Quad II output transformer

[Johan Potgieter]

In all the references re the Quad II output transformer, I could not find the primary inductance and leakage reactance listed - also not on ads for commercial versions.

Has anybody ever measured this and can you give me values? (For the primary the whole primary, i.e. inclusive of the cathode windings, needs to be measured.)

Thanks!

 

[jazbo8]

Here is a quote from yourself:

"I cannot add much more, except to perhaps give figures I measurted on the OPTs of 2 Quads in my possession (serial numbers 53125 and 49000 - would you believe it).

Inductance of full primary at 5V 50Hz: 39 Hy and 34 Hy
Leakage reactance at 1 KHz, full primary to secondary shorted: 11 mH and 11,2 mH
Anode primary (XZ) to shorted catode winding (UW shorted): 113mH and 106mH

Just as an exercise I calculated the anode/cathode winding leakage with some guesswork as to dimensions (but no back-engineering!) and got 118 mH, so everything seems to fit."

There are also other interesting data in the thread from cerrem.

 

[Johan Potgieter]

Jazbo!





Members: Permission to laugh!

It would appear that not only is my memory/filing system wanting, but also my search skills ....

Thanks friend! Thus duly mortified, I would still ask whether someone else perhaps also measured these characteristics (primary inductance and leakage reactance) and can confirm my findings. There appears to be some spread in values to what I have just found, but possibly not that unexpected for transformers. [The primary inductance (39H) is rather low for a p.p. KT66 application.]

 

[jazbo8]

Menno is your man.

122H @ 50Hz (180Vrms)

 

[Johan Potgieter]

Thanks again, Jazbo8,

Interesting reading; dusted off my relevant maths again!

I still have a Quad II with me, will go check. Not to belabour, but my measurements are with 'lowest' primary exitation, as in common practice with 5Vrms. That gives a lower reading than max. at 180Vrms. Mr Van der Veen takes the primary as Z-X, not clear that he also included the cathode winding UYW, which is also part of the primary. Otherwise enlightening reading.

So there, thanks for your responses!

 

[jazbo8]

You are welcome, yup, the drive voltage definitely changes the inductance reading, I guess he was interested in the maximum value. I do not think it included the cathode winding though. Are you trying to re-build some?

 

[Johan Potgieter]

Not right now - but these amplifiers keep coming my way for reconditioning. Just trying to get a complete set of notes in case, so that one can easily measure whether or not something is suspect.

(and by my own admission, myself the major reason for said notes being incomplete!)

 

[jazbo8]

According to Patrick Turner, the original Quad-II OPT's are pretty low quality, and he prefers to replacing them with modern, improved versions. But I am not sure if the "collectors" these days would agree with using non-original parts.

 

[cerrem]

A few things to keep in mind....
The primary inductance is not a constant for this type of transformer...
The inductance increases as the applied signal level increases...The inductance rises then peaks and then starts rolling off ... This is not the saturation of the core at this peak due to using E-I laminations as in the QUAD II OT.....
The leakage is a fixed number due to the winding geometry...
The capacitance will vary slightly based on the applied signal level....this will cause the transformer resonance to shift its 2nd order POLE lower as signal level increase...
When using a Network Analyzer to measure amplifier frequency response, it should be done at various power levels... The BODE plots will show this...

 

[trobbins]

The loading seen in the roll off region is also a large uncertainty. Perhaps best to dictate the need for a zobel to ensure nominal loading.

 

[Johan Potgieter]

Thanks all,

As Cerrem said. For comparison purposes many OPTs state primary inductance at 5V 50 Hz (or perhaps 60 Hz where applicable). Not to go into the whole theory again; I use this value to start, but also other exitations to check NFB stability. Cerrem gave a lot of detail in the earlier (quoted) thread.

I found 35 Hy at 5V 50Hz plus some 30 mHy at 1 KHz. Yes, somewhat mediocre, rather high leakage for all of 6 secondaries, but satisfied Sir Peter's demands. Further detail for rewind: Original laminations were cut to a different standard than lossless, more winding window area. It can be such a mess to retrieve the original core that in the past I decided to go for new, at approximate similar specs. I do not feel that will be cheating collectors when original 'originals' are not available any more.

Anyway, all said before. My thanks for members' reaction. Still a personal laugh in the mirror for having been unable to relocate my own posts!

P.S: Nominal loading is achieved by the low values of R10, R11.

 

[Johan Potgieter]

Follow:

Oops again - what is it with me and Quad OPTs

Just noticed I previously found a leakage reactance of 11 mH. Wiill have to check the present unit again.

 

[trobbins]

R10-11 only provides 570 ohm loading to 50% feedback portion of output turns - so there still appears to be a bit of range for stability margin to move around with resonance frequency and Q due to various speaker loading scenarios.

 

[cerrem]

Measuring primary inductance down in the mud like at 5V is not very useful and very inconsistent...
How are you measuring your leakage ??? LCR Bridge perhaps ??

 

[Johan Potgieter]

Sorry - did not follow up immediately.

The 5V exitation was arbitrarily chosen as representing the lowest inductance worth mentioning, for frequency response and NFB stability purposes. It seems to have been adopted since the days of the Williamson amplifier back in about 1946.

The leakage reactance was measured with a digital LCR meter, using the bridge principle I imagine. The frequency was roughly 1,2 KHz. Whole primary in series to whole secondary in series, shorted.

I have a definite one-off strange transformer on my hands, showing no signs of having been rewound. (The d.c. resistances of the windings are the same as with other transformers, so the problem remains as to where a different leakage reactance figure could be coming from. The value of some 27mH gives a low response peak at 41 KHz with secondary under load; the matter is somewhat complicated in that NFB is taken off a tap on the secondary and not the whole secondary. This can be compensated for but leaves a slightly uneasy feeling. Nevertheless.)