Philips 543A problem detector stage

[Radio1950]

Thanks for photo posts.

Leave sig gen at 200 uV, move CRO probe directly to speaker voice coil ie output.
Earmuffs on.
Volume up a little.

If CRO waveform shows distorted sine, turn volume down and check if distortion decreases, or varies with vol control
Then go back to the detector output with CRO probe, and move probe towards the speaker, following the audio path, with hope you will pick up something.
Leave volume control at a point where distortion is noticeable.


If CRO shows sine-wave at speaker, increase volume control and check if still sine.
Reduce Volume to comfortable level.
Then increase sig gen level to 2mV (two millivolts) and check at detector for audio distortion due to an AGC or RF overload problem.
Ignore any visible RF distortion (IF freq) at detector for now.

If none of the above shows up a fault, then you may have "other factors" at play.

By the way, distortion due to a basic AM detector fault is rare.
A detector resistor going extremely high can cause it
And the "diode" can be a "wet razor blade and lead pencil point" to function as a detector, ie anything with a front to back resistance of 10 to 1 or more will work, of sorts. Valve detectors can have a few more "woes".


good luck

[Sideband]

Originally Posted by Christoffrad

Thank you all.
Someone in the past has replaced resistor R14 (R13 Trader) in the V2 anode HT circuit and has replaced the 2K2 value shown on the service data with a 18K....Very strange.

Chris

I've come across things like this. Possibly misread an 18K and thought the orange was a red making it 1.8K as a 'near' value to 2.2k.

[Christoffrad]

Sorry to report that the suggestion by Sideband to add the 10M resistor between V3 screen and detector anode did not work (audio dramatically reduced) nor did the change of V2 HT resistor R14 to the correct value do anything to change the distorted sound, or gain.

[Christoffrad]

This is replying to Radio 1950 with results of his proposed tests

The distortion resulting from all the tests described on 21st Oct is negligible.
It just becomes noticeable when the generator modulation is set at max (80%).

The thumbnails show the distortion at the filtered output of the detector and at the loudspeaker when the generator is set at 200uV and the volume increased for 2.5V p-p at the loudspeaker (rather loud!) The first pic is the detector filter output and the second is the signal across the loudspeaker with 80%mod.

It seems to me that a single frequency of modulation does not show the problem. It is particularly noticeable on broadcast signals in the presence of significant bass where the bass seems to bottom the whole waveform as I showed on an earlier posting (20th Oct)

Your thoughts please!

Chris

[Radio1950]

You have done well.

Please do the 2 millivolt test, as described earlier, measure at detector with CRO.

Report back.

If you get bored, reconsider post #32 re speaker.

You will get a result; just hang in.

[Christoffrad]

Done that too. I didn't post any pics because the CRO displays a pure sine wave! About 7V p-p for 2mV with 80% mod at the filtered audio output from the detector.

BTW the FM output is fine

[PJL]

Originally Posted by PJL

You could check the value of R19 and R29. What is it like on VHF? Are weak stations better than strong ones?

Did you check these?

[Radio1950]

Please do a 2 millivolt test, with CRO monitoring speaker voicecoil.

Report back with a visual opinion about distortion.

If you can see distortion on CRO, it is more than 5%.
If you can't, it is less than 5%.
Approximate.

[Radio1950]

Maybe crossed posts.
If me, sorry.

If you get no distortion at the speaker with 2 millivolts test AM signal, try another speaker.as per post 32.



Then, test your sig gen.

Terminate in nominal load impedance, if in doubt just use 50 or 75 ohms, it will not matter.
Set sig gen to 1000 KHz, 75 % of max RF output level, 30 % mod, monitor with CRO with10x probe.

Observe mod on RF signal, report any possible distortion.

Increase sig gen mod to 80 %.

Observe any possible distortion.
Less than 5% is acceptable.
You cannot hear 5% distortion with music.
Caution! Opinion!

Please report back.

Do you have Application Audacity?
.

[Christoffrad]

R19 (Trader) R17 (Philips) measures 244k (220k data sheet) so OK
R29 (Trader) R28 (Philips) measures which I thought I had checked I have now removed one end completely and checked again and it measures O.C. (or at least greater than 20M) and should be a 10M
so I will change that and report back later (have to go out now)

I will also do the 2mV speaker test

[Christoffrad]

Here is the output of the Marconi sig gen at 100mV emf setting into 50ohm load so 50mV pd with 80% mod (CRO with x10 probe) pic1, 30% for pic2

Having removed the 10M resistor from V5 grid (R29 trader / R28 Philips) it was found to be reliably 10M (of course!) but I've replaced it anyway.

I have tried changing Loudspeaker but the distortion on broadcast material is still as bad)


Chris

[Christoffrad]

I may be homing in on this fault ...
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. Sometimes the restoration is such that the audio distortion has disappeared then when it becomes a little distorted or fades again touching the link brings about clear audio again. No mechanical movement is involved.

Could this be a dry joint on the IFT or an AGC line issue? I will make some further voltage measurements around V2 and V3

And no sooner have I written the above than I have to update this because the temporary solution no longer works..... audio has decreased and the distortion is back and touching the link with the probe now achieves nothing.
Voltage on V3 grid is -0.9V V2 grid is -0.5V in this condition

C

[frsimen]

Your photos show that the distortion at the detector isn't much different to the distortion showing on the 80% modulation envelope of the signal generator.

There is noticeably more distortion on the 2.5V p-p waveform measured at the loudspeaker. That tends to point towards there being a problem in the audio amplifier rather than the detector.

A frequent cause of distortion is the anode load resistor of the triode stage (V5 here) going high in value. In this set, this is made up from R29, R30, R31 (references from your posted circuit). They are also associated with the frequency shaping feedback circuit.

Depending on the setting of the tone control, this feedback circuit will provide bass boost. Have you tried adjusting tone control to see if that helps your problem?

Paula

[Christoffrad]

I may be homing in on this fault ...
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. Sometimes the restoration is such that the audio distortion has disappeared then when it becomes a little distorted or fades again touching the link brings about clear audio again. No mechanical movement is involved.

Could this be a dry joint on the IFT or an AGC line issue? I will make some further voltage measurements around V2 and V3

And no sooner have I written the above than I have to update this because the temporary solution no longer works..... audio has decreased and the distortion is back and touching the link with the probe now achieves nothing.
Voltage on V3 grid is -0.9V V2 grid is -0.5V in this condition

values should be V2 grid -1.5V and V3 grid -1.25V according to the data sheet

By the way, I have checked C49 (Philips) (C24 Trader) but perhaps it needs a re-check
C

[Christoffrad]

Thanks Paula
R29 and R30 (Philips) (R31 and R34 Trader) measure 55k and R31 (R30 Trader) measures 101k so no problem there. (measured with electronic resistance bridge after open ending each resistor)
The audio on FM sounds OK so I'd be surprised if the distortion is a audio amp problem.
Also you made an earlier observation about the cramping on V2 anode indicating a bias problem so I'm inclined to think the problem is up-front.

I have however commented that the bass switch makes little difference as it operates, so there may still be issues there.

[frsimen]

The resistors are in good health. Now you're beginning to run out of things to test!

At this point, the most sensible thing to do is to try a new EBF80. The EBF89 has the same pin out, with slightly higher mutual conductance and is generally a little beefier. Its capacitances aren't very different to those of the EBF80, so if you have one you could try that for a short test to see if it makes a difference.

Paula

[Radio1950]

No need for PM messages to anyone - just post here for everyone.

1 - The reason for asking about PC program "Audacity" is that as a last resort, when you have run out of other ideas, you may have to externally modulate your sig gen with "burst audio", and trace it thru.

2 - The reason that sig gen tests with AM are done at 30% mod is that this single tone mod level approximates the average mod level of speech. (And 1000 Hz is the geometric mean of 300 and 3300 Hz!)

3 - The thread is getting longer now, so you will have to forgive us if requests and comments are repeated.

4 - Before you started work on this receiver, did you notice anything that would make you think that someone else had worked on it previously?

[Radio1950]

You mentioned in post #27 and again in post #72, that you observed intermittent operation.
In post #72 you further mentioned that the AM broadcast speech returned to normal without distortion?

Before any more technical tests, you are best to follow Mr Ockham, and go hunting.

I suggest you tune the RX to a station with gentle speech (to avoid distraction), and at gentle volume, and go to the areas of the RX that you previously noticed a possible intermittent.

Use two things, a knitting needle, (this is a vintage forum), and a paint brush with non conductive bristles.
Use a "technician tap" around suspicious areas with your knitting needle, then use brush if required.
Wiggle all tubes, tap those IFTs, thump the chassis gently.
You are of course looking for the return of undistorted audio by listening whilst prodding.


If no joy.
Place receiver in sun for an hour.
No, hmmmm, maybe rather heat the receiver all over with a gentle hair dryer, and repeat.

As another poster has also said, you may have to use a metallic screwdriver around any earthing tags, to short the earth tag to chassis, not trusting any surface coatings.

If nothing found so far, have another, yet another good look around all components with good light and a magnifier for suspicious "stuff".
.

[Radio1950]

In post #74, you noticed the return of normal operation during probing.

Pursue this, (before resuming sig gen type tests).


Temporarily use a 0.01 uF cap to bypass the top (hot) end of C44 (AM IFT) to V3 IF Amp Grid 1, and observe result. It may be intermittent somewhere here.

This is C31 on some circuits.

Follow this line of approach around the AM IF.

You need to positively remove most doubt that the fault is just an intermittent, and to steer your thinking to bypass the FM IF path by using 0.01 uF bypass capacitors if required, as the intermittent could be in one of the FM IF transformer windings.
Reflow solder joints if suspicious.
I use a dab of flux gel these days on suspicious joints when reflowing, but you need the correct temperature solder iron tip for best results.
Sometimes it is best to remove all solder from a joint before inspection and a possible reflow.



Hopefully we don't have to interpret that complicated audio shaping circuitry to fault find; life is too short!


Looks like Philips put their heart and soul into the design of this thing.
Must have been a good receiver with good sound.

Enough for now, coffee calls.

Good luck.

[Radio1950]

Ref my post #78.
Use a plastic knitting needle.