Marconiphone T18DA - stations quiet

[martin_m0nxp]

Hi Folks,

Just looking for a quick pointer in the right direction if I may....

I have a Marconiphone t18 da, which I've got to the point of 'almost working'.

Symptoms now are:

* Audio stage seems fine - as I tune I get whistles etc... as I go over the dial. The UL41 seems to get very hot but apparently this is normal? There is some DC present across the output transformer (12V into 500R, so about 24mA of DC). Is this normal? I guess this is die to the way single ended valve amplifiers have to be biased?

* Very insensitive - I can't receive any 'real' radio stations, even with an external aerial, however I can receive the output from a local AM modulator.

* When using an AM modulator to generate a strong signal the audio output is very quiet - as soon as I tune to one side or another the off-tune whistle is much louder than the modulated music was.

* The dial lamp is very dim. I don't quite understand this one - the dial lamp is placed across one of the mains dropper resistors and in series with the rest of the set. It's supposed to be a 6.2V 220mA - I replaced it with a 6V 330mA (the closest I could get). The dropper resistors all measure the correct value and the set draws the expected 150mA (the schematic says 100mA for heaters and 55mA on the HT).

* The HT is a bit low. The schematic calls for 225V, I measure around 200V. Would this make a massive difference?

I've replaced all caps that potentially may have any large voltages across them. I haven't replaced all the caps in the front end, but I have measured them for shorts (none found). The caps in the front end are mostly 100pF silver mica. All resistors measure fine.

So my question is:
Is the lack of sensitivity going to be a problem before the combined mixer/oscillator or between the mixer and the AF amplifier, or a problem with the AF amplifier itself?

[martin_m0nxp]

Just replying to my own post: I've just realised the dial-lamp rating doesn't make sense. The set is supposed to take 150mA. This is drawn through the dial lamp and a 50R resistor in parallel.

This means that even as designed the 6.2v lamp would only have 2.7V across it. No wonder it's dim! I reckon the schematic must be wrong - I think it's supposed to be a 3V bulb.

[Nuvistor]

Appears to be instability in the IF amp or possibly mixer stages, the small value caps in the front end are likely good.
Check the AGC voltage, if it’s high that will reduce the radio sensitivity and another pointer towards instability.

[martin_m0nxp]

The schematic doesn't show any voltages on the AGC circuit... what's normal? I've checked the voltage divider from the 1st AF diode plate, and the subsequent decoupling cap (which I'd replaced) and they all check out fine.

...however I haven't replaced what is described as the 'AGC coupling' cap. It's between the diode plates of the 1st and 2nd IF stage (so the 2nd IF diode plate connects to the AGC line via this, the diode plate of the 1st IF doesn't have a coupling cap at all). It's a 50pF tubular ceramic. It measures very high resistance, but may have lost it's ability to act as a capacitor..... I didn't replace it as i figured it didn't have much voltage across it and ceramics are usually fine, but does this sound a likely candidate?

[cathoderay57]

Hi Martin. According to Trader Sheet 920 the dial lamp should be 6.2v 0.22A. You are unlikely to find an exact replacement but 6.3v 0.3A should work fine. Don't forget that the HT current also flows through the dial lamp so the total will be in excess of 150mA. The anode current of the UL41 in this set is supposed to be 33mA plus another 6mA via Grid 2. Therefore with a 100% valve and HT of 225V you should expect to see 7 to 8v across R17 the 200R cathode resistor. It would be informative to measure the voltages on all of the valves as this would give some clues as to the fault location(s):

V1: X142/UCH42 mixer anode: 153v ; grid2: 67v; oscillator anode: 60v.
V2: WD142/UAF42 anode: 153v; grid2: 67v.
V3: WD142/UAF42 anode: 36v; grid2: 39v.
V4: N142/UL41: anode: 215v; grid2: 153v.
V5: U142/UY41: Cathode: 225v.
Cheers, Jerry

[martin_m0nxp]

I still think the trader sheet is wrong on the dial lamp. The trader sheet shows the anode current on the rectifier as 55mA and the heater current is 100mA - so the total current through the lamp and the 50R resistor in parallel is 155mA.

With the resistor alone this would give 7.5V... however a 6.2V lamp rated at 220mA would have a resistance of 28R (once at temperature). In parallel with the 50R this gives a resistance of 18R, which at 150mA would have a drop of 2.7V.

I suspect a 3V bulb rated at around 100mA would be better suited - this would have around 2.8V on it so ideal.

The dial lamp is a fairly minor concern though (given it only has one of them I can't imagine it's even that effective!).

With regards to the UL41 - it sounds like it's all running as designed (albeit at around 24ma rather than 33mA - given the valve looks original I guess it's rather tired... the slightly low HT also explains this difference). I just can't believe how hot it gets though! I've ordered a spare off ebay, as I can't believe it can last any length of time at that temperature.

I'll check the voltages again - I'd like to get to the bottom of the lack of sensitivity so will go over all the previous checks I did in case I missed something.

[ms660]

I might be wrong but I suspect the calculation of the voltage across the bulb circuit is a bit more involved.

Lawrence.

[Nuvistor]

If you have been changing capacitors in batches it could be worth checking for mistakes.

[cathoderay57]

If you measure the valve voltages and they are reasonably close then at least you will know the DC conditions are right and that the valves are passing current. After that I would feed in an audio signal at the volume control to check for undistorted AF. Then feed in a modulated IF signal at 465kHz per circuit alignment procedure on Page 2 of the Trader Sheet. Do be extra careful that the chassis is connected to mains neutral and use an isolating capacitor in the earth connection to the sig gen. The symptoms tend to indicate a fault in the IF or RF stages. If you have not made any errors replacing capacitors then a systematic approach injecting signals as above then finally at RF should reveal where the signal path is failing. Cheers, Jerry

[martin_m0nxp]

Okay... so measured voltages:

These voltages were done with the HT at 205 (measured) instead of 225 (which the schematic calls for), so for most of the higher voltages I'd expect them to be ~20V down.

V1 Anode: 130 (expected 153)
V2 Anode: 130 (expected 153)
V3 Anode: 22.6 (expected 39)
V4 Anode: 174 (expected 213)

V1 Osc: 60 (expected 60). Radio whistles whilst this pin is probed
V2 Screen: 57 (expected 67)
V3 Screen: 23 (expected 39)
V4 Screen: 132 (expected 153)

Onto the AGC side of things:

I removed C26 - the 50pF ceramic between pins 3 of V2 and V3 and measured it - 51pF and no leakage, so I put it back.

I measured the voltages around the AGC line with no signal present:

@ Pin 3 of V2: -0.7V
@ Pin 3 of V3: 0.13V
@ Pin 6 of V1: 0.7V
@ lower pin of IFT1: 1V

I'll be honest I don't really understand how the diode section of the WD142 functions (is it a diode between the cathode and pin 3?) and I don't really understand how the AGC circuit works on this. Is it just adjusting the bias of G1 on V1?

I don't have a signal generator, so for the moment I don't have a way of generating a 465kHz signal to align it. If I get a bit further down the road I may make myself one....

I think the AF side of it is fine - using a strong signal from an AM modulator I get clean (but quiet) audio all the way through and out of the speaker.

[cathoderay57]

Nice work. The voltages seem reasonable and for the IF and mixer/oscillator it looks as though the coils in the anode circuits are conducting dc to he valves. Does swapping around the two WD142 valves make any difference?

[cathoderay57]

The WD142 is a diode between pins 3 & 4 (cathode). On V2 the IF signal at the secondary of the IFT is coupled to the AGC diode anode by C16. The AGC diode load is R9 across which a negative voltage should appear with signals present. AGC is fed to V1, and V2 grid1 via L7. A similar arrangement exists for V3 the detector but in this case L9 is coupled directly to the diode anode and R10 is the diode load resistor. When you feed in the modulated signal, does it only appear at one frequency (as it should) or does it appear as weak audio irrespective of the tuning setting?

[ms660]

Originally Posted by martin_m0nxp

I'll be honest I don't really understand how the diode section of the WD142 functions (is it a diode between the cathode and pin 3?) and I don't really understand how the AGC circuit works on this. Is it just adjusting the bias of G1 on V1?

Yes pin 3 is the AGC diodes anode, here's a simplified description if it helps, the signal for the AGC diode is coupled to the diodes anode via the 50pF capacitor, the diode conducts on the +ve half of the signal envelope making the anode -ve WRT the cathode (chassis) that voltage varies according to the signal strength, strong signal = more -ve, weaker signal = less -ve, the -ve voltage is the bias voltage for the mixers control grid (V1) and the IF amplifiers control grid (V2) those two valves are of the vari mu type, varying the grid voltage varies the gain, more -ve = less gain, less -ve = more gain, from that it can be seen that with a weak signal the gain of the AGC controlled valves is increased and with a strong signal the gain is decreased.

EDIT: Post crossed.

Lawrence.

[martin_m0nxp]

Ah... thanks for those explanations... I've been trying to work out where the negative voltages were being generated!

Is 0.7V on the grid of V1 expected or should this be negative?

With regards to feeding the modulated signal: It appears in only one place on the dial and tuning around it is fairly sharp. I can't really tell whether it's the correct place on the dial - the markings have worn off the dial itself. It was roughly in the middle though, which was about what I expected.

EDIT: I forgot to mention - I hadn't thought of attempting to swap the WD142s - nice thinking. Unfortunately I can't remove V2 from it's socket - not sure how much I dare tug on it!

[ms660]

Careful if pulling on rimlock valves, the retaining clip in the valve holders can damage the glass location pip on the valve, valve goes to air, not good, remove the clips and don't refit them.

Lawrence.

[cathoderay57]

As there doesn't seem to be any decent signal at the secondary of the IFT there will be no AGC generated so +0.7v on V1 g1 is not unusual. It sounds as if the oscillator is running. The possibilities for faults that spring to mind are an open circuit in L1 (L2 is presumed OK otherwise you'd get something on MW); what happens if you stick the aerial wire directly onto V1 g1?
Other possibilities are a duff diode section on V3 (hence suggestion of swapping V3 and V2); open circuit L7 or L9; C6, C7, C14, C15 drifted in value.

[martin_m0nxp]

Okay... so I've measured the resistance of L1, L7 and L9 - all low. Obviously I have no way of measuring C6, C7, C14 and C15.

Thanks for the tip regarding rimlock valves. I hadn't even seen the retaining clip, so now I know (and sure enough the valve came out easily once I'd rotated the clip).

I have now swapped V2 and V3 and seem to be able to pull in a single station now, albeit quiet. I don't know if swapping the valves had anything to do with it as I'd repositioned the antenna wire since I last tested. Maybe just re-seating the valves did it?

I'm still not convinced it's right - I'll experiment further with the modulator tomorrow.

I have, of course, now burnt my finger whilst pulling V3 out, as it sits right next to the impossibly hot N142.

[Restoration73]

Whilst agreeing with all the above, many of these sets may contain the original valves,
so do expect the cathode emission to be a bit lower after 60 years. Replacements are
easily found.

[ms660]

Originally Posted by martin_m0nxp

I'm still not convinced it's right - I'll experiment further with the modulator tomorrow.

What's the modulator?

Lawrence.

[martin_m0nxp]

What's the modulator?

I use one of these:

http://www.6v6.co.uk/index.php?seo_p...g-in-coils-kit

...it was a kit I treated myself to at Christmas. It's not amazing, but it'll do for testing.