Marconiphone T18DA - stations quiet

[ms660]

Cheers, that should do.

Lawrence.

[Nuvistor]

Does the modulator work correctly on another radio?

[martin_m0nxp]

So tonights fiddling....

I replaced the external aerial antenna coupling cap as it was a wax tubular I hadn't changed. On removal I tested it and sure enough it was fine. Others that have tested fine I've put back, but this one was so sticky I couldn't face handling it again, so it got replaced anyway.

I haven't changed any other caps to the left of V2 with the exception of the AGC decoupling cap, so there is still potential for a bad capacitor. I've measured what I can in circuit and none appear to be leaky, and I don't really just want to change them for the sake of it.

I tried feeding the antenna directly onto G1 of V1, both with and without a coupling cap and got far worse results than with it connected in the correct location.

Having swapped the two WD142s around last night I hadn't re-taken my AGC measurements, so I figured that would be an interesting experiment given the diode sections are now performing different functions. No significant difference (-1V on junction with IFT1 before the swap, -0.5V after the swap).

I then decided to see what range the AGC line went through (measured at the same point) with a really strong signal present. With the output from the modulator touching the internal antenna (no electrical connection) the voltage went from around -0.5V when fully de-tuned to -3.2V when tuned to the strongest point. I'm not sure what the expected range for this is (the trader sheet gives no indication), but it seems fairly consistent.

Whilst I had the modulator output very close (i.e. touching) the internal loop antenna the radio sounded pretty good. The volume of the music was what I would have expected and was slightly louder than the 'off tune' whistles if I de-tuned it. (Nuvistor: Yes - the modulator output sounds okay on a modern radio).

I wonder if everything is actually working as it should, it just needs aligning now, hence why it's a bit deaf.

So that leads me onto a question.... how clean (i.e. free from harmonics) does a test 465kHz signal need to be? I have an FPGA lying around I could easily use to generate a 465kHz square wave - would that do? what sort of amplitude does it need to generate? Do I need to modulate it with an audio tone?

[cathoderay57]

Hi Martin, for an aerial, are you only using the internal loop antenna or have you tried a long wire? Does the long wire give reception on any stations? If you cannot receive R4 LW 198kHz at all with a long wire that would be unusual for a set that is merely a bit deaf. It looks from the Trader Sheet as if the IFT trimmers are dust core slugs. Does it look as if they have been tweaked previously? This is not a rare phenomenon unfortunately. To align the IF you should ideally have a 465kHz sinewave oscillator amplitude modulated by an audio tone. Personally I wouldn't attempt it with a squarewave in view of multitudinous harmonics. If you get to the point of doing alignment be careful adjusting the dust cores (if that is what they are) as it is easy to break them apart with a screwdriver especially if they have been sealed with paint or wax. Better to use a plastic tool with a suitable flat filed on its end. Cheers, Jerry

[cathoderay57]

Rather than trying to construct a sig gen I recommend that you put out a request to borrow a suitable sig gen, together with a frequency counter (unless you could be convinced that the sig gen had recently been calibrated and was spot-on) on the Sets, Parts & Service Info wanted part of this Forum, from a member living near you. Alternatively someone at a local Amateur Radio Club would probably lend you suitable kit. As to the signal level required, variable output ideally up to 500mV I'd have thought.

[martin_m0nxp]

Sadly I fear you're right - the effort to construct a sig gen is never worth it (even if it only took a couple of hours) when you can buy them on Amazon for less than the cost of a UL41:

https://www.amazon.co.uk/KKmoon-1HZ-.../dp/B0748H1ZPM

The price is almost less than the cost of petrol to pick one up and return it to someone local. Not sure how good it is, but I can check the calibration of the output frequency using a scope - the rest I'm sure is good enough for this purpose. I've no idea if it can actually modulate the output signal - if it can't I'm sure something can be constructed to do so.

for an aerial, are you only using the internal loop antenna or have you tried a long wire?

I've tried both on MW and LW. The long wire can just about pull in a station on MW, but it's faint. I've so far not had anything on LW, but a modern radio in my kitchen doesn't really get much on LW with the same piece of wire, so I've sort of ignored that for now. I don't have a way of generating a LW signal myself.

Part of the problem with all of this is that there is a vast amount of electrical noise in my house. I've no real clue where it comes from (there's just too much going on to even hope of isolating it) but it covers the whole house and well into the garden. The modern transistor radios (perhaps with their narrower filters) can pick out the only 3 MW stations on air down here, but even then you wouldn't want to spend any length of time listening to them. At it's best the Marconiphone has just about been able to pick up the strongest one (with the same antenna used on the modern ones). Perhaps it just has less rejection so the AGC is pulling everything down due to the noise?

I have taken the radio to the back of the garden where there's the least amount of QRM, but even then it's not a great test as the noise could well be conducted back through the extension lead, and it's not really practical to tinker down there. The number of stations heard didn't change but the one faint station was clearer (but not louder).

be careful adjusting the dust cores (if that is what they are) as it is easy to break them apart with a screwdriver

I'll be honest - I'd been hoping to avoid touching them for this very reason (and also because I don't really have the tools or experience to perform an alignment). I do however have a trimmer adjustment tool - plastic with a small metal flat end.

The final problem I'm really having with all of this is lack of experience - I don't actually know how a 1949 radio should perform compared to modern ones. I don't have the experience to recognise important symptoms that might already be presenting themselves (e.g. is this sound normal as you tune from a station). This is all okay though - the reason I purchased this radio was for the very purpose of learning how vintage electronics worked and getting this experience.

So.... next steps....

1. Wait for the valves to arrive from ebay (the UCH142 has already arrived). Replace and see what makes a difference.

2. Make / borrow / buy something to generate 465khz modulated by 1kz and do the alignment procedure.

3. Find someone from here who is local and see how the radio performs away from my house and compared to a known working radio.

[cathoderay57]

Sounds like a plan. Since you have a scope it would be useful to see what the radio oscillator output looks like as you tune across the bands. It probably won't work if you connect the scope directly to the uch42 osc circuitry but if the scope is sensitive enough you should be able to hold a probe close in or make a small coupling coil to verify that the oscillator is running, particularly on LW but I guess if you have no LW reception it's a moot point. You seem to be in a lousy reception area. Here in Bristol R4 LW is loud and clear and there are about 6 stations on MW with useable signal. Good luck.