Output transformer. PYE P32QTG gramophone

[PeterPYE]

Just tested the output transformer and its got an open primary
Found a replacement off ebay, I'll have to wait and see if it will do the job.

What would be the cause of an open winding like this?

[Simon Gittins]

It might just fail anyway, the main thing to guard against is the output valve drawing too much current due to the audio coupling capacitor being electrically leaky.

[Simon Gittins]

I've just had a look at the circuit; there are two versions, both have a 0.005uF capacitor connected from the output valve anode to chassis and this is another possible cause of transformer failure. A 0.0047uF 1000V DC would be a suitable modern replacement.

[PeterPYE]

Originally Posted by Simon Gittins

I've just had a look at the circuit; there are two versions, both have a 0.005uF capacitor connected from the output valve anode to chassis and this is another possible cause of transformer failure. A 0.0047uF 1000V DC would be a suitable modern replacement.

Thanks for that, added it to the order list Mine is the later version btw. So that's not the audio coupling capacitor?

[Simon Gittins]

No, C25 (later version manufacturer's reference number) is the fixed audio correction capacitor that output pentodes need. In many sets, it's connected across the output transformer primary where failure just causes silence but in this case failure would stress the transformer.

The audio coupling capacitor is C22 0.01uF 350V DC. As this is a short superhet radiogram, there isn't the usual significant anode voltage on the other side of it but I would suggest changing both these capacitors anyway to reduce the risk to your new transformer.

[toprepairman]

Quite possibly it may have spent some time in a damp environment and the copper has corroded.

[PeterPYE]

Little update: Just got hold of another P32QTG radiogram. The service sheet says the OT should be 500 ohms on the primary and 18 ohms from tap to finish.
I'm measuring 320 ohms and 11.7 ohms. Would this be ok to use or have I got another dud?

[Sideband]

DC resistance of a transformer winding has no bearing on the AC impedance. The resistance measurements of windings in the manual can only be a rough guide. The only real test is to try it and see if it works (I think it will). The transformer may have been modified during production...for instance the primary might be wound with slightly thicker wire so the resistance will be lower. The overall RATIO of the windings (between primary and secondary) will however be the same.

[PeterPYE]

I've been testing the output transformer from the other set and this is what I've found. Please let me know if I've understood this correctly.

I applied an AC voltage to the primary until I got 1Vac on the secondary and the primary voltage was 63Vac. which is a turns ratio of 63:1.

So the impedance ratio is 63x63= 3969:1

The set's speaker resistance is 3R. x 1.25 to get the impedance = 3.75R

Primary impedance is 3969 x 3.75 = 14884R

I tested it with a signal generator as well (XR2206) which was fun

To calculate what the set requires I did the following:
The output valve is an EL41 with an anode resistance of 7KR.
So 7000/3.75 (speaker impedance) = 1867
So an impedance ratio of 1867:1
Sqr rt of 1867 = 43 So a turns ratio of 43:1

If i'm correct (which I doubt) there's quite a difference between what is actually in the set and the minimum requirements.

[Sideband]

Transformer ratio is calculated by dividing the Ra of the valve in ohms (7000) by the NOMINAL speaker impedance (3 ohms) and then taking the square root of that answer so....7000 ohms divided by three = 2333.3 (recurring). The sq root of that is 48.3 (only one decimal place) so the ideal ratio is 48:1. In practice anything from 50:1 - 60:1 would be fine....it really doesn't have to be spot-on.

I haven't checked the data for the EL41 so I assume your 7K is correct.

You can get an idea of the transformer ratio by applying an AC voltage to the primary as you have done BUT normally, the impedance or ratio would be measured at 1Khz...the standard frequency most of these things are measured at. If you used a 50Hz supply, you will get a different reading. Ideally you should use a signal generator set to 1Khz and a dual beam oscilloscope connected to the primary and secondary and adjust the signal generator output to give 1V p-p on the scope (connected to the secondary). You then measure the input reading to see what voltage input is giving you the 1V output.

All this shows that the transformer has different outputs at different frequencies hence the need for a 'standard' at 1Khz.

[Station X]

You've calculated the ideal turns ratio correctly near enough..

Either there's something wrong with your turns ratio measurement, the transformer is faulty, or it was never a good impedance match in the first place.

[Herald1360]

Try it anyway- it'll likely work well enough that you could only measure rather than hear any difference. If it's no good, you're no worse off than now.

[ms660]

The 7k figure in the valve data for Ra is when Vg2 = Va.

Lawrence.

[PeterPYE]

Thanks Sideband
Ok, here's my cheap little set up.

I'm getting a turns ration of 60:1 not too far off the variac measurement I made earlier. Although as I reduce the gain on the signal generator the P-P ratio seems to drop down to 50:1. I'm guessing that's to do with the OT being less efficient at lower inputs?

Looking at the scope trace, blue is the input signal and yellow is the output.
Why is the output signal not the same.

Sorry if I'm asking daft questions but I'm really enjoying learning about this stuff.

[Herald1360]

The output signal looks the "same" just a smaller amplitude on the screen, likely because the 20mV vs 1V trace sensitivities aren't exactly in the same ratio as the actual signals.

Also, it looks like you've set the trace positions so as to line up the negative peaks of each rather than having both traces' zeroed at the same level, which would show the lower amplitude trace centred "inside" the higher amplitude one.