Question about unequal OT primaries and push-pull

[Matt T.]

Many times when I measure a brand new (or known good) OT primaries, they don't measure exactly the same...i.e. the dc resistance of one primary 'end' to the center tap will be slightly more/less than the other.

Usually it seems that the out-of-phase signal of a PI is stronger than the in-phase signal (relative to the PI input signal)...or at least the power tube driven by the out-of-phase signal is working harder (verified by the highly scientific method of watching the tubes 'light up' as I play it hard ).

So...which is better...driving the 'higher resistance' side of the OT with the stronger PI signal or driving the 'lower resistance' side with the stronger PI signal. This means taking into account the NFB (if any) and swapping the PI/OT primary leads to get that part right.

[stokes]

FWIW I have never measured the dc resistance of any guitar amps OT primary and found them to match from one side to the other.As for one side of the PI's signal being stronger than the other,unless there is a huge difference,I wouldnt worry about it.You can order a PI tube with "matched" triodes,but in my experience,it wont make a big difference,since all the components are usually rated for a tolerance of +/- 10% anyway.Even if you were to use 1% tolerance components and a "matched" PI the whole thing is ending up going thru your OT wose primary doesnt match from one side to the other.I feel that with a guitar amp as opposed to a Hi-Fi amp that little non-linearity is needed to give the guitar amp its added warmth.

[MWJB]

As stokes says, I wouldn't worry about it as long as you don't have a huge discrepancy, say more than 15% (and of course, it sounds OK). One side of the primary will measure more than the other because it will have more wire in it, the same number of turns but because one side is wound first, then the second wound around it, a different amount of wire. Don't forget that your OT is not particularly concerned with this (DCR), it is only bothered about AC impedance. Ideally you would inject a small AC voltage (0.5VAC) on the secondary, then measure the voltage that develops at each end of the primary, to the CT, to identify a mismatch.

[Matt T.]

Thanks for the responses. I'm not worried about it...just curious and thought maybe I'd learn something. I realize the PI is usually unbalanced...and it appears that OTs are imbalanced a little bit too. Seems reasonable that having one imbalance reinforce/cancel out the other's imbalance might make a small difference in tone and I was just curious about it (without having to rewire an OT in one of my amps).

I was interested in the 'technical' reasons for choosing one way over the other...but sometimes what is 'technically correct' doesn't sound as good as 'technically incorrect' in guitar amps.

[Bruce / Mission Amps]

Matt... do not be confused by DC resistance measurements here.
Especially those made by an inexpensive DVMM!
The OT is a transformer and works in the realm of "impedances" not DC resistance.
There needs to be more wire on one side of the center tap in a PP OT in order to keep the impedance a better balance.

[Matt T.]

Thanks Bruce. I'm aware of the impedance vs. resistance thing...but not much past that. I was thinking; more resistance = more wire and more wire = different impedance. So while I said 'resistance'; what I was getting at was that the impedance appears to be slightly mismatched. Is it? If so...then my original question still holds...does it make a difference?

[Bruce / Mission Amps]

Well, I'm not a magnetics engineer so I can only speak in general terms.
However, using the term impedance (called 'Z' or Zed by some of us old guys) is the kind of like asking, what is the effect on an AC signal when the total amount of DC resistance, inductance and capacitance is all added up in a big lump ...(in the old days we actually use to call that lumped LCR).
Those individually are called reactance with respect to AC.
Capacitve reactance, inductive reactance and then of course DC resistance.The resistance in lumped LCR can change the Q (old term for a form of quality of the circuit) and it also effects the bandwidth.

AC resistance a little more complicated because you mostly want to call it reactance and current flows in both directions with AC, plus, the peak AC voltage is different then the RMS voltage, with RMS being the term to describe the amplitude of AC over a cycle portion where energy would be the same if it were really DC... I know ... confusing.
The Lumped LCR total also has to take into consideration the series and parallel effects of those resistances, inductive and capacitive reactances too.

So yes the DC resistance of the wire does have an minor effect but it is small compared to the other factors.
I would imagine in a perfect world the DC resistance would be zero so it has zero effect. The best transformers made would probably like to have the lowest DC resistance too.

In mildly distorted audio amplifiers, such as the cleanest guitar amp made, a slight mismatch of Zed across the OT primary, with respect to the center tap, is not very important. ha ha.
Probably less so for DC resistance if the Zed is close.

Unless you were using an O'scope to view and record absolute waveform, the differences with a slight mismatch are small enough that it is no big deal for the work we do.

[Matt T.]

Thanks Bruce. I figured it would be essentially 'negligible' but wanted to be sure. Time for me to get a 'scope...new thread.