WW2 Canadian Wireless Valve Testers

[David Dunlop]

These three photos show the final stages of the socket wiring for the adapter. With this stage of the project finally completed, there was no further need for the socket reference tags on the underside of the cover, so off they came.

An interesting realization came about while working on the wiring. The start point at the B9G socket was going to be the most challenging to solder. It was the closest point to the cable assembly, which sooner, or later, was going to be a challenge to work around, for getting wires into terminal holes and staying there, while also manouvering a soldering iron and the solder. I had decided to simply start with Pin 1 and work my way in sequence to Pin 9 and all should be OK. That worked for the first two pins before two things came to light. The first was the anticipated issue of the cable itself getting in the way. The second was my mindset.

The adapter I was modeling mine after was the Stark design that Bruce MacMillan is lucky enough to have in his collection. I had been following the wiring runs in the Stark so closely for the first two pin runs I lost sight of the fact this original Stark was a much narrower design than the case I was working with. I was using very short runs of wire between the sockets, which were adding to the problems of the loose ends staying put before soldering them in place and ease of access for the soldering iron. At that point, the valve lit and I became a bit more generous with the wiring runs. You can see that transition in the first two photos.

The third photo was taken after the cable assembly was finally positioned and all the continuity checks, within and between the four sockets, checked out absolutely correctly (several times). This last photo turned up an interesting point.

If you look at the reflection of the camera flash around the cover screw in the lower left corner, you can see a ‘cone of depression’ around the screw. When I spotted this and looked more closely at the reflections around all six cover screws, four in total showed this cone around them: a result of the screws being over tightened. I backed them off just until the cones disappeared and then checked the seam between the cover and case and it was all still snug, but at least the excess stress on the plastic cover was now gone.

The next step will be prepping the cable end to which the Octal Plug will be fitted.

David

[David Dunlop]

I have not got as far along as I had hoped with this work lately. I have been playing tag with a kidney stone again for the last couple of weeks, Grrr!

As noted in the attached two photographs, however, I have been able to trim the cable to the required length, bind it at the appropriate spot at the Octal Plug end and install the Octal Plug Cover on the cable (oriented correctly).

This morning, I was able to trim the 8 conductors in the cable back enough they will extend about one quarter inch beyond the ends of the plug pins when fully seated for soldering in place. I have also been able to arrange the conductors in the correct circular sequence to line up with their corresponding pins on the Octal Plug.

Next step will be to tin the exposed copper stranding, and bend them all to line up with the plug pins. Then the actual soldering of the connectors to the Octal Plug can begin.

David

[David Dunlop]

Clearly I have not had to do a lot of tinning for quite a while. I could not find my small tin of rosin flux anywhere and ended up making a quick run yesterday morning to a local electronics shop to purchase a fresh tin of flux.

With that on the workbench this afternoon, I have now completed the tinning of the eight leads to the Octal Plug. While checking the plug itself for any issues, I finally noticed that whoever the manufacturer of this particular plug was, they took the time to cast in the pin numbers on both the top and bottom of the plug base. Nice touch, as not all manufacturers seem to do that.

I did notice one other thing, which made the close inspection worthwhile. There was a noticeable variation in the size of the openings on the tips of the pins. Not a huge variation, but just enough that only 5 of the eight leads would slide easily through the openings. I took a small steel scratch all from the tool bin and gently inserted it into each hole to match them all to the largest sized opening visible. It was a bit of a challenge getting all eight leads into their allotted pins at roughly the same time. Seven of the eight seated quite easily and as each went into place, I used a pair of small needle nose pliers to gently bend the leads a few degrees to the side of the pin tips to hold each in place until they can be soldered home. The yellow lead, which had the greatest distance to travel from the end of the cable loom to its final resting place, took a little nudging to get it far enough out of its pin end to grab with the pliers, fully seat and get its holding bend applied.

Next step will be the final soldering of the eight leads to the Octal Plug pins and then I will use a small stone on my Dremel to clear away the excess leads. But not before I quadruple check continuity between the Octal Plug pins and the sockets. I’ve made enough of those hasty mistakes over time, I don’t have any desire to purchase another T-shirt!

David

[David Dunlop]

Another productive day at the basement work bench this afternoon.

I was able to get the leads to the Octal Plug pins all successfully soldered in place, and the excess trimmed away. Then it was a simple case of using a small stone on the Dremel to smooth the pin tips up.

The actual real challenge was getting the Octal Plug Cover to snap in place. It did not look like much but was a tough little sucker to pop into its final position. I was pleased to see that when the anti strain clamp was finally run home, it ended up exactly where I had hoped it would, at the mid point of the loom binding. The final photo shows how this adapter fits up with my Precision 612 Tube Tester.

Now I just have to sort out two of the tester controls to see exactly what output they are providing, and to what pin locations. A third control, for the Filament Voltages, Precision actually provided a table of voltage values for in their manual, for the 18 available switch points. Since the ARP-3 Valve is the most common in the 52-Set, I will focus on it first. The pin information for the valve is readily available, so it will just be a case of testing each pin on the B7 socket for each switch point on the two controls to see what end values are delivered and when I get the correct values for each pin noting the control settings. It will be basically the same process for the 12Y4G valve on the 5-Pin Socket.


David

[David Dunlop]

I would like to thank, once again, Bruce MacMillan, for bringing the existence of the original Stark valve adapter to our attention and so generously taking the time to post all the necessary details needed to make this clone a possibility.

It is that kind of interaction on this forum among its members that makes this site such a great place!

David

[David Dunlop]

It is interesting how we sometimes miss little details when we plunge into trying to understand bigger issues. It also doesn’t help if we find ourselves juggling several bigger issues at the same time.

Since the Wireless Set No. 52 evolved out of a British design, and the British based valves were retained in that process for the wireless set, a means to test those valves becomes important, and hence the existence of this Stark designed British Valve Adapter which could be used with North American tube testers.

The ARP-3 pentode valve is by far the commenest in the 52-Set and the interesting thing about the manufacture of these valves in Canada during the war is that all of them are marked “ARP-3 EQUIVALENT”. The wording does not seem that significant at first glance, but it screams volumes when you take the time to think about it.

The original British Military ARP-3 is nothing more than a direct copy of the prewar commercial 9D2 pentode valve. There was nothing preventing a company in Canada from getting a licence to produce said valve in Canada. In which case, all those Canadian valves would have simply been marked “ARP-3”. However, tooling up for production would have taken time and money to accomplish. Time being the important factor at the start of WW2.

So something in the list of available pentode valves already in existence in North America must have been a close enough match to work as an ‘equivalent’ to the ARP-3, thereby saving a lot of time getting production up to speed.

As this adapter project came closer to completion, I started looking to see what the history of the ARP-3 was, which was tricky. Lots of data is on the web about the ARP-3 but this data and photos are nearly always of the Canadian Marconi ‘equivalents’. This last weekend, I finally ran across an early wartime Brimar publication and in it found the technical details for the commercial 9D2 valve the ARP-3 was based on. These specs matched the Canadian Marconi ARP-3 equivalent perfectly. In very small print at the end of the article it said. “For further description see the 6K7G.” That was an eye opener! The 6K7G is part of the tube set in the Wireless Set No.19.

When I pulled out my RCA tube guide for North American tubes, the data for the 6K7G was indeed a match for the 9D2, except for the Heater/Filament Voltage. The 6K7G was rated for only 6.3 Volts. The 9D2 and Marconi ARP-3 Equivalent were both 12.6 Volts. So then I checked the 12K7G pentode and there it was. Same specs AND the 12.6 Volt Heater/Filament rating. I then carefully traced out the octal pin arrangement for the 12K7G tube through the British Valve Adapter wiring diagram I had just worked on and arrived at the correct pin arrangement for the B7 British Valve Base. So now i know when the adapter is connected to my tube tester with the Marconi ARP-3 valve in place, I just have to set the tube tester to read a 12K7G tube and I am in business.

I was so pleased with this discovery, I sent an email to a friend in Quebec to tell him about it. His reply was, “Did I not tell you that last Summer in an email?”

After several minutes review of old emails on the computer, there it was. He had! But I had lost that little tidbit while wrapping my head around a bigger problem we had been working on.

Silly me.

David

[David Dunlop]

I have now been able to pull together the appropriate settings for my Precision Series 612 Tube Tester to enable me to test the ARP-3 Pentode valves for my 52-Set Project, using the British Valve Adapter I was able to build.

A long overdue Calibration Test of the tube tester revealed no serious problems, and i was able to do three consecutive tests of the same ARP-3 valve and get readings within two points for the three readings. Quite pleased with that.

David