Can Someone Check My Maths - OT Question

[CraigGa]

Earlier I posted excitedly that the transformers for my Benson Monarch build had arrived. That quickly turned to confusion when I realised that the PT bore no resemblance to the one I specced out although the note with them was exactly as I had specced them.

I emailed the company suspecting that I had just been sent someone else's PT but it was late in the working day and I haven't had a reply yet.

However I decided to test the OT and I think that's wrong too but I'd prefer someone confirm my method before I mention that too.

The OT spec was for 8K/8Ω to suit a pair of 6V6s
Using the method someone recently posted on http://www.radioremembered.org/outimp.htm

I connected an AC wall wart across the primary outers leaving the b+ connection open, connected an 8Ω resistor across the secondary and measured the voltages
Primary 10.35Vac
Secondary 0.41Vac
That gives a voltage/turns ratio of 25.24
Square that gives an impedance ratio of around 637
So an 8Ω load would be reflected as 5096Ω

I would expect a voltage ratio of 32 to give me the required impedance ratio of 1000.

Can anyone see if I've done something wrong here?

I doubt that I will have any trouble with the company and I think that I've just got someone else's transformers but is it one or both of them?

Thanks in advance.
Craig

[CraigGa]

I just had a thought that I'm in the UK so the supply was 50HZ, is that ok or should I be using a higher frequency?

Craig

[martin manning]

I get the same result as you did. Are there multiple secondary taps? If so, are you sure you used the right secondary leads? 50Hz should be fine for this kind of check.

[CraigGa]

Thanks Martin,
No there's just a single 8Ω tap

[blackeye]

connected an 8Ω resistor across the secondary

I don't know if this helps, but whenever I've tried this, I don't use a resistor. I just measure the full unloaded voltage on the secondary. I'm not sure, but maybe that 8-ohm load is dragging down your voltage?

[Phil_S]

connected an 8Ω resistor across the secondary

I don't know if this helps, but whenever I've tried this, I don't use a resistor. I just measure the full unloaded voltage on the secondary. I'm not sure, but maybe that 8-ohm load is dragging down your voltage?

Same here, no resistor. Just hook your meter to the transformer, one probe to each transformer lead. Do not connect the transformer leads. You just want to unloaded output voltage to use in the calculation.

[CraigGa]

Actually I did it without the resistor first but I thought that was the wrong way,
Without a resistor it worked out as

Primary 10.38Vac
Secondary 0.46Vac
That gives a voltage/turns ratio of 22.565
Square that gives an impedance ratio of around 509
So an 8Ω load would be reflected as 4072Ω

Slightly different but still not what I would expect.

Craig

[Phil_S]

The math suggests it is 8KΩ:16Ω, though it could just as easily be 4KΩ:8Ω. If you are concerned about accuracy on the low voltage reading, run the test in the other direction, applying the 10V supply to the secondary. Generally, in my experience, it will be OK to do that. If you think 230V will be too much, then see if you can tap the 6.3V or 5V filament supply on a power transformer for the test.

[sluckey]

I like to inject 1V on the secondary. Keeps the math simple.

[CraigGa]

Thanks for all your replies,
I tried applying the AC to the secondary and got the same results but I'd not considered that it could be a 8KΩ:16Ω transformer.

I did receive a reply to my email so hopefully we'll get to the bottom of it tomorrow.

Craig

[trobbins]

Craig, you will get a slightly more accurate outcome by applying the AC wall wart across half the primary (B+ to one plate terminal), and measuring the voltage across the other half-primary winding (B+ to other plate terminal) - then double that measurement as the better P-P voltage to use in impedance calculations. That is preferred, as the voltage across the 'driven' primary winding section drives excitation current in to that winding which adds some error in to the voltage transfer ratio. The other benefit is that the secondary voltage will be twice what you measured, and so is likely to be more accurately measured by your voltmeter (given that voltage may be a low level for your meter to measure accurately).

[CraigGa]

Craig, you will get a slightly more accurate outcome by applying the AC wall wart across half the primary (B+ to one plate terminal), and measuring the voltage across the other half-primary winding (B+ to other plate terminal) - then double that measurement as the better P-P voltage to use in impedance calculations. That is preferred, as the voltage across the 'driven' primary winding section drives excitation current in to that winding which adds some error in to the voltage transfer ratio. The other benefit is that the secondary voltage will be twice what you measured, and so is likely to be more accurately measured by your voltmeter (given that voltage may be a low level for your meter to measure accurately).

Thanks,
I tried your method and came to an impedance ratio of 490 which still makes it look like a 4KΩ:8Ω or 8KΩ:16Ω transformer.

Craig

[roberto]

If I wer you I'd contact the supplier explaining what happened and asking for a free substitution.

[CraigGa]

If I wer you I'd contact the supplier explaining what happened and asking for a free substitution.

That's already being dealt with Roberto and it doesn't look like it will be a problem. The PT is obviously wrong because it has the wrong number of taps I just wanted to prove that I had someone else's OT too.

Craig

[CraigGa]

My replacement transformers turned up today so I can finally get on with my build.

I tested the new OT using my previous methods and got an impedance ratio of 987:1, spot on for a 8KΩ:8Ω transformer.

Thanks for all your help

Craig