Questions about Output Transformer Ohms and Balance

[HaroldBrooks]

For the last 6 months or so, I've been checking output tube bias by measuring the ohms between the plates and the center tap (with amp off), and then the voltage drop for same. I get current, and then multiple the milliamps times the plate voltage minus the voltage drop.

I got to thinking,

1. why is there resistance in the output transformer ? I thought output transformers were reflective of ohmage depending on winding ratios. Or is it just the impedance of the winding on one side I am measuring ?
2. Why are some output transformers ohm measurements different from left to right, and some not ? How much difference is considered "normal" ? Can a transformer show a large imbalance and still function correctly going forward ?
3. How does this variation in resistance from right to left of an output transformer effect tone and the output tube's operation ? It seems to make a difference, all else held equal, as far as I have seen. (Recently been making a lot of bias and screen voltage changes, as well as trying different tubes and pushing their limits with the imbalance intentionally, or trying to balance things).
4. My thought is that different resistances to the plates on a tube pair should change the response of any given tube pair under a load, and perhaps dynamically, based on input gain and level of transients and sag.

The Zenith amp I am referring to has a rather large difference between both sides of the output transformer, that is 316 ohms vs 454 ohms respectively after playing for a while, and I've checked the transformer after thrashing the amp for about a half hour with horrific gain, and it's not hot, and the temp is fairly even around the windings, and most importantly (to me) the amp sounds fantastic either running clean, or distorted.

Thanks for any insights !

[Enzo]

Got a spool of hook up wire in your shop? Is the inside end of the wire accessible? Measure the resistance of this reel of wire. It won't be zero. A transformer is a coil of wire around an iron core. The wire has resistance. That is what you measured.

Transformers function by transforming between IMPEDANCES, not resistances. You measured resistance, not impedance.

Transformers can be uneven between plate leads due to their construction. There are two ways to wind one with a center tap. One way is to use two pieces of wire and wrap them around the core together. Each piece becomes half a winding. We join one end of each to make the center tap. because the two wires are the same length, they will measure the same resistance. Oh there might be a small difference, but it won't matter to the circuit. The other way is to wind one long piece of wire around the core. When you get to half the turns, you connect a center tap wire to the winding wire, then continue to wind the rest of the wire. That results in one side being tight around the core, while the other half is wrapped around the first. The number of turns will be the same for both sides, which is what counts. But the outer windings will be longer as they have to go around a wider wrap, so the second half being longer wire will have a larger resistance.

If I recall, the little Peavey Classic 30 output transformer had one side at about 90 ohms and the second side at about 120 ohms. Perfectly normal..Uneven resistance will make for uneven idle current, but the turns ratio is not affected, so the thing works for audio just fine. Your 316 versus 450 is a very similar difference in terms of percentage.

This has little effect on tone. Consider the impedance of a transformer might be 6000 ohms, having the two halves differing by a few tens of ohms will not make a difference.

Fir a given tube current yes idle current will be a little different. if it bugs you adjust for the average between them, as in if we get 36ma and 40ma, consider it as 38ma for both. Remember this is just a guitar amp, nothing precision here.

To the tubes, this difference is miniscule.

[Helmholtz]

why is there resistance in the output transformer ? I thought output transformers were reflective of ohmage depending on winding ratios. Or is it just the impedance of the winding on one side I am measuring ?

The OT primary consists of several hundred turns of thin copper wire. The DC resistance is the copper resistance. The DC resistance of transformer coils is an unwanted but unavoidable property. The DCR causes power losses and raises wire temperature.

Smaller OTs typically use more turns and thinner gauge wire, so more DCR.

Why are some output transformers ohm measurements different from left to right, and some not ? How much difference is considered "normal" ? Can a transformer show a large imbalance and still function correctly going forward ?

Different DCRs of the 2 halves of the primary are a consequence of coil geometry. Typically one of the halves is the innermost winding followed by the secondary winding and at last the other half of the primary. As outer turns have a larger circumference, the outer primary half needs more wire length for the same number of turns - so its DCR is higher.
For proper operation it is essential that both primary halves have the same number of turns, meaning equal voltage amplitudes. To test for equal number of turns disconnect and separate all OT wires, feed 6.3VAC heater voltage to the 8 Ohm secondary tap and measure left and right primary voltages (couple of hundred volts!) for symmetry.

I have measured DCR differences up to around 30%. Larger OTs show less difference and much lower DCR (meaning less losses). A 100W Marshall OT measures 15.0/16.7 Ohm.

A difference of 100 Ohm will cause the plate voltages to differ by around 1%, so no effect on idle currents.
With AC signals the DCRs adds to the much higher reflected load impedance of 1k to 2k per tube, so the total load difference is small.
In other words, no significant signal imbalance caused by DCR difference.

[überfuzz]

Interesting points... I'm a tad curious. Is this for a Hifi-amp?

Are you trying to balance the two sides of the signal, or signals? You might get a good, lets call it, overall result by tinkering with the inverter's anode resistors. What kind of inverter are you using on this amp?

[HaroldBrooks]

Interesting points... I'm a tad curious. Is this for a Hifi-amp?

Are you trying to balance the two sides of the signal, or signals? You might get a good, lets call it, overall result by tinkering with the inverter's anode resistors. What kind of inverter are you using on this amp?

It is a hi-fi amp (Zenith), but I spotted this amp as a version that's really close to a guitar amp, I.E. 12ax7 PI, 6V6 push pull, and I've altered some of the values to improve the distorted tone even further, with the help of Helmholtz and others.The amp is now in my top 5 tone and response wise, who would have thought ! Particularly when it's matched to my vintage 12" Jensen cabinet.

It's basically a raw chassis at this point, and it's super easy to work on so I've been at it nearly every night over the past few weeks running some loose "Experiments", and it's yielded some great insights for me, along with the help I get from everyone here. With the exception of my Marshall, this is the only other amp I've had that uses negative feedback and more importantly a "Presence" circuit that ads even more to the high gain clarity. Here's the simple schematic :

I don't know what type of PI arrangement it is, and please feel free to comment on that.

I reduced the size of the coupling caps, and changed other resistors to play with screen voltage, frequencies for the global feedback, etc... Great returns on this one !!!

I also tried several sets of 6v6 tubes of various vintage and strength as pairs, and in combinations while measuring plate dissipation, voltage, and screen voltages. I am mostly concerned with the amp's tone and response at fairly high gain settings (think 1970's to 80's rock), and it shines in that arena.

On this amp in particular, I've noticed a clearer more defined distorted tone when there is a mismatch of power tubes. One now runs with 12.6 watts dissipation or so, the other 10.3 watts with plates around 375vdc and screens up at 360vdc. That yields the best sound IMHO so far. Matching output tubes deteriorated the clarity, and made the sound more homogeneous, and closer to a "New" amp that contains no such folly of numbers, or "Gasp" a modeling amp. Another thing is high screen grid voltages, in conjunction with the power tube imbalance. On this amp, it sounds much better. So much so, that I had to live with using a bigger screen resistor now (2.7k) to stop the glowing screens cold on just the one tube. There is a trandeoff for this, and I can feel it in the response, as it's all become a bit more compressed, but the wild and "Hairy" quality of running the higher voltages, at least on this amp, is not something I would let go of, regardless if I gave up some dynamics because of the large screen resistor.

I posted these OT questions because I didn't understand the wire wrapping geometry on the impedance of the primary sides, but now I do thanks to Enzo and Helmholtz, and I see this is a normal state on some OTs and nothing to be concerned about in the least.

I also tried using a larger OT I have as a backup as a test, after disconnecting the original, and that made the sound worse !!! I only judge my results by me ear, provided I haven't screwed anything up too much and made a grenade of the whole thing. So there I go, the original OT sounds better, and this amp likes high voltages and a power tube imbalance, and finally a modern tube for the PI position. I had imagined an older RCA might be better, but again, it's not to my ear !

Nice thing about having lived a while, you get to know what you like right away.

Thanks for any help regarding the PI arrangement and the amp in general !

[Justin Thomas]

I'm just asking to play alomg with this phase inverter question, but, is that a form of Paraphase? I've never seen one drawn like that... Just a guess. Don't pay attention to me.

Justin

[Helmholtz]

Yes, its's a self-balancing paraphase PI.

[trobbins]

DC bias imbalance in the OT primary half windings can significantly increase distortion at low frequencies in hi-fi amps. At some point of imbalance that may well become noticeable and may partly contribute to your enjoyment factor for guitar use. That experience sort of aligns with other other vintage guitar amps that purposefully used say a 60W rated OT in a 100W guitar amp, where the low frequency end distortion becomes noticeable at some point.

[überfuzz]

...I don't know what type of PI arrangement it is, and please feel free to comment on that...

Seems to be a paraphase inverter. I've never tinker with one of those...

Are the out of phase signals unbalanced? If I were trying to balance this kind of inverter I'd start with the anode resistor, or maybe put a pot on the signal path to ground to try and dial in the balance..