Philips 543A problem detector stage

[Radio1950]

Hi again

when you solve that mystery.


Expected IF and RF sensitivity.

I believe you should achieve 70 - 100 uV EMF at the mixer G1 for 10 dB S/N.

And RF should be about 60 uV EMF or less at the antenna input for 10 dB S/N.

Pls ignore my previous expected test levels which were "Sig Gen Volts PD".


Have a look at my thread https://www.vintage-radio.net/forum/...d.php?t=186094


Good luck.

[Christoffrad]

Thanks again for comments.
I think I'm getting about 2mV sensitivity still into V2 G1
Today I checked the 100pf capacitor feeding V2 G1 (Lifting one end to measure it's value and proving that sensitivity is not changed as it is reconnected with the wafer switch)
I also checked the resistances of the IFT winding (L21 and L22 Trader) (L28 and L29 Philips) they measure 8ohms and 5ohms just as they should.
I also looked for any lack of decoupling on C23, C38, C40 and C41 (C21, C27, C32 and C33 Trader) looking with the scope --A bit tricky but no obvious signs. Also tried bridging an extra decoupling C on each in turn (though I couldn't reach the first C) Suggested by Peter.
Also looked for oscillation on anodes and grids with scope via 220K resistor to avoid loading (suggested by Paula) but could see nothing other than the pick up of local oscillator. Also no evidence of any 10.7 radiation (using another FM receiver to act as a pick up - as close as I could physically get it.)

So everyone... have I overlooked any suggestions so far? I'm up for further checks and measurements.
BTW I have a Ferranti set that is somewhat similar. It has occured to me whether it would be possible to link between them as a method of circuit elimination. But of course the stray inductances and capacitances etc would probably make this out of the question.
Incidentally, the strength of Radio5 here is strong and as a consequence there is no audio distortion (unfortunately they don't broadcast any music so I can't be sure) but I think the focus should NOT be on the audio distortion.
I have re-connected everything (except for retaining the external C on the IFT) just too make sure everything is still as it was.


C

[PJL]

Did you put a link across the 10.7 IFT secondary?

[Christoffrad]

Further tests have been done today based on test method proposed by Paula (Thank You) using 220k resistors in series with the x10 scope probe to reduce loading and whilst keeping the scope sensitivity constant, observing the amount by which the input signal level needs to be adjusted to return the reading to the same displayed level.

with the oscillator grid and AGC capacitor shorted to chassis. Sig Gen set to 470kHz no mod. All scope measurements at fixed setting on X10 via 220k for 1 div p-p

V2 grid - Sig gen 900uV
V2 anode - sig gen 5mV
V3 grid - sig gen 120mV
V3 anode - sig gen 640uV

V2 gain = x180 = 45dB
IFT Insertion Loss = x 24 = 28dB loss
V3 gain = x5.3 = 14.5dB

This seemed to confirm yet again that the IFT is at fault. Finally tonight I managed to remove the IFT can and make measurements of the coils and capacitors:

Primary 908uH (895uH with can) resonating capacitor 108.7pf (109.2 with can) (110pf service sheet ) measured range :- 533uH to 1.057mH
Secondary 618uH (610uH with can) resonating capacitor 132pf (135pf with can) (195pf service sheet) measured range:- 234uH to 594uH
Followers of this thread will note that the secondary capacitor has been corrected to 195pf but set performance did not restore.

See thumbnails of the IFT

I can't see anything wrong here
Unfortunately I don't posess any suitable IFT that I can try in place.

BTW there is no evidence of any oscillation and the V2 HT decoupling C has been checked. The 10.7 IF and switch has been previously eliminated by being completely isolated in a previous test.

C

[Mr 1936]

Hi Christoffrad

If by now you haven't decided to take up stamp collecting, here are a few thoughts.

The photos have reminded me that this is a "side by side" coil arrangement, where the tuning slug positions perhaps have less influence on coupling than the "end to end" style. The inductances I calculated for a previous post were 570 uH secondary and 1000 uH primary. These are near the top of your measured ranges. Clearly, it's important that you get a genuine resonance which will occur at two places, one either side of the "core in centre-maximum inductance" position. If there's only one peak that's a warning sign (but you knew that anyway).

It seems to me that this radio has two states which it can jump between, 1) High sensitivity but signs of oscillation. 2) Very low sensitivity, with apparent heavy loading of the secondary of the IFT. This "bipolar disorder" isn't the sort of thing normally given by failed components, which don't mend themselves.

The pictures of the IFT don't show anything wrong that I can see. Even if the wax had absorbed loads of water, it would be lossy all the time, not just sometimes.

It would be nice to measure the IFT completely out of the radio circuit, using a Sig Gen and scope. The nuisance is that the source and load impedances it is designed for are so far removed from the cosy world of 50 ohms that it creates difficulties. I suggest (and I haven't tried this) that you place three 100 k resistors as a series string in series with your Sig Gen output and feed it to the primary. The idea is to reduce the stray capacitance of the resistor bodies by using several. Your scope probe can be direct on the secondary, maybe deliberately reducing the tuning capacitor a bit to allow for scope probe capacitance. If you get a proper double peak on both cores on the bench, that puts the IFT in the clear. The resistive impedance at the input to the primary will depend on the coil losses (unloaded Q) and I'd expect something like 100 K so the signal generator OC voltage will be potted down by a factor of about 4 at this point

[PJL]

Originally Posted by Christoffrad

V2 grid - Sig gen 900uV
V2 anode - sig gen 5mV
V3 grid - sig gen 120mV
V3 anode - sig gen 640uV

Probably me being stupid but what is the test procedure?

1. The sig gen is always connected to V2 grid and you are moving the scope probe to each test point then adjusting the sig gen level to get 1V peak to peak?
or
2. The scope is permanently on ? and you are moving the sig gen to each of the test points and then adjusting the sig gen level to get 1V peak to peak?

Due to the sig gen loading, you can only sensibly measure overall stage gains by injecting the signal on a grid where it is isolated from the next IFT.

[PJL]

I would recommend we go for a simple test and assume the 220K series resistor will reduce the loading enough to measure around the IFT's.

a) Realign the IF now you have reassembled it.
b) Set the sig gen to say 500uV and connect via capacitor to V2 grid. Leave it there for the remaining tests and do not adjust the amplitude.
c) Using the 220K resistor (leads must be kept short) measure the peak to peak on V2 grid, V2 anode, V3 grid, V3 anode. You can change the scope input sensitivity to take the reading.

[PJL]

You could independently test each stage in-situ to eliminate feedback:

a) Put the sig gen to V2 grid as before. Lift the link between AM and FM IFT1. Pull V3 out as the grid will be floating. Put the scope on V2 grid and note peak to peak level.
b) Put the scope on AM IFT pin you have lifted connection from. Peak the AM IFT and note the peak to peak level.
c) Replace the link. Put V3 back in.
d) Put the sig gen on V3 grid. Lift the link between AM IFT2 and V3 detector. Put the scope on V3 grid and note peak to peak level.
e) Put the scope on AM IFT2 pin you have lifted connection from. Peak the AM IFT2 and note the peak to peak level.
f) Replace detector connection.

NOTES:
The scope should always be used with the 220K series resistor.
Readings of the sig gen output are required as a baseline as the 220K will attenuate the 470KHz signal and the sig gen on V2 grid will be loaded by the IF filter.
The set will need alignment again.

[Christoffrad]

At last some real progress here.
Having removed the 1st IFT (yet again) I carried out some measurements of the Qs prior more invasive work (thanks Paula).
By measuring 3dB points -- Primary Q = 86, Secondary Q = 150
Then by a method of forming a series resonant circuit of the tuned circuit, feeding via 50ohms with 1ohm shunt across the series network and measuring across the C via x10 probe:
Primary Q = 100, Secondary Q = 113
Whilst a Q of 150 seemed a bit unlikely these tests indicated that there was nothing obviously wrong.
I proceeded to re-solder the litz wires (previously suggested) and clean between the tags with Servisol. I also fitted the two ceramic capacitors forming the required 195pf resonance for the secondary into the can. I left the original C in place with one leg lifted.
After refitting the IFT I retuned the whole IF according to the service data. The set now essentially works however it was not possible to peak the IFT1 secondary without oscillation.
When set for acceptable stability it means that the set now has a sensitivity of about 250uV emf (measured driving 50ohm sig gen directly via 0.01uf into the aerial). But this is real progress!

Using a 10n ceramic capacitor to temporarily bridge some decoupling capacitors in turn, I have eliminated C38 C40, C41 (Philips) (C27, C32, C33 Trader). Also tried extra decoupling on the main HT rail. I'm reluctant to pad the IF secondary since the set should be capable of working without.

C

[Radio1950]

Well, from a PM, and no more posts, it seems that you have a working receiver.
Perhaps not as sensitive as some other types.

Your RF sensitivity of 250uV EMF directly into the antenna circuit is the same level as 125uV PD.

I am assuming that your original fault "detection distortion" is gone, and that the cause of this observed symptom is insufficient RF level into the detector circuit?

Is this journey's end?

If so, well done.
We all learn.
Thanks for sharing.

[Christoffrad]

Yes you are correct, The only remaining query concerns the Tone/bass boost switch. I have measured that it has a 3dB effect approxinately which seems rather little. So I would be grateful for any views on this. All paper (black tar) capacitors have been changed.

I would like to thank all those who made contributions to this thread. All your comments were much appreciated.

Chris