UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Radio (domestic)

Notices

Vintage Radio (domestic) Domestic vintage radio (wireless) receivers only.

Closed Thread
 
Thread Tools
Old 12th Nov 2021, 10:45 am   #141
Radio1950
Hexode
 
Join Date: Apr 2017
Location: Buderim, Queensland, Australia.
Posts: 428
Default Re: Philips 543A problem detector stage

I also thought that alignment is being carried out first.

My subsequent suggested approach was for after this, as it is a fault finding approach, if nothing was found earlier during or after alignment, or there was doubt about where to go next.
Radio1950 is offline  
Old 12th Nov 2021, 11:49 am   #142
Christoffrad
Hexode
 
Join Date: Jun 2018
Location: Cambridge, Cambridgeshire, UK.
Posts: 260
Default Re: Philips 543A problem detector stage

Thank you for your continued postings.
Today I embarked on IF alignment. The Philips data requires alignemt in the order:
S33, S32, S28, S29, S32. This corresponds to Trader cores L26, L25, L21, L22, L25
The cores were found to be very close to optimum with the exception of S28 (L21) the core of which had to be screwed much further in and even having done so did not have the sharp peak of the other cores. eg half a turn makes little difference whereas the other cores tune sharply. I am therefore suspicious about the S28 (mixer anode load) tuned circuit. I think the change in audio output was about 3 or 4 dB and having repeated yesterday's measurement at 800kHz via the aerial the set has marginally increased sensitivity from 90mV to 45mV.
I note for the front end tuning Philips advise the use of a dummy aerial. I found a circuit here: https://www.vintage-radio.net/forum/...ad.php?t=56666
So I intend to build it.

Question1: I have been using the generator unterminated and quoting emf values. Would you advise that I include the resistive matching pad version of this dummy antenna circuit and use my 50ohm termination load/ T connector from now on for the front end alignment?

Question2 : Or do you think that the behaviour of S28 (L21) is so suspicious that it should be investigated first. You will recall that S28 is the coil fed by R14 (R13 Trader) which had been changed (with it's decoupling cap) by someone previously to a much higher value (now restored to 2k2).

Last edited by Christoffrad; 12th Nov 2021 at 11:57 am.
Christoffrad is offline  
Old 12th Nov 2021, 3:01 pm   #143
PJL
Dekatron
 
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
Default Re: Philips 543A problem detector stage

If you have to screw the core fully in then that needs investigating as that implies it is not resonating. I am out at the moment so will look when i get back.
PJL is offline  
Old 12th Nov 2021, 4:41 pm   #144
Christoffrad
Hexode
 
Join Date: Jun 2018
Location: Cambridge, Cambridgeshire, UK.
Posts: 260
Default Re: Philips 543A problem detector stage

The 'tuned' position of S28 is now sitting about 1mm proud of the former (measured to the bottom of the slot) having needed about 3 or 4 turns to reach some degree of peak, compared with between 2 and 4mm for the other cores (which peaked sharply).
Having tuned many cores during my time working on narrow band VHF and UHF receivers this one feels distinctly odd. And there must be a reason why it needed to be retuned so far..... like it's associated C has gone OC perhaps. And why did the series resistor get changed ... it seems to hint at a dramatic historical event!
Christoffrad is offline  
Old 12th Nov 2021, 5:24 pm   #145
frsimen
Heptode
 
Join Date: Jan 2011
Location: Croydon, London, UK.
Posts: 766
Default Re: Philips 543A problem detector stage

It is usual for IF transformers to be tuned to the outer peaks, so you have yours set correctly now. That the peak isn't very marked is almost certainly an indication of a fault.

An open circuit capacitor would shift the core setting for a proper peak a long way from where it should be. A leaky capacitor will damp the Q of the tuned circuit, as you did for the bandwidth tests.

A resistance measurement of S28 may give a clue as to the cause of the trouble. The Trader sheet states that the resistance is supposed to be 8 ohms. If it measures much higher than that, there is a problem with the winding itself, probably a break in one or more of the Litz wire strands and that will lower the Q of the tuned circuit. If it measures much lower than 8 ohms, the associated capacitor is extremely leaky.

The resistance measurement won't spot a capacitor with a leakage of more than a few 10s of ohms. That will be bad enough to wreck the Q of the tuned circuit and greatly diminish the gain of the mixer.

I don't think the capacitor is open circuit, but it would do no harm to connect a 100pF capacitor across the winding as a test. If that doesn't restore operation following an adjustment of the core, it would appear that the IF transformer is the cause of your problem.

Some IF transformers are easy to dismantle, I don't know the type you have here, but if it can be opened up you will be able to check the capacitor.

Paula
frsimen is offline  
Old 12th Nov 2021, 6:15 pm   #146
Christoffrad
Hexode
 
Join Date: Jun 2018
Location: Cambridge, Cambridgeshire, UK.
Posts: 260
Default Re: Philips 543A problem detector stage

I have just noticed an error on the Philips data that has the coils showing on the two layout diagrams contradicting each other.
The coil defined as S28 on the circuit and underside layout is S29 on the top chassis layout.
The core that has the suspect tuning is in fact S29 ie the secondary of the first IFT. So everything I have prevoiusly said about S28 should be taken as S29 which is L22 on the Trader sheet.
So the top chassis drawing is wrong. S28 is the coil towards the RF front end side of the set (Trader is correct)
I've measured the resistance of the coil. It is correct 5ohms.
Luckily there is reasonable access to this IFT and it should not be too difficult to remove and check it's C. Or I could just assume that the C is open circuit (a bit risky) and see what happens if I tack another in situ.

see thumbnails

Gods are working against me on this one!

C
Attached Thumbnails
Click image for larger version

Name:	543parta.jpg
Views:	48
Size:	51.1 KB
ID:	245654   Click image for larger version

Name:	543partb.jpg
Views:	44
Size:	103.7 KB
ID:	245655   Click image for larger version

Name:	543partc.jpg
Views:	44
Size:	143.1 KB
ID:	245656  

Last edited by Christoffrad; 12th Nov 2021 at 6:36 pm.
Christoffrad is offline  
Old 12th Nov 2021, 6:39 pm   #147
G6Tanuki
Dekatron
 
G6Tanuki's Avatar
 
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 13,951
Default Re: Philips 543A problem detector stage

Having followed this thread, I'm wondering if you have a dry-joint-which-is-sometimes-acting-as-a-diode within the problematic IFT.

The 'disturb it and it works OK for a bit' thing could be the dry-joint becoming less-dry following electrical/mechanical provocation; the odd tuning-behaviour, distortion and signal-loss would be explained by diode rectification in the dry-joint and associated harmonic-generation/non-linearity.

I'd be popping the IFTs out and reflowing all the soldered joints inside the cans.
__________________
I'm the Operator of my Pocket Calculator. -Kraftwerk.
G6Tanuki is offline  
Old 12th Nov 2021, 7:06 pm   #148
Christoffrad
Hexode
 
Join Date: Jun 2018
Location: Cambridge, Cambridgeshire, UK.
Posts: 260
Default Re: Philips 543A problem detector stage

Yes, Access is not too bad on the underside of this IFT so I think I'll pop it out tomorrow.
I trust there is concensus......?
Christoffrad is offline  
Old 12th Nov 2021, 10:49 pm   #149
PJL
Dekatron
 
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
Default Re: Philips 543A problem detector stage

Quote:
Originally Posted by Christoffrad View Post
during today's measurement the audio faded away two or three times. I found that by touching the link between the two secondary windings of the first IFT with the scope probe, as well as creating a loud crack also restored operation.
Before dismantling the 470KHz IF, I would be tempted (if it is easy) to wire out all the 'extra bits' from the circuit by bypassing the 10.7MHz IF and switch. Lift the link from the 465 IF to the 10.7 IF and move the other connection to the 10.7 IF to the 470 IF and do this for the secondary and test and if this does not improve things try the primary.

Last edited by PJL; 12th Nov 2021 at 10:59 pm.
PJL is offline  
Old 13th Nov 2021, 3:13 pm   #150
Christoffrad
Hexode
 
Join Date: Jun 2018
Location: Cambridge, Cambridgeshire, UK.
Posts: 260
Default Re: Philips 543A problem detector stage

Here is the latest finding on the trail.
This is the IFT removed from the set and one end of capacitor C37 (C37 Trader) lifted for measurement. As you will see, it measures 130.4pF. According to the parts list this capacitor should be 195pF.
Without the capacitor the IF coil measures 607.2uH. This suggests a resonant frequency with 130pF of 566kHz. With 195pF this becomes 462kHz which of course takes into account that I have already moved the core. For resonance at 470kHz with 195pF the inductance needs to be 588uH which will probably correspond to the core returned to it's original position.
This crearly accounts for the lack of correct tuning.

I won't say i've fixed it until I have it fully tested but this now looks very promising.

My previous posts described conditions where I have been able to change the set's behaviour by touching the connections of this IFT with the scope probe so there is some uncertainty about this capacitor inclining me to isolate it. No details are given in the Philips service data for the capacitor type but I suppose a polystyrene replacement is appropriate? I have seen polystyrene caps used in this role. Does this type have the right temperature coefficient and stability?
This cap is buried inside the can. I have carefully raised the retaining crimps to remove the can but it does not readily open.
Next question: How important is it for this cap to sit inside the can? would it be more prudent to just isolate the old cap and fit the replacement externally to the coil tags?

Chris
Attached Thumbnails
Click image for larger version

Name:	Philips 543A IFT1cap.jpg
Views:	43
Size:	53.1 KB
ID:	245700  

Last edited by Christoffrad; 13th Nov 2021 at 3:22 pm.
Christoffrad is offline  
Old 13th Nov 2021, 3:25 pm   #151
PJL
Dekatron
 
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
Default Re: Philips 543A problem detector stage

You would need to replace it with a high stability capacitor, mica or HS ceramic types are normally used. I would also run over the joints with a soldering iron whilst you are there.
PJL is offline  
Old 13th Nov 2021, 4:03 pm   #152
Christoffrad
Hexode
 
Join Date: Jun 2018
Location: Cambridge, Cambridgeshire, UK.
Posts: 260
Default Re: Philips 543A problem detector stage

I'm sure I have some of the small tubular ceramic that we used at Pye/Philips for this sort of application and I can do a crude temperature check. I don't know the voltage rating but I think would have been suitable for 30W VHF Tx.

C
Christoffrad is offline  
Old 13th Nov 2021, 5:53 pm   #153
Mr 1936
Heptode
 
Join Date: Mar 2017
Location: Romsey, Hampshire, UK.
Posts: 511
Default Re: Philips 543A problem detector stage

Hi

If it were me I'd use silvered mica (+50 PPM tempco), with a NP0 type ceramic as an alternative. I see no objection to external mounting on the IFT pins for one capacitor, but wouldn't put both capacitors there just in case they coupled to each other. Assuming the cap is shorted by the winding, the DC voltage rating isn't overly critical.
Mr 1936 is offline  
Old 15th Nov 2021, 7:12 pm   #154
Christoffrad
Hexode
 
Join Date: Jun 2018
Location: Cambridge, Cambridgeshire, UK.
Posts: 260
Default Re: Philips 543A problem detector stage

After a day of recovery I resumed my fight with this 543A today.
Having removed the IFT and measured the secondary inductance of the IFT and descovered the low value of it's tuning capacitor (130pf instead of 195pf) I first decided to assess the temperature characterstic of the IF coil. Using the bread proving setting of 40degC on our oven I warmed the IFT and having lifted one end of the 130pf cap. I used a cheap thermometer to confirm the temperature. At about 45deg I quickly connected my LCR and found that the inductance had only changed by +1uH over 30deg

I made up 195pf with two ceramic plate capacitors and checked the temperature characteristic similarly. The change was +0.5pf over 30deg
So I concluded this acceptable, (a change of 1uH and 0.5pf would change the resonance by 1kHz)
I insulated the leg of the old cap and reassembled the IFT circuitry with the capacitors added externally.
I tuned the IF strip again according to the service data, injecting 470kHz into V2 grid. This time whilst the secondary core tuned and I could find two resonances (as expected as the core tuned through) the peaks were once again not sharp as they should be. I was therefore not surprised when the performance was unchanged (still the low sensitivity and subsequent audio distortion)
Here are some figures:
With the volume control set for mid position and scope measurement of the amplifier output into 5ohm dummy load...
1) 13mV emf (30% mod) is required into V2 grid to achieve 1.4V p-p audio
2) direct connection of generator into aerial 30% mod on 800kHz requires 25mV emf for 1.4V p-p audio

In other words little has changed from previous measurements.
As the IF seems to remain so deaf I have not bothered to adjust the front end.

Chris
Attached Thumbnails
Click image for larger version

Name:	Philips 5543a IFTcap.jpg
Views:	30
Size:	72.1 KB
ID:	245889  

Last edited by Christoffrad; 15th Nov 2021 at 7:26 pm.
Christoffrad is offline  
Old 15th Nov 2021, 8:50 pm   #155
frsimen
Heptode
 
Join Date: Jan 2011
Location: Croydon, London, UK.
Posts: 766
Default Re: Philips 543A problem detector stage

The attenuation of the volume control at mid setting is still quite high, so allowing for that, the sensitivity is more like 1mV for 1.4V p-p (50mW in this case). Still rather deaf, but rather better than previously.

One fault down, but there is still another one lurking somewhere. One set of measurements that could be made is to check for HT leakage onto the grid circuits of the IF valves.

To do that, remove both the ECH81 and the EBF80. Measure the voltage present on the pin 2 of the each of the valves.

It should be zero but if there is a leakage problem on the ECH81 valve holder, you may find a positive voltage. A good clean up around the valve holder may help reduce any voltage that is present. The pin arrangements on the EBF80 make it less likely that this will be a problem on that valve.

Next, plug in the ECH81 and measure the voltage across the AGC capacitor. It should still remain close to zero. If it goes positive, it would be worth trying another ECH81.

Plug in the EBF80 and recheck the voltage across the AGC capacitor (ideally with no signal applied). It should be close to zero or slightly negative. If it goes positive that may indicate a problem with the EBF80.

Final test, monitor the screen voltage (pin 1) of the ECH81 as you tune across a strong signal from your signal generator. Check that the voltage becomes rises on a strong signal. If it doesn't, check for continuity between the valve pin 2 and the valve holder. You have already proved the equivalent pin on the EBF80 is making good contact in an earlier test.

The above may not identify a problem, but at least they will eliminate some potential suspects.

Paula
frsimen is offline  
Old 15th Nov 2021, 9:29 pm   #156
PJL
Dekatron
 
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
Default Re: Philips 543A problem detector stage

I also suggest bypassing the 10.7MHz 1st IFT and switch. You could lift the link wire from the 470KHz to the 10.7MHz IFT, lift the wire from V3 grid and wire it direct. This would help eliminate some other stuff.
PJL is offline  
Old 15th Nov 2021, 11:36 pm   #157
Christoffrad
Hexode
 
Join Date: Jun 2018
Location: Cambridge, Cambridgeshire, UK.
Posts: 260
Default Re: Philips 543A problem detector stage

I wish I could be more sure that the IFT problem is fixed. When tuning the coil with the changed capacitor, there was no sharp peak. It was like tuning an over coupled circuit falling sharply at the edges with a slight sag between the slight peaks.
I know it's illogical but do I need to remove the old C altogether? Does the new C need to be inside the can?
Thanks Paula / Peter for your suggestions. Tomorrow is another day!

C
Christoffrad is offline  
Old 16th Nov 2021, 12:57 am   #158
PJL
Dekatron
 
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
Default Re: Philips 543A problem detector stage

You did do the alignment with maximum volume and adjusting the generator output to maintain a quiet listening level? This is to avoid the AGC interfering with alignment.

I assume your 13mV test into V2 grid was injecting 470KHz? As we are looking at the IF performance, I suggest all measurements of sensitivity are done at maximum volume with 470KHz into V2 grid as Paula suggests.
PJL is offline  
Old 16th Nov 2021, 1:18 pm   #159
Christoffrad
Hexode
 
Join Date: Jun 2018
Location: Cambridge, Cambridgeshire, UK.
Posts: 260
Default Re: Philips 543A problem detector stage

Actually the answer is Yes and No
Yes the test was done at 470kHz but not at full volume. (mid volume)
At Full volume setting and 470kHz into V2 grid 30% mod the generator needs to output 1.4mV emf to produce 1.4V p-p across 5ohm dummy loudspeaker.
Christoffrad is offline  
Old 16th Nov 2021, 4:08 pm   #160
Christoffrad
Hexode
 
Join Date: Jun 2018
Location: Cambridge, Cambridgeshire, UK.
Posts: 260
Default Re: Philips 543A problem detector stage

Today I have carried out a series of tests suggested by Paula and Peter. Note I have not yet built the dummy aerial so in the case of the test stated above and the tests today fed to the aerial, this is a direct connection (via 0.1uF) from the signal generator (unterminated).

1) Checking for HT leakage onto grids: V2 and V3 removed: Result - there is no measurable voltage (perhaps 1mV)

2) voltage on AGC capacitor (C49 Philips, C24 Trader) - With V2 (ECH81) re-inserted measured voltage = -0.35v
Then with also V3 re-inserted, measured voltage = -0.42v

3) V2 screen volts as strong signal tuned through (measuring pin1 of EBF80) Screen v0lts measured with scope: No signal = 70v (exactly as service sheet)
5mV EMF input from signal generator gives screen = 75v
15mV gives 85V
50mV gives 110V
100mV gives 130V

Next the link between the FM and AM secondaries on the first IFT was lifted and the FM IFT and associated switch bridged out.

The grid bias on V2 and V3 was then measured :
V2 grid = -0.8v
V3 grid = -1.1v

There is no change in reception of broadcast stations.
Christoffrad is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 1:25 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.