leaking tub capacitor and straightening-up the component mounting and soldering. It still looks sloppy but that's original wiring for the most part.
I removed the top 2uf tub cap because that was the one that was leaking oil. I left the other tub cap installed since it wasn't leaking oil (yet) but disconnected from the circuit. I electronically replaced tubs with a 4uf 400vdc non-polarized capacitor just in case I want to try DC operation sometime. I dismounted all of the newer components and added sleeving to the leads. Then the components were remounted neatly and soldered correctly with real SnPb solder. A 16-2 AC power cable with mil-grade 2-pin plug was installed. Although the VP-3 seemed to be working fine before the clean-up operation, it sure didn't look like it would have worked very well at all. Now, with the underside looking better it's surprising that there doesn't seem to be any change in how the VP-3 operates.

NOTE: This VP-3 had Canadian Marconi 6X5G rectifier tubes installed when I got it. These tubes tested good and I've been using them for quite a while since they are marked "Canadian Marconi." However, it's fairly well-known that USA-manufactured 6X5G tubes that were built just before WWII and were used in the 1940 Zenith radios of that time commonly developed cathode to heater shorts that very often caused the high voltage winding of the power transformer to go "open circuit" mainly because Zenith radios of that time period were not fused. Supposedly, whatever the fault was in these initial USA 6X5G tubes, it was corrected with the 6X5GT version. Most of this is "Zenith-collector Urban Legend" type of information. But, I've seen a lot of 1940 Zenith radios with replaced power transformers, so maybe there is something to the information. Since the VP-3 chassis tube ID indicates "6X5GT" I've replaced those Canadian Marconi 6X5G tubes with USA-type 6X5GT tubes.   

 

CSR-5 SN:394 Re-Restoration

Earlier Restoration Dissatisfaction - I obtained CSR-5 SN:394 from a fellow collector that brought it to the Western Historic Radio Museum in Virginia City, Nevada in 2010. The receiver had been cosmetically restored earlier by another collector-restorer. This earlier owner had the cabinet and panel powder-coated in a grayish-brown color that was substantially darker than the light greenish-gray or beige wrinkle that has been found on most of stand-alone CSR-5 receivers. The collector that was selling the CSR-5 to me had obtained the receiver in "unfinished" condition and needing some electronic work that he had performed. The VP-3 power supply was included in the purchase.

Electronic Work - The "re-cap" job that had been performed by the former owner wasn't particularly sympathetic and had used "yellow jackets" (poly-film caps) and IC electrolytics. When installing replacement capacitors, many were mounted over the resistors on the component boards rather than going through the extra work to mount all of the capacitors on the backside of the component board as original. Most of the capacitors were connected using "hook joints" that were soldered rather than dismounting the old component completely and installing the replacement component directly to the terminals. Some components were left "hanging in the air" rather than using the correct lead dress as original. Almost all of the original resistors were also replaced with no regard for type or vintage. There were a few changes in component locations for some reason. Although the receiver did function (and that was probably all that was desired,) underneath looked amateurish and sloppy.

Cosmetic Work - Besides the mayhem under the chassis, I had always been disappointed at the color choice for the powder-coat paint job. Although it certainly could be argued that the CSR-5 receivers are found in quite a variety of color schemes, many of which are vintage, meaning that some color variations do have RCN provenance. However, the dark brownish-gray color doesn't compliment the great CSR-5 aesthetics, though I suppose that's just my subjective opinion. Then there's the silk-screening that at first glance appears to be not centered at the control locations. For quite a while I thought the silk-screening had been carelessly done and that accounted for the "off-center" look. However, when I acquired an original CSR-5 SN:665, I was able to compare the original silk-screening and I saw that the "off-center" nature of the nomenclature is original and, therefore, the silk-screener of SN:394 had been accurate in the locations of the nomenclature. But, the size of all of the nomenclature is too small and not the original style font. Also, the nomenclature color had to be white because of the dark panel. The smaller size and the white color of the nomenclature actually enhances the "off-center" appearance making it look more apparent.

Once I had obtained the receiver, I repainted the dial escutcheon black wrinkle finish because it had been flat black with white dry-transfer rub-on "block" letters that weren't an original RNC application and were obviously an amateur's attempt. I also made new cables for the power supply to receiver hook-up since the cables that had been used were made from very small gauge wires that didn't look sufficient to provide the power supply connections without a substantial voltage drop on the 12vac line. I didn't do anything else to the receiver except use it for a couple of test-listening sessions and then put it on display in the Western Historic Radio Museum in Virginia City, Nevada.

Unfortunately, although a repaint would certainly be possible, the silk-screened nomenclature is virtually impossible to duplicate and have the results be close to the original and appearing professional. However, when looking at many of the CSR-5 and CSR-5A receiver photos on the Internet, there are examples with much worse paint-jobs and much worse nomenclature reproduction attempts (like the CSR-5A with the DYMO label nomenclature.) This cosmetic restoration is in excellent condition and that's a major benefit.

Correcting Some of the Electronic Mayhem - Fast-forward to 2020,...I had closed WHRM in 2012 and moved to Dayton Valley, Nevada. The CSR-5 had been stored for the past seven and a half years in my "cool room" in the shop in Dayton Valley, a well-insulated and very dry storage room where the temp extremes are from 30F to maybe 75F and it's always dark,...unless I turn the lights on. I had just recently gone through a Marconi/RAF R1155 and its impressive multi-colored, semi-circular dial got me interested in that large multi-colored "half-circle" dial on the Canadian Marconi receiver. I had to extract it from the "cool room" and bring it into the house and upstairs to the radio repair lab to take a better look at it and reacquaint myself with how it performed.

As far as performance, the CSR-5 still worked as it had before,...okay, but seeming like it could do much better. The tubes tested okay but the band switch was very erratic and needed cleaning with DeOxit and a small paint brush. All controls were given the DeOxit treatment. The split-gears in the tuning gear box were stuck together so a flush with WD-40 loosened them and that eliminated the backlash. The BFO and NL toggle switches seemed non-functional but a "spray down the barrel" with DeOxit cleared up the problem-causing internal oxidation. I was positive that neither of the former owners I knew about had performed an alignment. If a full IF/RF alignment hasn't been performed, doing one will certainly improve performance significantly. The IF adjustments and RF L adjustments all have lock-nuts so be sure to loosen them when adjusting. When snugging-up the lock nut be sure to kept an eye on the receiver output to verify that the adjustment doesn't change with the lock-nut tightening. With this CSR-5, all of the alignment adjustments were far enough out that quite a bit of improvement was gained by the procedure, especially the tracking which now easily meets the 0.5% accuracy tolerance specification (marker stations are right-on frequency but the dial resolution is very limited.)

I checked the "problem C123" location since this receiver was a CSR-5 with a serial number lower than 816. C123 was entirely missing. It's not a critical component but it should be present to take care of high frequency transients and protect the output transformer primary. NOTE: On early CSR-5 receivers, C123 was connected to the wrong terminal on T4 and required correction. It's likely that by now all CSR-5 C123 connections have been corrected. C123 should be a 2200pf capacitor (square mica-type) connected to pins 4 and 5 on the audio output transformer T4. Also noted were two of the connections to T4 output xmfr weren't soldered and several others were "tack soldered" connections. All connections to T4 had to be completely redone. I checked over all of the "new" solder joints under the chassis and corrected those that looked questionable (a lot of 'em.) The RF Gain control connection to chassis needed to be redone. I shunted the series "re-radiation resistor" in the antenna connection (it had been changed from original 1K to a 10 ohm carbon composition resistor, but I shunted it anyway. Interestingly, the original rack-mount CSR-5 has the 1K re-radiation resistor installed and it doesn't seem to have any detrimental effects on reception.)
 


CSR-5 SN:394 in 2010 - data plate on front panel

 Cosmetically, one of the former owners had mounted the CSR-5 data plate on the front panel (Why? So it could be "seen" while the receiver was a shelf-queen. See photo to the left. Note that this photo was taken with tungsten lamps illuminating the room and that really brought-out the "brown" in the powder-coat paint.) The data plate should have been mounted to the top lid of the cabinet (as original.) The data plate had been installed on the front panel using drive pins. That required a C-clamp and spacer to push the pins out of the panel. I remounted the data plate to the cabinet lid (luckily, the original holes with screws and nuts hardware were already present.) The non-original holes in the front panel were filled with epoxy and then touched-up with custom matched paint.


Serial number "394" stamped into chassis's rear left-side
It's worth noting that on the data plate, while it's for a CSR-5 Type 110930-Z, it has the SN:682 stamped and yet stamped on each side and on the rear chassis apron of the receiver is "394" which I assume is the actual receiver's serial number. Cabinet swapping during RCN depot repair might account for the discrepancy. The RNC depot scene was probably comparable to the U.S. Army Signal Corps depot repairs of BC-348 receivers where it was very common to find the receiver installed into a non-matching cabinet, e.g., a BC-348-Q installed in a BC-348-R cabinet. It's also possible that maybe there never was any attempt at CMC to assure that the receiver chassis serial number matched the data plate serial number on the cabinet. One note is that the CSR-5 receiver chassis will not fit into a CSR-5A cabinet due to the rear chassis locating pins inside the "A" cabinet.

All of the front panel screws were new, non-original type, black Phillips head screws. This was another example of using non-authentic hardware just because it's easy to find and looks new. The original screws used on this receiver were round head slotted machine screws although some CSR-5 receivers will have binder head slotted screws installed. As long as all screw types match, it's an indication of original hardware or a careful restoration. I replaced all of the screws with the vintage correct types that were "raven finish". The external tooth locking washers were used to match the other locking washers used in the receiver.

I wanted to set up the CSR-5 with the shock mounts. I found a set of Barry shock mount feet on eBay and purchased the 1"x 1" mild steel angle locally. The angle brackets were built and then painted wrinkle finish using VHT Gray Wrinkle Finish. The VHT Gray has a lot of brown in the mix and it was a close match,...well, maybe one shade more towards gray but close enough. I've set CSR-5 SN:394 up as the station receiver with an ART-13A transmitter. Maybe not an original combination but both pieces are WWII veterans and work together nicely.


Canadian Marconi Company Type CSR-5 110930-Z Table Cabinet Version SN:394/682  from 1944
This photo shows the receiver after the subsequent "touch-ups" that I had to perform on the older cosmetic restoration. The Barry shock mounts and the side brackets can be seen, also the cabinet side venting.

Performance - Nov 2020 - Overall, CSR-5 SN:394 is a good performer that is fully capable of operating as a station receiver in an amateur vintage military radio set up (providing the particular Mil-Rad Net is operating AM on 75M.) The receiver is robustly-built and that provides excellent stability which is helped by the VR tube and filament loads on the LO and BFO tubes resulting in very little drift after a short warm-up. Sensitivity is competitive, especially if the receiver has been "gone through" and has had a complete IF/RF alignment and is used with a full-size, resonant (or tuned) antenna. Selectivity is very good if the crystal filter is operating correctly. It's a minor negative point that there's no front panel Phasing control for the crystal filter. The IF alignment instructions have C-64, the phasing trimmer, set to minimum bandwidth and that's generally the desired function for a Crystal Filter.

Marker frequency stations, such as WWV, are "right on" the correct index indication on the dial. On the down side, the dial resolution is vague at best, especially on SW. This is due to the wide frequency coverage of each tuning range and is typical and expected of any WWII vintage receiver. The logging dial and logging scale on the dial will provide an accurate method of frequency reset ability.

I'm using an eight inch Jensen speaker mounted in a Northern Electric Co. loudspeaker housing that looks a lot like an old RME speaker housing with a Northern Electric data plate (a property tag, actually) installed. But, it matches the CSR-5 quite well. To match the 10K output impedance, I'm using an 8K ohm to 8 ohm matching transformer spare that originally was for the Signal Corps LS-3 loudspeaker.

T-R function is accomplished using the Send-Rec "AC receptacle" on the back of the chassis apron or one can use the SEND-REC toggle switch on the front panel and then utilizing the remote standby contacts on the T-R relay inside the ART-13 transmitter. The ART-13 also has a vacuum receive antenna switch for receiver isolation during transmit.

The CSR-5 is a "heavy-weight" and when installed in the cabinet with the shock mounts weight is up around 75 pounds. The receiver chassis weighs about 50 pounds. The all-steel cabinet is very heavy, weighing in at about 25 pounds. Then add the VP-3 power supply at about 20 pounds but my cables are long enough so the VP-3 sets on the floor. It all adds up to a formidable weight experience unless the pieces are disconnected and moved separately including removing the CSR-5 chassis from the cabinet.

630M Operation? - I've been using the CSR-5 on 75M for several months now (this was in Nov 2020.) It's set up with the ART-13A from SAAMA. It's the same ART-13A and CU-32 combo that I used on 630M a couple of years ago. I had a 630M hiatus due to an ART-13A power supply modification that involved using a separate linear 25A power supply for the 28vdc supply. It turned out that in CW, this 25A power supply would "kick out" its breaker immediately with the first "dit." Voice operation was better but still random "kick outs" did happen. In November 2020, I purchased a used Lambda SWS600-24 switching supply that was rated at 25A and could be adjusted to over +30vdc. This power supply has solved all of the issues I had with using the linear supply. So, now 630M CW operation is possible again. A couple of years ago, I was using the Hammarlund SP-600VLF as the receiver and the receiving antenna was a homebrew 6' remotely-tuned loop. However, the SP-600VLF has been moved downstairs and I now have the CSR-5 as the station receiver. I needed to do a test of the CSR-5's MW reception performance. Also, the homebrew loop is history and I've been using a Pixel Shielded Magnetic Loop for MW reception. So, the test was to see what I'd receive using the CSR-5 and the Pixel Loop.

Actually, performance was a real surprise. I tuned through the 300kc to 420kc part of the MW band to see what NDBs were present. Quite a few, with LLD 352kc in Hawaii and YXL 390kc in Ontario, Canada being noticed as pretty good DX. The "blowtorch" NDBs like MOG 404kc or QQ 400kc were extremely strong. Several of the 25W NDBs were received out to the midwest. So, this indicated that the CSR-5 was a capable receiver when using the Pixel Loop. How about what was on 630M? From 472kc up to 479kc was occupied by about five JT9 data signals, one of which was very strong. I did hear some very slow CW around 473kc but the signal wasn't on but a few seconds. What the test indicated was that 630M CW using the CSR-5 with the ART-13A and operating with a scheduled QSO is very possible. Due to the JT9 data users occupying most of the 630M band, calling CQ on CW is usually a futile effort. Setting up a "sked" is the best method to assure a 2X QSO on CW. 

NOTE for 2026 LW Signals: This 630M testing was performed in the 2020 Long Wave Season. Listening on 630M has changed significantly since 2020 and very seldom are any CW stations heard, regardless of the type of equipment used. Virtually all stations on 630M are using digital modes that require special equipment tied in with computer monitoring. If any "real-time" CW stations are active they must rely on a scheduled contact. Calling CQ is a futile effort that results in nothing. Listening to NDBs has also changed significantly. Most of the NDBs I listed as received in November 2020 have been decommissioned and are no longer transmitting. In 2026, expect about a 75% reduction in the number of NDB stations as compared to the number of NDBs operating in 2010 and about a 50% reduction compared to the number of NDBs operating in 2020. NDBs are NOT used for navigation in the USA and are only active as an airport ID or are kept active as a tradition of air navigation. A couple of signals that can be used for LF testing would be LORAN E on 100kc (Loran-E Master Station "M" is located in Fallon, Nevada and runs 400KW to a 625ft ground-isolated vertical and is located only 50 miles from my QTH,...what an incredible signal!) Also, ALS162 on 162kc is a phase-encoded time signal running 800KW from France that can usually be received easily at night and sometimes during the daytime. Some cities have Public Service BC on 530kc that usually consists of winter road conditions or summer tourist information. These are usually 10W transmitters with a limited antenna so their BC range is extremely limited. I can pick-up the PS-BC 530kc from Carson City, Nevada about 20 miles away. And, of course, there are still a few NDBs operating in the 200kc to 425kc part of the spectrum.

NOTE For 2026 Discoveries: This CSR-5 has some slight instability that shows up as varying pitch on CW signals and also when tuning with the BFO on. It's probably also happening in the AM mode but isn't really noticeable without a heterodyne to actually hear the pitch changes. I've never cleaned the rotor contacts on the tuning condensers and I doubt that were cleaned by the former owners. That should be performed. The receiver wasn't used for about five years so the band switch also seems to need another DeOxit treatment. That also needs to be performed. I probably should go through the tubes and test them carefully. It does seem to be a common task to have to "go through" a receiver or any electronic equipment after long period of idleness.

I cleaned the tuning condensers rotor contacts and also cleaned the band switch contacts. This cleared up the pitch changing that was happening in CW. Weird as it seems, after owning this CSR-5 for 15 years, I just now noticed that the tube socket IDs for the 2nd RF amplifier and 1st IF amplifier have indeed been changed from the original 6SK7 to the 6SG7 (handwritten in ink.) I don't know where I got the idea that I had installed 6SK7 tubes in place of the 6SG7 tubes but after discovering the tube socket IDs, I checked the tubes,...and indeed they were 6SG7 tubes. Earlier, I had changed R27 to 270 ohms thinking I had a 6SK7 IF amplifier tube and, of course, the receiver's IF stage practically oscillated with so much gain. I changed R27 to 390 ohms which was an improvement but the IF stage was still prone to nearly go into oscillation. I'm going to change R27 to the 470 ohms (maybe even 500 ohms) that was in the circuit when I got the receiver. More on this as I discover things in this receiver that I obviously over-looked in 2020.  Apr 25, 2026
 

Alignments - Use Modern Test Gear - Like all WWII vintage radio equipment alignment procedures, the CSR-5 alignment procedure was written for the test gear that available at that time. Most of the alignment procedure is burdened with nowadays useless procedures for accurately determining the output frequency setting of an analog RF signal generator or other now-useless diversions from doing the actual adjustments necessary to the receiver. Using vintage test equipment will easily double the time needed for an alignment and will not benefit the accuracy at all. A modern synthesizer signal generator is a tremendous help because the frequency set up is direct. Just punch in the frequency and the amplitude and it's ready to go. No need for a separate digital frequency counter to verify the frequency of an analog dial type signal generator. Almost all synthesizer generators will also do sweep frequencies by just punching in the parameters of the sweep. Normally, a SYNC output is available. Most modern oscilloscopes have an XvsY function that will easily allow a sweep alignment of the IF. The old method shown in the manual was adjusting for coincident between two traces while XvsY allows viewing the IF passband directly. When doing the RF tracking, a synthesizer really speeds up the process. Just punch in the frequency and amplitude needed and do the adjustments necessary to the receiver. Some information in the manual will be necessary when doing the alignment using modern test gear. The manual's procedure will provide the specific trimmer or inductance adjustments and hopefully the location of the adjustment. The receiver's chassis has everything marked but sometimes the silk-screening has worn off or was painted over by the RCN. The manual also provides set-up frequencies for alignment. When it comes to monitoring the receiver output, the manual uses the vintage method of measuring the audio output level. This requires the input RF signal to be modulated (usual mod-f is 400hz) so that the audio output meter has an AC signal to rectify to have the meter indicate "peak-to-peak" voltage, or a DC voltage derived from the "pk-pk" audio level. Most audio output meters required a fairly strong signal level with the AF Gain level usually indicated be fully CW or full-on. Of course, a loudspeaker couldn't be connected up so the audio output meter usually had a load that substituted for the loudspeaker. It's really a lot easier to monitor the AVC bias voltage. This then allows the RF signal generator to be a pure sine wave without modulation. The AF gain can be turned all the way down. Much easier and a lot quieter. The conclusion is,...modern test gear is much easier to use, it's a lot more accurate, it's much smaller and lighter weight. There's really no advantage to using vintage test equipment,...except for the VTVM. A really good VTVM will provide an accurate way to "adjust to peak" and provide an easy to see, large display with a moving needle that's easy to see when making adjustments that require setting a "peak" value.

CSR-5 Alignment Details Including Sweep IF Alignment

Equipment used: HP3325A Synthesizer Function Generator up to 20mc, FNIRSI DPOS350P 2 Ch Oscilloscope, 50mc Synthesizer (used above 20mc when needed) and DFC, ME-26D Military VTVM, Leader LOB 505 Oscilloscope with XvsY capability (if needed,) a General Radio 1192-B Digital Frequency Counter (if needed,) a HP-3312A Function Generator with Sweep (if needed) and an HP606B analog RF signal generator up to 50mc (if needed.)

IF Alignment - The manual's first paragraph in the alignment procedure is basically a set-up to find the Crystal Filter's crystal resonant frequency. Even using an analog function generator and a digital frequency counter is easier and more accurate than what is described here. As far as connecting the HP3325A through a .01uf capacitor to the Mixer grid and setting the 3325A output frequency to 575kc and then stepping up the amplitude for a moderate level signal,...that's fine. The procedure has you do a "peak" IF alignment at this point just to be sure the IF is close to 575kc. The next steps essentially describe setting the SELECTIVITY to 4 (narrowest bandwidth) and varying the generator frequency to find the actual active frequency of the crystal in the Crystal Filter. The HP3325A can be "stepped" by .001kc (or less) just using the "arrow" buttons to "zero-in" on the actual active frequency of the crystal. The output of the receiver can be listened to for the "swishing" sound and the ME-26D can be monitoring the AVC line for a maximum indication. BFO should be off. The HP3325A will indicate on its display the crystal frequency and that should be written down for later reference. Once the active frequency of the crystal is known, now the IF can be aligned again using this exact frequency. Since I'm using a synthesizer generator, I can easily reset to the exact frequency anytime. Another IF "peak alignment" is performed. Then a sweep alignment is performed. Since the object of the sweep alignment is a "flat top" symmetrical passband, the sweep alignment is performed in SELECTIVITY position 1. The oscilloscope X channel is connected to the Diode Load/AVC line ahead of the 1 meg resistor. The Y channel is connected to the synthesizer sync signal. The sweep rate should be about 25hz and the frequency start should be around 300kc and the stop should be around 700kc. Since the IF is already peak-aligned, all that's necessary is very slight adjustments to shape the passband. The standard method for sweep alignments would be to inject the sweep signal at the grid of the last IF amplifier and adjust the output IF transformer for a symmetrical passband. Then move to the grid of the proceeding IF amplifier performing the same steps. Finally, the sweep signal is connected to the Mixer grid and the 1st IF output transformer is adjusted. Switch between 1 and 2 on the SELECTIVITY and make very small adjustments to have the passband shape symmetrical (as much as possible) in both positions. The Crystal Filter isn't adjusted during a sweep normally. The thing to remember is doing sweep IF alignments is that you probably won't be able to adjust the IF passband for the classic "flat-top steep-sides" but get as close as possible. Also, adjustments to the IF transformer trimmers is very slight, don't make any major changes to the adjustments. The goal is a symmetrical passband with an amplitude that's very close to that of a "peak alignment." The Crystal Filter is adjusted as described in the procedure. The Crystal Filter is only used in SELECTIVITY positions 3 and 4.

RF Tracking - The synthesizer generator should go through a dummy antenna to the receiver's antenna input. There are three types of dummy antennas described in the manual, two using RC or RLC components and one that just uses a series R or C depending on the Band being adjusted. It's not critical because the last step of the alignment is to adjust the Antenna trimmers for the antenna used on each specific Band. The hassle of the RF tracking alignment is that the receiver has to be on its side to be able to access the 2nd RF and Mixer C-trimmers. Use the VTVM to monitor the AVC line voltage. The manual's procedure is okay to follow for specific frequency inputs, dial settings and which trimmers of L-slugs to adjust. RF Tracking is easy but tedious because there are so many adjustments. Also, when "tracking" Band A, be sure to verify where the image is because at the frequencies involved, it very easy to align to the image instead of the correct frequency. Also, the HFO is below the tuned frequency on Band A, so the image will be 1250kc below the tuned frequency. If aligned to an image, none of the tracking on Band A will be correct across the entire range, so it's usually obvious.
 
 


Type CSR-5 110480-Z  SN:665 - R.C.N. PATT: 3AU/17

CANADIAN MARCONI COMPANY
CSR-5  Receiver (from CM-11)
Type CSR-5 110480-Z -  R.N.C. PATT.3AU/17
Chassis SN: G0 H50  Actual SN: 665
 


Found this CSR-5 listed on eBay in February 2026. I watched it for a month or so as the seller kept reducing the price. When he reached $255, I went ahead and hit the BIN. The seller was in Escondido, California. He had listed the receiver as a CSR-5A receiver but it was obviously a CSR-5. I think he based the "A" assumption on the RCN tag on the side panel that indicates the following "R.C.N. PATT. 3AU/17" indicating the receiver was installed in the CM-11 Transmitter-Receiver. However, both CSR-5 and CSR-5A receivers were used in the CM-11 and both were identified as pattern 3AU/17, hence the misidentification. Note that the panel is NOT wrinkle finish but is a smooth, semi-gloss light gray color. The light gray color in semi-gloss smooth finish paint was standard for the CM-11 equipment.

Data Plates, Serial Numbers and R.C.N. Tags - SN:665 has a small metal tag mounted on the right side panel of the receiver with "R.N.C. PATT.3AU/17" on the tag. The CM-11 transmitter had a similar tag installed on its side panel with "R.C.N. PATT.3AU/18" on its tag. The presence of the 3AU/17 tag on this CSR-5 is confirmation that the receiver was part of a CM-11 Transmitter-Receiver combination. This receiver was originally rack-mounted but someone removed the top cover and also removed the two D-zus mounts on each side panel at the back. This top cover removal was very common with rack-mounted equipment and a result of the misplaced belief that the top cover retained heat and caused failures. It's very seldom that rack-mount receivers, especially when going through maintenance facilities several times over many years, were able to retain their original top covers. Unfortunately, the data plate was mounted on the rear of the top cover and the data plate had the actual serial number stamped on it. I thought the actual serial number was lost with the top cover until I discovered it well-hidden. 


Not the actual Serial Number?

Actual Serial Number? - For about one month, I thought G0 H50 was the serial number of this CSR-5 receiver. I mean, it's stamped on the rear apron of the chassis,...a logical assumption, right? Well, G0 H50 didn't fit any formats of serial numbers used on CSR-5 receivers, so it was a bit of a mystery. Finally, I had the receiver upside-down on the bench and happened to notice a number stamped on the inside of the right side-panel,...665. An actual serial number? It fits the format but not the location. But, what else could it be but the actual serial number? Photo to the right.

See photo below-right showing the complete CM-11 Transmitter-Receiver,...well, complete except for the ZM-11 power supply that mounted underneath the operating table. The bottom unit is a CSR-5A receiver, the middle unit is the CM-11 transmitter (a potent 100W CW transmitter using an 813 PA, less output power in MCW or Voice) and the top unit is the Antenna Tuner.

Initial Inspection: Problems Found

1. Yellow power cable exiting the rectangular hole that originally had the two pin Jones connector. This was a fairly modern three conductor cord that had the black and white wires connected to the Power switch wires using crimp splices. Removed.
2. Cinch-Jones male two-pin is missing (two blade pins and one round locating pin.) Connection for Power switch from VP-3 external power supply. The round Cinch-Jones version is very common but the rectangular version with the round locating pin is not as common but still fairly easy to find.
3. 3rd IF transformer shield-can has a gouged open area at the top corner of the shield. This can easily be repaired but it does require disassembly of the 3rd IF transformer to remove the shield-can for the repair. This looks like a very involved process to remove the shield cover from the IF transformer (but it wasn't.)
4. The two large knobs are installed in wrong positions. Knob with index "dot" should be in Range switch position. Switched positions only to discover that sometime in the past someone filed a "flat" on the shaft so the knob wouldn't slip. Too bad they filed the flat in the wrong position so that with the correct knob installed and aligned with the index-dot to the correct range letter the set screws are nowhere near the flat. So, they drilled a new index dot that is in the correct position but didn't fill it with white paint and then didn't paint the original dot index black. It doesn't matter too much as long as the range switching mechanism is lubricated along with the switch detent, the range changing should be easily accomplished and not loosen the knob. With the lubrication, the range switch operation is easy enough that the knob doesn't move when changing tuning ranges.
5. White backing plastic index that mounts behind the logging dial has a large chip out of the upper left side. Not visible with the front panel mounted.


CM-11 Transmitter-Receiver
From Bottom to Top, Rcvr, Xmtr, Ant Tuner
photo from: www.jproc.ca
6. BFO toggle switch is a replacement that is "raven finish" and doesn't match the other toggle switches. I'll change to the correct type of switch. (switched the switch. Installed a matching A-H nickel toggle switch.)
7. Right side grab handle slightly bent inwards. Will have to remove to straighten.
8. All of the 12 volt 1 C.P. had blackened glass bulbs. They all had good filaments but I replaced them with NOS #53 bulbs since some of the black deposits were quite dark.
9. Range E color on dial is light blue but color on indicator is light green. Range E on SN:394 is a bluish-gray color on dial and gray on the indicator. Faded colors?
10. Four of the five small knobs are missing their white index "dot." Looking at vintage photos it's difficult to tell on CSR-5 receivers if the "dots" were filled or not.
11. Tested all tubes on the TV-7 tube tester. All tubes tested >50% above minimum acceptable. Several of the tubes are Marconi brand or they have "Made for Canadian Marconi" stamped in red on the tube (metal octal type.)
12. The 2nd RF and 1st IF tubes changed from 6SK7 to 6SG7 with nomenclature changed on the chassis to match the CSR-5A circuit and schematic using hand-written ink. The 6SG7 has a gm of around 4300 where the 6SK7 had a gm of about 2300. The tubes' pin-outs are nearly identical except that the Suppressor Grid-G3 is tied to the Cathode-K internally with the 6SG7 but they are not tied together internally in the 6SK7. The G3 and K are tied together in the receiver wiring however, so the only significant difference is the transconductance of the tubes. The 6SG7 is a semi-remote cutoff pentode and the 6SK7 is a remote cutoff pentode and this difference can affect how the AVC operates. The cathode resistor of V4, the 1st IF amplifier, is shown as "factory select" with the range being 100 to 400 ohms (in CSR-5A documents.) This CSR-5 has an original-looking 100 ohms installed and that value provides maximum gain with the 6SG7 tubes (although, maybe 100 ohms was correct for the CSR-5 and a 6SK7 tube but then why wasn't R27 changed when the upgrade was done?) 

NOTE: There is considerably more information on component differences between the CSR-5 and the CSR-5A in the last section of this article (section "Back to the Bench - Apr 21, 2026.") How those differences relate to performance of the receiver and how the changes can affect the use of the receiver nowadays, particularly as an amateur vintage military radio station receiver are covered in that section.


As Delivered Condition - Not bad

 Front Panel and Dial Cleaning - Mar 29, 2026 - Removed all of the knobs. Dismounted the all of the toggle switches, phone jacks, crystal holder and the RF Gain pot. Four 8-32 BH slotted screws mount the front panel which was then removed. Cleaning is much easier with the front panel separated from the receiver. The front panel was very dirty, although it really didn't look like it was. I used Glass Plus, WD-40 and finally Isopropyl Alcohol to remove all of the different kinds of crud that was on the front panel. None of these cleaners had any effect on the nomenclature silk-screening.

Cleaned the dial with Glass Plus. The dial was also very dirty and required several cleanings to remove all of the dirt. Some dirt deposits were also in the colored scales and that required Glass Plus and a soft small paint brush to clean out of the engraving. Cleaned the logging dial carefully but the silk-screen ink wasn't affected by the Glass Plus. Cleaned with white plastic index plate. Cleaned the tuning range indicator.

Cleaned the front part of the chassis and lubricated the range switch dent and the gear bearings for the range indicator. Lubed the control shafts. Remounted the front panel to see how things looked now,...big improvement. Front panel color went from a greenish-grayish color to very light gray.


Top of the chassis before any work or cleaning was performed. Note the one dial lamp mount that's bent and the lamp socket is dangling next to it. Also note the yellow power cable that was "hard-wired" using crimp-splices to act as the Power switch wiring. Also, the gouged hole in the 3rd IF can top. The RF deck has a silicone coating that appears to have been applied over all of the alignment adjustments and also over the existing grunge. Alcohol dissolved this coating but it also dissolved the silk-screen nomenclature. As can be seen in this photo, the RF deck silk-screening wasn't in very good condition to begin with.

Cleaning Disaster,...well,...a mini-disaster - Mar 30, 2026 - I cleaned the knobs and reinstalled them. I started cleaning the chassis, beginning with the RF deck. This appeared to have a coating that was applied over dirt, maybe. Nothing seemed to remove the dirt, except Isopropyl Alcohol applied with a Q-tip. Then I applied more alcohol using a small paint brush and in a few seconds the coating wrinkled and lifted. It appeared that the coating was silicone-based due to the way it "balled up" in clumps as it was removed. It didn't appear to be from WWII but maybe some later application that was over the entire RF deck, including ALL of the alignment adjustments. Originally, a thin coating (probably lacquer) was only on the silk-screened nomenclature. Since the alcohol didn't seem to affect the silk-screen nomenclature, I started to remove the silicone coating. Well,...I thought the silk-screening wasn't affected. Actually, if the alcohol was left on the silk-screen lettering long enough, it did dissolve the lettering too. It takes longer,...but it will happen. So,...don't try to clean the chassis with anything other than WD-40 or Glass Plus. This now absent nomenclature on the RF deck doesn't really affect anything electronically,...just visually. The alignment instructions show where the adjustments are located and I have the other CSR-5 receiver as a reference.


Underneath is mostly original except for minor RCN maintenance that included modification from CSR-5 to CSR-5A

Still,...it's too bad I didn't test the Isopropyl alcohol more thoroughly before applying massive quantities on the RF deck but this type of "rubbing alcohol" is so mild you can rub it on your skin for a "rub down" so I didn't think it would have the effect that it did. And, as mentioned, this was some later silicone spray applied over everything on the RF deck, dirt, alignment adjustments and all,...not an original coating.

Power ON Test - Mar 31, 2026 - I used a 12vdc 2.5A power supply and clip-lead connected the positive to the Cinch-Jones bottom pin of the power connector. Negative was connected to the center pin. I used another clip-lead to short the POWER wires (simulates the POWER switch being ON.) I switched on the +12vdc and the dial lamps slowly illuminated and I could see the tube heater in the 9002 tube. I put the Lambda 25 on the bench and connected +200vdc to the top pin of the Cinch-Jones connector and switched on the Lambda. The CSR-5 had a loudspeaker connected and a ten foot long test wire for the antenna. Background noise and no hum since I was using the Lambda. The controls were scratching and dirty but I was able to tune in 15mc WWV coming in strong. Tuned a few 20M CW hams. I tuned down to 40M but since it was about 0830hrs, nobody was on,...but then, I was only using a 10 foot wire for the antenna. Not bad for an initial power-on test. Later I connected the Collinear Array for the antenna. Copied hams on 40M and on 20M. I had to reset the BFO knob. I also had to spray some DeOxit down the barrel of the AVC switch to get it working.

NOTE: I based my decision to not recap this CSR-5 on the experiences of Gerry O'Hara VE7GUH and his restoration of a CSR-5 that he wrote up in a SPARC article "Restoring a Marconi CSR-5 - A Canadian WWII Classic." In his write-up Gerry related that he tested the original capacitors that he pulled while doing the recap of the receiver. He found that all but one capacitor tested good. I think it's comparable to the USN RAO-7 receivers that seem to operate quite well on all original parts. During the later part of WWII, the best quality components were usually provided to manufacturers,...not always (like Micamold) but most of the time. Of course, this applies only to receivers that have been well-taken-care-of and were stored indoors their entire existence. It certainly doesn't apply to receivers that have been stored for decades in an unheated shed with a leaky roof and a voracious rodent population. At any rate, I'll see if any problems develop from this decision, but I doubt they will. SN:G0H50 appears to have always been kept indoors and is in excellent physical condition. I almost never perform a "wholesale replacement" of all capacitors in WWII vintage gear,...well, unless they're Micamold capacitors. The CSR-5 uses most Aerovox capacitors and that's a pretty good brand. Sprague is the best as far as functional longevity but the key is the environment that these 80 year old components have had to survive. Here is a link to Gerry's SPARC article:  https://sparcradio.ca/wp-content/uploads/2020/03/Canadian-Marconi-CSR-5-Receiver.pdf

Detailing - I pulled the cover on the tuning condenser and used a small paint brush to apply DeOxit to the rotor contacts. The shield over the three gang condenser is easy to remove but the shield over the 1st RF/Ant stage has two screws on top of the chassis and then two nuts under the chassis that need to be removed before the shield can be dismounted. I cleaned the rotor contacts on the RF condenser also. Afterwards both shields were remounted. I used a small paint brush to apply DeOxit to all of the band switch segments. Replaced several missing 6-32 BH machine screws for the bottom covers. Tested on 20M, copied several SE USA hams on USB.

The Dilemma of the Cinch-Jones Connectors - Luckily, the three-pin male Cinch-Jones (C-J) connector for the B+, Ground-Chassis and the 12vac tube heaters is present. However, the two-pin male C-J connector is missing. This connector is for the two wires that run up to the POWER switch and allow turning on the VP-3 power supply from the receiver. A large yellow power cable had been connected directly to the original wires using crimp splices probably because the original C-J connector had broken. I removed the yellow power cable and have been powering the CSR-5 using a +12vdc 2.5A power supply and the Lambda 25 for the +240vdc B+ that are connected to the 3 pin C-J connector pins using clip-on test leads. Once I have a proper C-J male connector installed then I can use the Canadian Marconi VP-3 power supply to run the CSR-5. So far, it's been difficult to find the proper connector in the size that fits the opening, or L=0.875" x H=0.50" and the connector body has to have a mounting flange that allows mounting in a horizontal opening for the connector with the flange mounting holes being vertical oriented. Once the proper C-J connector is found, the rest is easy. Although, my other CSR-5 receiver uses a non-original (what else?) ten-pin C-J connector with the five wires required connecting to the single connector. The receiver-end of the cable from the VP-3 also has a mating ten-pin C-J connector but the power supply end of the cable does have the original three-pin C-J connector for the VP-3 but the two-wire connector is some sort of mil-grade AC plug that fits into the original C-J receptacle. I could make a second power cable with the correct C-J connectors for SN:665 and then I'd be able to operate either receiver using the VP-3. Maybe later,...I could modify the CSR-5 SN:394 to have the correct C-J connectors and modify the receiver-end of the cable to use the correct C-J connectors.  Two sets of two-pin male and female C-J connectors ordered and one three pin female C-J connector ordered - Apr 2, 2026

Tube Heaters now on Hammond Transformer - Apr 1, 2026 - I came across this transformer while out in the shop looking for,...what else?,...Cinch-Jones connectors. It was a Hammond transformer in the original box and was a type 266 with dual primary windings and dual 6.3vac secondary windings. The transformer was rated at 12.6vac at 4 amps. The primary windings had to be connected in parallel for 117vac operation and the two secondary windings had to be connected in series for 12.6vac. I used an AC "zip cord" with molded plug for the primary connection to the house AC line (with "bucking transformer" to lower the AC line from 122vac to 116vac.) I used test clip-leads for the secondary connections to the receiver. For the 3 amp current load, I used two clip leads in parallel for the heater connection and an 18ga. 12" test lead for the ground connection. CSR-5 performance seems a little better, more stable, SSB signals easier to tune in on 20M. Also, solid copy on Trenton Military VOLMET on 15.035mc USB (I know Trenton Mil's operating frequency,...the CSR-5 dial only indicated some frequency slightly above 15mc. Typical of any WWII vintage receiver, if an accurate determination of tuned frequency was necessary then the operator would use a Heterodyne Frequency Meter to make that determination.)  NOTE: This combination of the Lambda 25 and the 12vac 4A transformer has provided very clean reception on the higher frequencies. On 20M CW, the heterodyne tones are clean and distortion-free. SSB signals are easy to tune in and sound very good.

Verified Reception on all Tuning Ranges - Apr 2, 2026 - I used the HP606B as a signal source but I connected its output to a 10ft wire to act as the radiator rather than connecting the HP606 directly to the CSR-5 antenna input. However, the CSR-5 was connected to the Collinear Array outdoor antenna but I didn't match the antenna to the tuning ranges, except that it was matched to 20M. Signals were received on all tuning ranges. On E and F ranges, the LF and MW part of the spectrum covered, signals were noticeably at a lower level, although still easily heard. These two ranges have reduced sensitivity specs, as indicated in the manual, so the somewhat lower response was expected. I was surprised at the tracking since it appeared to be about as good as my other CSR-5 that I aligned just a few years ago.

BFO Switch - Apr 3, 2026 - The BFO toggle switch was a "raven" finish,...a black finish mainly for black wrinkle finish Navy equipment. It was an Arrow-Hart switch, which is correct, but the wrong color. The soldering quality (or lack of it) was an indication that the switch was a replacement but something that was probably accomplished in the 1950s. I looked through the toggle switch bin and found a nickel-plated A-H toggle switch that was SPST, as needed. I decided to pull the front panel since it makes the switch replacement so much easier. I had to recondition the wire ends as would be expected since this is the third switch installed in this position, the original, then the black switch and now this nickel switch. Wiring was simple since there are only two wires going to the switch. While I had the panel off, I adjusted the position of the logging dial because it was just slightly rubbing on the backside against the white plastic index piece. I only moved the logging dial out about 0.060" and that was sufficient to eliminate the rubbing. I reinstalled the front panel and knobs.

CSR-5 BFO is Different than CSR-5A BFO  - Since the manual that is available is for the CSR-5A, it describes a BFO that has an iron-core trimmer adjustable-type inductor. It also shows that type of inductor on the CSR-5A schematic. The manual's alignment procedure indicates a BFO frequency adjustment the uses the air variable C for the front panel BFO control but utilizing this adjustable iron-core inductor to "trim-in" the zero-beat frequency. However, the CSR-5 uses a different BFO inductor without the iron-core. The inductor is a "honeycomb wound" air core inductor. There isn't a trimmer adjustment for the BFO, at least, not as described in the manual. The older BFO is adjusted by finding zero beat with 575kc un-modulated signal injected into the Mixer grid using the front panel BFO control (adjusting the air-variable C that should be "half-mesh" at 575kc.) Once BFO zero is found, the BFO knob is loosened on the shaft and set to "0" and the knob set screws snugged-up. Check that the 575kc oscillation from the BFO changes by increasing frequency if the knob is rotated CW and decreasing frequency when the knob is rotated CCW. In both directions the BFO should change frequency by about 4kc. Photo right shows what the old CSR-5 BFO looks like with the shield cover removed.


Older BFO as found in CSR-5 Receivers

3rd IF Transformer Shield-can Gouge Hole - Apr 4, 2026 - Repairing the hole in the 3rd IF transformer shield-can required dismounting it from the chassis. First, the bottom cover was removed that covered the IF section of the receiver to verify that the two mounting nuts for the shield-can were easily accessible. I didn't want to unsolder the IF transformer, so I removed the three, side-mounting screws that hold the IF transformer assembly inside the shield-can. There isn't enough clearance to remove the shield-can unless the iron-core slug adjustments are screwed almost down to the bottom. I measured the length of screw-threads showing before I adjusted the slug-cores so I would know approximately where to reset them when finished. With the mounting nuts removed, the three side screws removed and the slug-core adjustments screwed in, the shield-can was easily dismounted. I used wooden dowels to push the soft aluminum into place to close the hole and remove the dents in the can. I applied epoxy to the crack that had been the hole. After the epoxy cured, I painted the area of the repair with silver paint to match the color of the aluminum. After the paint dried, the IF transformer was then reassembled and mounted. Then the two slug-core adjustments were adjusted out to 0.312" threads showing. The receiver was powered up and the 3rd IF transformer "peak" adjusted. The three photos below show the process.


3rd IF transformer shield-can with gouged hole


Shield-Can Removed showing the IF transformer assembly


3rd IF transformer shield-can repaired

More Band Switch Cleaning - Although the CSR-5 is working fine on all tuning ranges, it's a bit erratic on Band A, the highest frequency tuning range. I was trying to receive some 15M ham signals and noticed that if I "rocked" the band switch knob, I could hear some ham stations. No amount of "rocking" though seemed to be a permanent solution. Since Band A is at the end of the switch rotation (mechanical stop) I can't really get a good "swiping" action to help clean the switch contacts. A little more De-Oxit judiciously applied should help.

Cinch-Jones Plugs - Apr 6, 2026 - The C-J three-pin female cable connector arrived today. It was a perfect fit. I wired 12.6vac, Chassis-Ground/B- and +250vdc B+ to three 16 gauge wires about 36" long. At the moment, the wire ends are soldered to the 12.6vac 4A transformer with one transformer secondary lead connected to Chassis-Ground/B-. The other transformer secondary lead is connected to the wire that goes to 12.6vac input. The B+ wire is connected directly to the Lambda 25 B+ binding post. A second wire from the Lambda 25's B- binding post is connected to the transformer's Chassis-Ground/B- wire for the B- return. Tested and the receiver functions correctly. The two-pin C-J male plugs arrived the next day (Apr 7) and these won't fit the rectangular opening on the back of the CSR-5 without a little bit of modification. Of course, I wouldn't modify the CSR-5 opening,...I'd modify the connector to fit. This is actually easy since the embossing of the connector body actually defines the proper size to fit the CSR-5 rectangular opening. I used a hacksaw to remove about 0.060" on all four sides to achieve the proper size. Then the pins are too long and will contact the chassis. I had to bend the pins to have some clearance. Finally, I had to come up with a bracket for mounting the now correct size C-J connector. A strap type of bracket was made from thin aluminum. It was made to wrap around the back of the connector (between the two pins) and was secured to the connector body with epoxy. By having the bracket wrap around the connector its strength is greatly increased since the mating plug has to be pushed in and pulled out with each move of the receiver to a different location. Also, the modification of this C-J connector has to take into consideration that there are two additional components and a tie strip mounted inside the rear connector housing,...clearance is minimal inside. The modified C-J connector fit inside the connector cover with no problems. It was mounted with vintage screws to match those used for the other C-J connector. Since someone had added the yellow power cable with crimp splices, now the original wires were too short to reach inside the connector cover. I had to splice about 3" of wire to reach inside the connector cover and be able to solder the wires to the C-J terminals. Each splice was covered with black sleeving and the wires were pushed against the chassis side wall. I had to chase the chassis threads for the mounting of the connector cover and then use vintage and matching screws for mounting. I tested the fit of the female C-J connector and it was perfect. Finally, I was able to test the operation of the C-J pins to the SUPPLY ON switch. NO Continuity. I sprayed DeOxit down the barrel of the toggle switch and operated it back and forth several times. I then had continuity at the SUPPLY ON switch and also had continuity at the C-J pins. Apr 8, 2026 

VP-3 Power Supply Sharing - I only have one VP-3 Power Supply. By alternating compatible cables, either CSR-5 receiver could be operated using the single VP-3. Another possibility would be to build an AC power supply to operate the rack mount CSR-5. I already have a good 12vac 4A transformer. I have all of the parts necessary for building a +250vdc 110mA supply. I also have a nice-size, unused chassis and a NOS Bud cabinet for the project. The only advantage of a second power supply would be if I wanted to set-up both CSR-5 receivers and operate either one without having to exchange receiver power cables. The disadvantage is, of course, having to build a power supply,...the sheet metal work is a PAIN. The cables just "plug-in" so it's not that involved of an operation to swap cables and "share" the single VP-3. It's just that the VP-3 usually sets on the floor under the operating desk so crawling under the desk to swap cables might be considered another type of PAIN.    

The cable for SN:665 is 4' long. It has two 16 gauge wires for 12vac and chassis-ground. One 18 gauge wire is provided for +250vdc B+. The POWER ON wires are two 16 gauge wires that are actually a heavy-duty AC zip cord. As with the first cable, this second cable has all five wires taped together to be one cable but with two different connectors on each end. All connectors on this cable are the correct C-J types, as original.

The CSR-5 SN:394 already has its power cable. I'm not really satisfied with the installation of a 10 pin C-J connector on the receiver. Eventually, I might replace the 10 pin C-J connector although the housing has been severely modified. A butch-plate will be required to use the original housing. When looking at the housing, I think it's actually a power transformer top cover. If I can find a power transformer cover the correct size, that might be an easy solution. Also, on the SN:394 cable, the connector for the POWER ON is a mil-grade AC metal plug. These AC plugs do fit into the C-J connector that is the original receptacle on the VP-3. I have changed the mil-grade AC plug to the correct type of C-J plug (two blade pins and one round locating pin) on this cable. I might consider changing the 10 pin C-J connector to the original type connectors IF the decision is made to redo the entire recap job.

NOTE: After actually performing this "crawling under the bench to switch cables" method of powering two receivers from one VP-3 power supply, I've had to re-think the project of creating a second power supply. As mentioned, I have all of the parts,...even the C-J connectors. Although a solid-state approach would simplify construction, I'd opt for the dual 6X5GT rectifiers as the VP-3 uses. Also, AC-only operation would certainly make sense and a lot less work. An improvement would be to use dual-section filtering. That would involve using two filter chokes and three electrolytic filter capacitors. Dual-section filtering is very effective at reducing hum to an absolute minimum. Since I have the proper Greenlee punch for the 6X5GT octal tube sockets, the sheet metal work involved is relatively easy. Although I don't have the proper rectangular punch for the C-J connectors. A good tool for making square or rectangular openings is a sheet metal "nibbling tool." These hand tools allow removing small slices of metal in a very controlled manner so creating C-J connector openings is fairly easy. Also, since the chassis I have is aluminum, rather than steel, it's much easier to perform all of the cutting, drilling and punching needed. Unfortunately, the "never-used" BUD cabinet is steel but there won't be an excessive amount of metalwork required except for the panel and that's flat sheet metal so the work involved there is fairly easy. I'll update here if this project actually progresses beyond the planning stage.
Quick and Easy Top Cover - Finding an original top cover for the rack-mount CSR-5 would next to impossible. This is just a quick method to provide some type of top cover for the CSR-5. This is definitely not a permanent cover,...just something that adds a look of "completeness" to the receiver until an original top is found (unlikely that will ever happen.) The R-390A top cover is just about the same size that's required to set on top of the CSR-5 side panels. The only problem is original R-390A covers are needed for R-390A restorations - not for something like this. However, about 15 years ago, Fair Radio was selling reproduction R-390A covers,...cheap. I bought a set back then. These were absolutely terrible reproductions that were made of very thin sheet metal (much thinner than the original covers) and the covers didn't fit without some modification and a lot of bending. I never even used the top cover,...until now. An R-390A top cover can just set on top of a rack-mount CSR-5 and it fits okay "as-is" without any modifications. Just set it on top and it looks pretty good. However, if you have one of these crappy repros, a better fit is easy to accomplish by clipping-out the front hole in each side lip to allow the cover to clear the CSR-5 top cover mounting posts. An even better fit can be accomplished by removing about a six inch section of the top cover's front lip to allow the cover to clear the dial lamp reflector and slide under the front panel lip. This reproduction top cover had a gold iridite dye (not easily removed) so I painted it with silver spray paint for a more appropriate appearance.


Rack-mount CSR-5 SN:665 with the non-original, "Quick and Easy" Top Cover Installed
IMPORTANT NOTE: I would never, NEVER, NEVER have done this modification to an original R-390A top cover,...no matter how bad of condition the original R-390A cover was in. I only did this mod because this particular R-390A cover I had was a cheaply-made reproduction that didn't even fit an R-390A anyway. But, as mentioned, even without any modifications, an extra or spare original R-390A top cover can just set on top of a CSR-5 and it fits pretty well and it will improve the looks of a rack-mount CSR-5 that's missing its original top cover. It's just a temporary fix until an original CSR-5 top cover shows up,...as unlikely as that's going to be. Also, the slight modifications to the repro cover mentioned aren't visible if or when the cover would be "bent-to-fit" and installed on an R-390A,...well,...other than the silver paint job. And,...even if the repro cover was used on an R-390A, it would only be considered a "temporary" installation anyway,...to be replaced as soon as an original R-390A cover could be found.
 
SN:665 Performance using the VP-3 - Apr 14, 2026 - Up to now I've been running this CSR-5 with the 12vac 4A transformer and the Lambda 25 bench power supply. Now that I've gone through the VP-3 and have built a compatible power cable with the correct C-J connectors, I'm ready to see how the CSR-5 runs with an original-type Canadian Marconi power supply.

When first powered up, the CSR-5 didn't have any BFO. I tuned the BFO control 180º and heard a very weak heterodyne. I switched power supplies going back to the Lambda and the BFO still wasn't operating correctly. I replaced V9 (BFO tube) with a NOS 6SK7 CRC/USN tube and the BFO then operated normally. I retested the old 6SK7 and was surprised that it tested just at "minimum acceptable" when two weeks earlier it had tested good (it's also unexpected that a "minimum acceptable" tube in a BFO circuit would not function,...but tube testers don't catch all types of failures since the "tube under test" isn't operated in the tube tester anything like it is in the receiver.) I switched back to the VP-3 and everything was then operating correctly. I was test-listening on 20M. With the Lambda, heterodynes were very clean with no modulation. The VP-3 also allowed for clean heterodynes on 20M. SSB signals were easily tuned on 20M. If there are any differences in operating the CSR-5 on the Lambda compared to the VP-3, they seem to be insignificant and not really noticeable. I measured the B+ voltage at the output of the VP-3 while it was operating the CSR-5 and the DVM indicated +254.2vdc. I'm running the VP-3 AC input using a bucking transformer so the AC input is 116vac.

Later in the afternoon I monitored a SSB station on 20M that was working a variety of other ham stations around the country. I had only let the CSR-5 warm up about five minutes. I didn't have to retune this SSB signal at all and I listened to him work about seven or eight stations for a duration of about 15 minutes. I was surprised at the stability on 20M and especially involving a SSB signal. The amateur station was located in Corpus Christi, TX and was a vehicular mobile that was stationary but parked near the water of the Gulf (he was surrounded by massive amounts of salt water that accounted for his strong signal for only 100 watts to a mobile vertical antenna.) He did try to work a VK (Australia) but he couldn't copy him well enough and I couldn't even hear the VK.


CSR-5 Receivers with ART-13A Transmitter
On the Air with the Rack-Mount CSR-5 and the ART-13A - Apr 19, 2026 - I used the SAAMA ART-13A with the table cabinet CSR-5 in 2020 with very good results. I used the combination for several months which indicates that I liked how the CSR-5 performed with the ART-13A. Although I've listened quite a bit with the rack-mount CSR-5, I haven't used it installed in an amateur vintage mil-rad station until now. With the cabinet version of the CSR-5, I had the remote standby operational and I now have the remote standby set up for the rack-mount. The most difficult part (it's not really that difficult) of the remote standby set up is accessing the NC contacts on the ART-13A sending relay. The U8/U connector on the ART-13A provides direct connections to the NO-ARM-NC contacts for receiver remote standby. Pins 23 and 24 are the ARM and NC contacts and are accessed with 14 gauge "pins" that slide into the U8/U connector receptacle sockets #23 and #24. These 14 gauge pins are soldered to a 4 foot long zip cord that has an AC plug on the other end. The AC plug is installed into the remote relay socket on the rear chassis apron of the receiver. The receiver has to have the SEND-RECEIVE switch in the SEND position for remote standby operation. When the ART-13A PTT is actuated, the CSR-5 instantly goes into standby. The CSR-5 receiver antenna coax is connected to the REC terminal on the left side of the ART-13A. This REC terminal uses the internal vacuum switch to isolate the receiver during transmit. The ART-13A set up was originally for the BC-348 receivers used in the WWII bombers equipped with the ARC-8 radio gear. Since the CSR-5 is essentially functioning in the same manner, the ART-13A can be interconnected with the CSR-5 and both work together nicely. The use of the ART-13A PTT to control the remote standby of the CSR-5 allows for quick break-in and that allows me the ability to use this mil-rad station when I'm operating as net control for the Nevada Vintage Mil-Rad Net. The ART-13A uses an AC operated power supply of my own design and it allows the transmitter to have an output power of 120 watts that is routed through a Viking KW Matchbox to the Collinear Array antenna.

When operating on 75M sensitivity isn't really very important since even the simplest of receivers are sensitive enough for 75M nets. There are two other reception factors that are important though,...how selective is the receiver and how well does it function in an RFI-noisy environment. The selectivity needs are dependent on the time of operation. The Nevada Vintage Mil-Rad Net operates early Sunday morning when activity is fairly low but propagation is adequate for about a 300 mile radius. Usually 6kc bandwidth is enough selectivity to cope with adjacent QRM but finding a WWII receiver with 6kc bandwidth is sometimes a challenge. The CSR-5 has a Crystal Filter that is brought into the IF circuit in SELECTIVITY positions 3 and 4 and these positions are usually adequate for the normal QRM. Also, "off frequency tuning" can be applied for "dodging" minor QRM. Just tune a couple of kc higher or lower to reduce the adjacent frequency QRM. RFI-Noise is a serious problem on 75M that sometimes prevents any reception at all. Metropolitan areas and even some urban areas are becoming an unrelenting noise source that nothing can be done about. Some receivers are better than others in coping with RFI-Noise. Also, antenna types can reduce some types of RF noise. Shielded Magnetic Loop antennas are particularly effective for coping with an RFI-noisy environment. In severe RFI-noise environments, remote reception using SDR-Internet accessible receiver stations may be necessary. Any WWII vintage receiver that has a Noise Limiter circuit, that NL circuit was designed to reduce auto ignition noise. A pulsing, repetitive noise "spike" that was easy to remove with a clipper-type NL circuit. Most NL circuits for that time period work fine on auto ignition noise,...but, of course, there hasn't been any auto ignition noise produced in decades. The roar of SCRs, modulated furnace blowers, switcher power supplies, lamp dimmer controls, neon pilot lamps, even some LED lamps are all RFI noises that the Clipper NL doesn't respond to. When I used the cabinet version CSR-5 in 2020, I didn't have any problems with either QRM or with RFI-Noise. 

April 19, 2026 -  I was net control for the Nevada Vintage Mil-Rad Net operating on 3.974mc AM mode starting at 0650hrs. This is what I noticed about the CSR-5 SN:665 operation during the one hour long net. It appears that the AVC isn't working very well or maybe not at all. It did seem to slightly control the gain when on the bench but that was listening to 20M stations. Now, on 75M, the AVC doesn't seem to work and the RF Gain has to be reduced on strong stations. Possibly some of the AVC resistors may have drifted in value allowing just enough AVC bias to be available for 20M signal strength but not enough negative bias to control 75M signal strength. I didn't check the resistors, so that will need to be done. I think it's obvious that the non-functional AVC is also responsible for the necessity of keeping the RF gain set to below 15 and sometimes below 10 for strong stations. I didn't notice any frequency drift but I'm always retuning with every transmitting station change so I wouldn't notice frequency drift anyway. QRM was never a severe problem but once or twice an SSB signal would be close enough to be heard in the passband. A slight retuning to one sideband or the other (QRM-dodging) easily eliminated the interfering signal. I have to say though, this QRM was just the occasional adjacent frequency artifacts that are heard when using WWII receivers and not heavy, deliberate interference. I never even had to switch the SELECTIVITY off of 1, the widest bandwidth. Audio quality was excellent. I would say that SN:665 performed very well on its initial "sea trial" and only has one obvious issue and that's the weak or inoperative AVC on 75M.

Differences in Performance Between the CSR-5 and the CSR-5A
Back on the Bench - Apr 21, 2026 - In checking over the CSR-5 SN:665, I performed a DCR test on all of the resistors and all were easily within 20% of the marked value except for R59. In looking at the available Internet information, I can't really find a specific "parts list" for the CSR-5, only for the CSR-5A. I checked all of the tag boards and the resistors all agree in value with the CSR-5A "parts list." If this receiver was modified by the RNC to be a CSR-5A, then the resistors are more-or-less correct. I can see that at least half of the resistors are not original,...but they aren't brand new either. If their values have been changed, it was done while the receiver was still in active use by the RNC. The change from 6SK7 tubes for the 2nd RF amplifier and the 1st IF amplifier to 6SG7 tubes that would operate at twice the gain would seem to require an adjustment in cathode resistance and possibly screen voltage. For V2, the 2nd RF amplifier, that would require some change in either the 300 ohm cathode resistor or the 1000 ohm screen dropping resistor, although these are CSR-5A values and the CSR-5 with a 6SK7 installed might have had different values. Definitely, R27, the cathode resistor for V4 in the 1st IF amplifier is specified as "100 to 400 ohms" in the CSR-5A parts list, implying that it was a selected value for CSR-5A receivers. However, SN:665 has what looks like the original R27 installed and it's a 100 ohm resistor. Without a CSR-5 parts list and, even if I found a CSR-5 schematic it wouldn't show component values, I can only guess what value the resistors were for V2 and V4.

Out of curiosity, I replaced the two 6SG7 tubes with NOS tested-good 6SK7. There wasn't too much difference in reception except the AVC now seemed to work much better (remember, R27 is still 100 ohms.) I used WWV 15mc for a high frequency strong signal and 4mc with an HP606B strong signal source for a low frequency. I could definitely hear the difference when switching the AVC off and on. When adjusting the RF gain, with the AVC on, at above 15 there isn't anymore increase in sensitivity since the AVC has control at that level on strong signals. When tuning off of strong SW-BC stations, I can hear the AVC tc delay as the AVC bias is reduced (goes less negative) and the RF/IF gain increases, indicating that the AVC is functioning correctly with the 6SK7 tubes installed. I think the problem is using 6SG7 tubes and then R27 being 100 ohms, there's too much IF gain and that overloads the AVC.

Comparing SN:394 to SN:665 - Since I can't find a CSR-5 "parts list" the next best thing is inspect CSR-5 SN:394. After putting SN:394 on the bench and removing the bottom covers, it was unfortunately very obvious that nothing could be determined from SN:394 because virtually every resistor on the two IF tag boards, along with most of the resistors in the receiver, have been replaced. The R-values installed are what the CSR-5A parts list shows for the R-values and that fits with the 6SG7 tube IDs and tubes. A comparison between the two CSR-5(A) receivers follows:

In comparing SN:394 to SN:665,...SN:394 has R27 installed as a fairly new style Allen-Bradley 470 ohm 5% resistor. In SN:665, R27 is the original 100 ohms (maybe original, measures 118 ohms - Cathode R for V4, 1st IF amplifier.) I changed the 470 ohm resistor in SN:394 to 220 ohms. This resulted in the receiver almost oscillating as SW-BC stations were tuned. I then increased the R27 value to 390 ohms and that seemed to work better. In SN:665, I changed R27 to 270 ohms and the receiver performed just about the same as with the 100 ohms R27. I should change R27 to ~500 ohms and put the 6SG7 tubes back in,...and I did, Apr 27th. 

Other discrepancies are,...SN:394 has R34 (part of the Cathode divider for V5 2nd IF) amplifier that measures 840 ohms (one of the few original resistors but the color code damaged and can't be read with confidence but appears that it was 500 ohms - green?, black, brown.) I changed R34 to a 390 ohm resistor. No obvious change in performance. In SN:665, R34 is an original resistor that measures 398 ohms (should be 400 ohms, so it's okay.) 

In SN:665, R59 (the other Cathode divider resistor for V5) is an original 400 ohm resistor that measures 602 ohms, a 50% drift so it should be changed since it's somewhat affecting the cathode bias on V5. I changed R59 to a 390 ohm resistor. No noticeable change in performance. Cathode bias measured 2.1vdc and is shown as 2.5vdc in the manual voltage check list.

In trying various resistors and gain control adjustments on SN:394, it seems the resistor that determines best performance is the value of R27 acting something like an IF Gain control. If 6SG7 tubes are going to be employed, the value of R27 should be hand-selected for best gain without overloading the AVC and without the IF oscillating. The expected value would be between 100 ohms and 400 ohms and probably a lot closer to 400 ohms,...maybe even 500 ohms. I don't think the value of R27 was that critical when using 6SK7 tubes but it should probably still be a value of about 400 ohms. This higher value tends to keep the IF stage gain reasonable and stable. An alignment of the IF also might prove beneficial to stability and performance.

Were the CSR-5A Changes to Benefit CW Reception? - I think the 6SG7 tubes were used for maximum sensitivity when operating in the CW mode, that is, with the AVC off, the BFO on and riding the RF gain for best ratio of BFO injection to signal strength,...especially on Bands B and A. It's also possible that, with CW being the primary mode of reception, the Crystal Filter was used often and the additional RF/IF gain when using the Crystal Filter may have been a significant help with narrow bandwidth CW reception. Since this CW mode set-up has the RF gain controlled manually with no AVC, the problems with over-loading the AVC won't be encountered. For better AM-Voice operation with AVC controlling the receiver sensitivity, I think the 6SG7 can be used if R27 has a minimum value of 400 ohms,...500 ohms is better. Certainly during WWII and for quite a long time afterwards, CW was the primary mode of communications for the USN and probably for the RCN also. It seems logical that the CSR-5A changes were to benefit the receiver's performance in that primary mode of marine communication. The same sort of CW reception design decisions can be found in the WWII USN RAO-7 receivers where, certainly when used by the USN, the RAO was never used with AVC and always was considered a CW receiver to be operated without AVC. When using the RAO-7, it becomes obvious that the AVC has way too much negative bias and when the receiver is operated above 10mc in the AM mode, the AVC responds to the received noise rather than the tuned signal with the AVC cutting back the sensitivity. This only happens with the AVC on when searching for weak to moderate AM signals. In the CW mode, with the AVC off, reception is normal on all bands. The CSR-5A wouldn't be the only "Navy Receiver" that was enhanced for CW reception to the detriment of AVC-controlled gain for AM reception. It seems, that nowadays, the best performance and operation for the CSR-5, many of which have been "partially" converted to CSR-5A receivers, will depend on the individual's listening preferences. For CW ops, 6SG7 tubes with the value of R27 at the lower end of the range "100 to 400 ohms." For AM with AVC control, 6SG7 tubes with R27 value between 400 to 500 ohms.

Other Questions - Although R27 in SN:665 looked like the original resistor, there was a lot of "gloppy" soldering on the terminals and the cloth-insulated wire lead condition and resistor lead wraps didn't look original. That's unfortunate because, if it was definitely original, then one could assume the the CSR-5 used 100 ohms for R27 and then later the CSR-5A made R27 hand-select between 100 ohms and 400 ohms. I installed a 270 ohm resistor as a test and that value seems to work fine with 6SK7 tubes (I'm contemplating changing this 270 ohm value to around 400-500 ohms for better IF gain stability and going back to 6SG7 tubes. I did this change Apr 27,2026 and the performance is much better with 6SG7 tubes for V2 and V4 and with R27's value at 500 ohms - Band A perked-up quite a bit.) So, I guess the question would be, was R27 a 100 ohm resistor in the initial CSR-5 receiver? Was R27 changed to "hand-select" between 100 to 400 ohms for the CSR-5A with 6SG7 tubes? Was a subsequent alignment of the IF section part of the upgrade?

The VOLUME control installed in SN:665 appears to be original. It measures about 1.5 meg with the parts list value shown as 1 meg. With this particular potentiometer, with a full CCW setting and the arm reads open to the grounded terminal. A slight movement makes contact with the resistive element with this initial resistance being about 25K to ground. The pot appears to be a linear taper (okay for CW.) The listenable VOLUME level is between 0 and 1 with 2 being very loud. This is in the AM mode with the AVC on, RF gain at 20 and tuned to a strong SW-BC station. CSR-5 SN:394 has had its VOLUME pot changed. The VOLUME adjustment on SN:394 works great with normal volume being achieved around 4 in the AM mode. I don't like to deviate (if possible) from original components and changing the original VOLUME control in SN:665 isn't going to happen. I'm sure there's nothing wrong with the potentiometer and it's part of the original CSR-5 design. The question here would be, do all original CSR-5 volume control pots function in this manner of seemingly designed for CW operation where the VOLUME would be set above 5 and the sensitivity controlled manually with the RF gain?

The BFO described in the CSR-5A manual and shown on the CSR-5A schematic isn't the type used in either of my CSR-5 receivers. My assumption is that the BFO circuit was changed to have an iron-core inductance that was adjustable along with the BFO capacitor for tuning the BFO on the front panel. The CSR-5A BFO adjustment L3 location isn't shown in the manual. Was L3 accessible from the top of the chassis on a CSR-5A?

It's obvious that the RCN changed several resistors over the period of time that CSR-5 SN:665 was in active use. However, SN:665 is still basically an "original" receiver since it was never "hamstered." I did change the 6SG7 tubes back to 6SK7 tubes but after considering that the receiver is going to be kept as original as possible, I've returned the 6SG7 tubes (with an adjustment of R27 to 500 ohms for best stability.) I wonder if any other owners of CSR-5 receivers actually have receivers that were never "upgraded" and are still using original type 6SK7 tubes? In such a receiver, I wonder what the value of R27 would be?
 
 

For More CSR-5 and CSR-5A Info - go the Jerry Proc's fabulous webpage located on his extensive website on Canadian Radio History and Royal Canadian Navy ships. URL is www.jproc.ca for the extensive website and www.jproc.ca/marconi/csr5a.html for the webpage specifically for the CSR-5 and CSR-5A. Much of the information presented in this article about the CSR-5 was found on Jerry Proc's website.

Also, an excellent article on restoring the CSR-5 receiver was written by Gerry O'Hara VE7GUH "Restoring a Marconi CSR-5 - A Canadian WWII Classic." Gerry's write-up has a lot of detail about the formidable task of replacing the visible capacitors by "restuffing" the original cardboard tubes with new polyfilm caps and then using brown hot melt glue to seal the cardboard tubes (most capacitors are well-hidden so it was a limited quantity that were plainly visible and required preserving their original appearance.) Also, a note on a few of the inaccessible capacitors. Gerry also built-up vintage-looking electrolytic capacitors for a believable under-chassis appearance. Also, some info on recreating the appearance of WWII IRC resistors using more modern components. If you're contemplating rebuilding a CSR-5, reading Gerry's write-up will be extremely helpful.
Here's a link to Gerry's SPARC write-up:  https://sparcradio.ca/wp-content/uploads/2020/03/Canadian-Marconi-CSR-5-Receiver.pdf
 
 

Return to Home Index

 

 

Radio Boulevard
 Western Historic Radio Museum

 Vintage Radio Communication Equipment Rebuilding & Restoration Articles,

 Vintage Radio History and WHRM Radio Photo Galleries

1909 - 1969

- 60 years of Radio Technology -

 

 

This website created and maintained by: Henry Rogers - Radio Boulevard, Western Historic Radio Museum © 1997/2026