[EdHolland]

Well, I am still not satisfied with the short wave performance. Something is amiss, according to my "Sixth sense". All the coils seem intact and in good condition. Additionally, I got out the trusty Heathkit grid dip meter, and checked that the antenna circuits were tuning in reasonable track with the oscillator. Next on the checklist will be to look at all the voltages again on the mixer, check resistors, mica caps and bandswitch.

I'm happy enough at the moment that there is still some tinkering to do

[EdHolland]

A a bit more background research on this radio while I wait to be able to continue work (more pics later).

I am now practically certain that the manufacturer was Gilfillan. Comparison with other models by this manufacturer (Thanks Google images) reveals one or two details in common. Not the least of which were the list of patents decal, the use of unterminated wires for the antenna and ground connections, and the particular tube shields that match the shouldered shape of the components. Also type and style of capacitors.

Some probing around yesterday indicated there was little, if anything amiss with the converter tube components (replaced osc grid resistor). Voltages appear to be correct within reason. Cleaning the bandswitch improved band B, and WWV on 5 MHz is coming in stronger. Band C still seems poor, but perhaps that is how it is. Need to test on the long wire again. I've ordered a replacement 6A8 converter anyway, since I like to have spares.

Cheers

Ed

[Phlogiston]

I am not sure if this will be any help, but, here is a WAP that was built by Gilfillan:

https://www.russoldradios.com/blog/corona-boredom

[EdHolland]

  That's funny Russ, I spent a happy while enjoying your postings on this already!

Beautiful job on that poor radio, and I agree with your point about careful rust removal and not painting a chassis.

More work cleaning the bandswitch seems to have helped the SW ranges. So there is progress.

[EdHolland]

Well, I am still confounded. Band C (5-18 MHz) is still really deaf. Sometimes it seems to crackle in briefly, but fades away again. Here is what I have done:

Band switch cleaned thoroughly, and connections verified for the band positions. Seemingly no problems here.
This band is the only one that does not respond to connection of an antenna wire. Medium wave and SW1 (A and B) both show an increase in signal when a temporary antenna is connected. Band C/SW2 gets a crackle, but no increase in signal. All oscillator sections are working

Here is where it gets confusing. The antenna coil for band C seems to be intact, and the band switch connects it in when the switch is in C position. But adding the temporary antenna to the antenna terminal, or the appropriate lug on the coil does not result in an increase in received signal. However, if I slip a temporary coil over the windings, connect one end to the chassis and the other to an antenna, then the set begins to recieve properly.

The detective work continues...

[EdHolland]

Lots more head scratching, not much more progress.

One thing I notice, looking at the design of the coil. As wired, The primary for broadcast is 34 ohms, the primary for Band B (up to 5 MHz) 0.5 ohms and the primary for Band C (up to 18 MHz) is 2 ohms. Measuring their inductances, band B has less inductance than band C. This looks the wrong way around for the two SW coils. More-over, the higher frequency coil is close wound from fine wire, where typically it should be spaced a little, and larger gauge wire for the higher frequency.

I am wondering if the antenna coils for SW are mis wired? It seems possible to change the wiring over and retest - nothing permanent need be done.

[RodB]

Suggestion, if you have a rf signal generator try sending a modulated signal at the middle of each band and see if it comes in at the right place on the dial.

[EdHolland]

Thanks Rod,

I've got the oscillator section working fine, it tracks +460 kHz all the way up and down relative to the dial, which is spot on the WWV frequencies, and receives a generator signal perfectly. The problem (if there is one) is the antenna side of things on the other coil set. One thing that I think might be a clue is that the trimmers for the antenna circuits don't seem to do much - the high frequency band C in particular, the trimmer has to be all the way loose, otherwise sensitivity is squashed.

More on this tomorrow...it is a head-scratcher.

[RodB]

My thinking is that if a signal of a certain strength at 8 or 9 MHZ can be received in the antenna on band B then a 14 MHZ signal at the same strength should be received on band C (with some variation). If not, then the antenna circuit isn't doing the job. Of course, you probably already know that but it might verify that you're on the right track. I have always figured that antenna transformer design is a science and some radio designers are not scientists. Keep searching. Rod

[EdHolland]

That is certainly worth investigating - I'll have a look. It can be a bit confusing, as the generator signal (Heathkit) can find its way to the radio via the antenna, the mains etc. First thing is to be sure just how the radio is "hearing" the generator. Worth playing around.

Thanks,

Ed

[EdHolland]

This morning I tested the antenna coil miswire theory. It was not the answer, but easy enough to verify. I have another thought on this, but will run some more tests before reporting back.

Cheers

Ed

[EdHolland]

While I am thinking, here are some more pictures. The only thing missing from the chassis are the tall "Solar" electrolytic caps. I am still deciding whether to cut these open and restuff, or replace them, leave disconnected and add capacitors under the chassis.

Anyway, here is a view of the recap work:

   

And here is the top, valves back in place and playing on AM.

   

[Paul Philco322]

Gee the chassis cleaned up very nicely. Hope you get it all to your satisfaction soon. 

Paul

[EdHolland]

I should mention two things about the under chassis pic:

1) The electrolytics are only tacked in place (safely enough) for a bit of testing. They won't stay like that

2) The push button and adjustable coils for the AM presets are not installed. They were tested briefly over the weekend, and work nicely. For now, I want them out of the way while I continue testing.

[EdHolland]

Some possible insight today. I didn't get long with the radio, but had a moment to switch on and poke about, literally, touching low voltage signal points in the front end and IF chain. Twice, I was able to surprise the set into a state of improved sensitivity, suggesting that something is either unstable, or there is an intermittent contact.

First thing will be cleaning the switches and valve sockets again, rotor contacts on the tuning cap etc. then another look at some voltages. Helpfully, even though I am schematic-less, valve data sheets give a good idea what to expect, plus example circuits and component values.

Something is slightly amiss, but it is playing hard-to-get. It is times like this where it is good to be doing this as a hobby, rather than for a customer