Wireless Set No. 52 Canadian - Project

[David Dunlop]

The 3rd set of resistance tests for the Sender were completed yesterday and overall, the results were very consistent with the two sets of tests run last Spring.

Of none electrical note, the rust on my VTVM Testing Procedures is now shedding in huge chunks. Most of the red flag readings I took earlier that were ‘Infinity’ in nature produced excessively high, but real values this time around. This was primarily due to me focusing on taking advantage of the different testing ranges on the meter to see if and where I might actually get a useful reading. It is amazing what one can forget when being an infrequent user of helpful equipment.

The majority of the good readings were either identical to the earlier ones, or +/- very, very close, so that was nice to see as well. The red flags are all clustered in the same circuits as earlier and the other big plus was nothing new showed up that might have been related to the overheating event that happened a few weeks back, so I have inflicted no new damage…so far.

The dark side of all this, however, is my next step in the process. I now have to go back to the Sender Circuit Diagram and methodically trace out each red flag circuit to identify all the related components. There are likely a few resistors that have dropped off in value, but most of the problems are probably going to be with the PIO Capacitors that have transformed themselves into resistors. The problem now though is the 52-Set Circuit Diagram for the Sender is very compact for easy reading/tracing. One sooner or later has to resort to using a magnifying glass, or lamp, which narrows down your field of view, compounding the problem. A whole different experience from working with 19-Set diagrams.

The challenges don’t stop there either. Once I have identified all the components that need further investigation, I have to find them on the chassis and test each. The ones tied to valve socket terminals are the easy ones. It is the remote ones tied to a terminal panel, tucked behind another component, six inches away that create the nightmares.

Good thing it is a long Winter in these parts.


David

[David Dunlop]

As expected, and not withstanding current seasonal activities, the analytical work involved in tracking down problems identified in testing electronics is moving painfully slow for me. The price I pay, I suppose for not doing enough of it over the last few decades. In any event, progress is actually being made.

The initial point of interest was that across all three resistance tests, the problem results were nearly all of excessively high values, and most were directly related to tests in which the Mode of Operation Switch on the Sender was set to the R.T. Mode. A number of these results, of course, showed up with Pin to Ground readings for the 813, which was a little scary initially. Then I finally realized the Speech Circuit would be the common factor across all these problem results.

The Speech Amplifier in the Sender is V1J, an ARP-3 Valve, and in close relation to it is V5D, the Modulator, the position of which is taken up by a 6V6G Valve.

In the tests result, V5D, showed no real problems and it shares some of the circuit components of V1J. V1J, on the other hand showed good results until tests involved it be active in the RT Mode and then results went off the charts. So V1J now has my undivided attention.

The two problem circuits for V1J include 3 capacitors, one of which is electrolytic, and 5 resistors. One of those resistors shows normal values off RT Mode and three of the others are shared when required by V5D, so may be OK. The stage of my work now, is finding where all eight of these components are located on the actual chassis and testing each individually. An interesting challenge in its own right.

In the two photos today, the first shows the V1J socket at the bottom, partially hidden by the T2A transformer (easily removable), with the V5D socket directly above it. I suspect all of the components needing examination will be either tied directly to the pins of these two sockets or hiding somewhere along the board in the second photo which is fastened to the lower rear edge of the Sender chassis with three screws.

Hard to do, I know, but it would have been nice if the capacitors could have been mounted on the boards with the useful specs data showing. I have found more than enough in 19-Sets and other equipment, including this 52-Set, where the only readable information is the capacitor makers name or “MADE IN CANADA” and the side with the critical specs is securely fixed against the circuit board.

The journey continues.


David

[Chris Suslowicz]

Originally Posted by David Dunlop

As expected, and not withstanding current seasonal activities, the analytical work involved in tracking down problems identified in testing electronics is moving painfully slow for me. The price I pay, I suppose for not doing enough of it over the last few decades. In any event, progress is actually being made.

The initial point of interest was that across all three resistance tests, the problem results were nearly all of excessively high values, and most were directly related to tests in which the Mode of Operation Switch on the Sender was set to the R.T. Mode. A number of these results, of course, showed up with Pin to Ground readings for the 813, which was a little scary initially. Then I finally realized the Speech Circuit would be the common factor across all these problem results.

(Snipped a bit.)

The two problem circuits for V1J include 3 capacitors, one of which is electrolytic, and 5 resistors. One of those resistors shows normal values off RT Mode and three of the others are shared when required by V5D, so may be OK. The stage of my work now, is finding where all eight of these components are located on the actual chassis and testing each individually. An interesting challenge in its own right.

High values in the resistance tests will not be caused by faulty capacitors - they should have infinite resistance (waxed paper/paper in oil/ceramic/silver mica types) or a low resistance that increases rapidly as they charge up (electrolytics). Capacitor failures are usually low resistance (electrolytics where the oxide dielectric layer has dissolved), moisture entry (waxed paper), or short circuits (silver mica), or else open circuit failures. High resistances are more likely due to resistor ageing or dirt/corrosion on switch and relay contacts - so it's worth using switch cleaner on those if the results of the testing are "off" in one position.

Capacitors should be checked for low resistances (may need to lift the wire at one end so you're only testing the capacitor and not any associated resistors), and reformed (electrolytic types) or replaced (all types) if leaky.

Hard to do, I know, but it would have been nice if the capacitors could have been mounted on the boards with the useful specs data showing. I have found more than enough in 19-Sets and other equipment, including this 52-Set, where the only readable information is the capacitor makers name or “MADE IN CANADA” and the side with the critical specs is securely fixed against the circuit board.

The "PC" number is visible on a lot of them though, so if you've got the actual parts list that will tell you the value and rating. PC = Procurement Catalog(ue), and it's common to all equipment so you can use the numbers from the WS19 if they're in that manual.

It might be worth making a list of PC numbers and the specification as you work through the restoration - looking up visible numbers is a lot easier than dismantling the set to get at the hidden side (though you may need to do that anyway if it's faulty and has to be replaced).

Best regards,
Chris.

[Mike K]

This guy has many excellent tutorials on repairing old tube equipment

https://www.youtube.com/watch?v=lLQThhf3Brc

[Chris Suslowicz]

Originally Posted by Chris Suslowicz

(Snipped a lot.)

The "PC" number is visible on a lot of them though, so if you've got the actual parts list that will tell you the value and rating. PC = Procurement Catalog(ue), and it's common to all equipment so you can use the numbers from the WS19 if they're in that manual.

It might be worth making a list of PC numbers and the specification as you work through the restoration - looking up visible numbers is a lot easier than dismantling the set to get at the hidden side (though you may need to do that anyway if it's faulty and has to be replaced).

Having looked at the 1945 Identification List (does not contain PC numbers) and the WS52 User Handbook. (Which only contains CMC part numbers!) I don't think my suggestion was very helpful - though the User Handbook does list what the various capacitors relate to, which may be some help.

Chris.

[David Dunlop]

Hi Chris.

Actually, you did jog me in the direction of checking the 19-set documentation for possible cross references, which I had not thought of to this point in time. So Thanks on that!

I then went back into the 52-Set Operators Manual to see how useful the data there might be and discovered of 51 different capacitors in the 52-Set, only 5 were actually identified by PC Number. As you noted, all the others had Canadian Marconi Part Numbers. Be interesting to see if the CMC Numbers were actually stamped onto the caps used in the assembly process. I shall find out shortly I suspect.

On the bright side, however, I did stop this time and read the prefix notes Marconi included at the front of the Parts Index at the back of the Operators Manual and noticed they had included a reference to the fact quite a few of the capacitors used in the 52-Set were also used in the WS No. 9 and WS No. 19 and those parts supplies can be drawn on to effect repairs to the 52-Set, if necessary.

I will keep you posted with what I find hiding on the chassis and this might even allow me to play a bit more with some photo software, of which I have explored very little over the years.

Cheers for now,

David

[David Dunlop]

Good Evening Mike, and thanks for the interesting link.

I actually purchased a new LCR Meter similar to the one shown in that video a few years ago, primarily for the ease of testing capacitors that are still in circuit. The conventional method wherein one needs to disconnect one end of a cap to test it can be a real PITA if the cap is difficult to access to start with and then you find it is working just fine and have to reinstall it. Been there, done that and hate the T-Shirt.

The two pieces of ‘older kit’ he had were interesting for their in situ testing capability of capacitors. Way back in 1946, the Solar Manufacturing Company produced their Model CF Capacitor Exam-eter. Each meter came with its own calibrated manual and a set of test leads, one of which was a matched lead similar in approach to the one in the video somebody had bodged. My leads were missing, but there was enough data in the manual, along with some very astute notes made by the Radio Engineer who originally owned my meter, that I was able to replicate a new pair of leads.

This meter can do a staggering number of tests on capacitors, as well as resistance testing, AC and DC Voltage checks and a few others. I have done some resistance testing with it out of curiosity and it gives results very close to what I get with my vintage RCA and Simpson VTVMs.

I have somewhat limited mains supply at my work bench and two things came to the forefront in purchasing the new LCR Meter. It is battery driven, so no plug in requirements and, it will eventually give me a means of comparison for capacitor test results from the vintage Solar meter.

If there is a dark side to all this, I must remember to always be careful with wayward voltages when using a modern digital test meter on vintage valve driven electronics.


David

[David Dunlop]

Lately, in the time spent stepping away from the resistance analysis of the Sender, in order to clear my head, I have been thinking about the reasoning behind the Canadian Marconi Company changing the name of the piece of canvas they designed to provide some degree of moisture protection to the front of the 52-Set.

When you first look at it, everything makes sense. It rolls up and down in front of the Set, just like a ‘curtain’ in a window. It is waterproof, but absolutely not weatherproof. If the operator was caught outdoors in the rain working a 52-Set, I would not want to be that operator. In the rain, nothing about a 52-Set would be ‘covered’ with this simple piece of canvas.

Why Marconi chose a light olive green canvas for their curtain, may simply have been a factor of product availability at the time they needed to get the curtain into production. Khaki Tan may just not have been available. Nothing about any of this appears to be sloppy work on the part of CMC. They clearly knew it would be unwise to operate a 52-Set exposed in wet weather. If it was not going to be installed in a fixed building, or an enclosed vehicle, CMC went to the trouble of ensuring a Wireless Tent Kit was available to operate the 52-Set. Marconi had covered all the bases.

Photo evidence suggests the first curtains that used nickel plated hardware was a bad idea that Marconi identified and fixed, switching to matt black hardware in production as soon as possible. To date, no 52-Set components have turned up bearing 1945 dated data plates, so if seems logical production of the 52-Set ended a some point late in 1944. This would have included full sets, kits and a appropriate amount of spare parts. However, the February 1945 Master Parts List identifies this canvas item as “COVERS, Waterproof No. C2” but advises items may be found identified as ‘curtains’ and this is to be ignored. Identical information is found in the July 1948 revision of the List. It seems likely, therefore, the identification of ‘COVERS, Waterproof No. C2’ is a pure piece of supply system paperwork, simply for standardization purposes. It is highly likely, items were never manufactured bearing this identification.

I took a look in my 19-Set Parts Lists and found the following five covers:

COVERS, Waterproof No. 1 ZA 2952 for Carriers No. 21
COVERS, Waterproof No. C1 ZA/CAN 1076 in Universal Carrier
COVERS, Waterproof No. 5 ZA 10317 for Carriers No. 23
COVERS, Waterproof No. 16 ZA 10411 for Variometer
COVERS, Waterproof ZA/CAN 1362 for 19-Set Cdn in Carrier Universal


So in spite of the fact the 52-Set ‘CURTAIN, Waterproof’ is indeed a ‘curtain’ and does not cover the set at all, as any of the above noted items would, bureaucrats decided the name must change, not Canadian Marconi.

The three attached photos are of the following COVERS in my collection: No. 1, No. 5 and No. 16.


David