Choosing an Output Transformer

[Bob M.]

Hi,
I'm building an amp for myself using 2x 6V6 power tubes and I have to choose an output (audio) transformer. I have three candidates in my transformer box: an Allen TO-26 (made for 2x 6V6), an older (cloth wire) original, but used, Fender A-041318 and an newer Fender type, bust also used (with plastic covered wire) NSC 041318 (this could be a third party Fender copy, not sure). All three of these measure really differently from one another when measuring ohmage between the leads. I'm wondering if there is a bench test of some sort I can perform on these, besides hooking them up to the amp and hearing which sounds the best, so I might pick the most likely candidate to sound best. I know I can choose by reputation or by which sounds best to my ears but I'm wondering if there is a scientific way of determining which has the best chances of sounding great before the trial and error listening test.

Thanks,
Bob M.

[olddawg]

You can swap them in with wire nuts unmounted once the amp,is operational. That’s what I would do. Imhe the beefier ones tend to have more bass response.

[TomCarlos]

You didn't mention what type of speaker(s) you intend to use and the impedance. Do you know for certain that the transformers you have are a match to the speaker impedance you intend to use?

I believe the Allen Transformer has both 8 and 16 ohm taps.

The Fender A-041318 (and equivalents) are 8 ohm but you might find one that is also 4 ohm.

The NSC 041318 is also an 8 ohm.

And all of these are rated for 25 watts.

[Chuck H]

This^^^^^^^^^^^^^^^^^^^^^

I don't think I've ever done a build that didn't obscure a couple of the OT mounting screws. In that case it's hard to have things "operational" before choosing, UNLESS, the mounting holes are the same or close enough that you can slot a couple of the screw holes (or whatever else might work) to make the chassis universal for any of the three. Then you could try each before committing. That's what I would do. And I'll tell you why. If I committed to one without actually hearing the other two I would always wonder. I don't think I could live with that

[olddawg]

Well.. you can mount one the transformer using one screw for test.. or even a spot of Velcro. I’ve done both. As long as you have enough flywire you can put the OT anywhere. Many old amps had them mounted on the speaker basket.

[Helmholtz]

If the amp has global NFB I recommend to keep OT wires short and twist them to avoid oscillation. Also the "correct" secondary wire must be grounded.
I would also ground the OT frame/core for safety reasons - mounting with at least one screw would achieve that.

[Bob M.]

Thanks for your opinions. I'm looking for a scientific test to measure these transformers. I know I can hook them up, one way or another, and do listening tests (yes, I know what I like) but I'm trying to save myself the hassle of doing that and find a less subjective method of evaluation. We're dealing with the largest voltages in the amp here - it's not like subbing in another value of tone stack cap for audition - so some amount of care must be taken. In the past, and with other builds, I have mounted each candidate, played the amp for a while, then for a few gigs and then tried the next one. Eventually, I'd arrive at the parts that worked best with my amp, circuit, speaker and cabinet but that was over time. This time, I'm trying to shorten this trial listening period by using more objective scientific methods, if they exist.

All three of these output transformers were designed to work with 2x 6V6 power tubes with a 8 ohm secondary output (Yes, the Allen has both 8 ohm and 16 ohm taps). The cab I have chosen for this project has a cutout for a 12" speaker which will be 8 ohms and about 50 watt handling power. I've built this very amp before for a friend of mine so it's somewhat tested and pre-troubleshot. What spurred me to ask about a test is the ohmage on the primaries of all three of these transformers vary quite a lot from one to another. I've had these transformers in other amps in the past (but not the Allen TO-26, as it's brand new, but I've used the TO-26 in other builds (successfully) in the past) and I know them all to be working and to sound like an output transformer from a Deluxe Reverb amp. What I think I'm after here is some finer nuance of what the numbers can tell me. If not, I'll put my ears to work.

Bob M.

[Mick Bailey]

The actual parameters of an output transformer can vary considerably and the most basic measurement of DC resistance can be very misleading. Actual impedance values also vary with different manufacturers and may appear to be sub-optimal compared with the ideal value given in the tube spec sheet. In the end a scientific analysis may not give you the best sounding choice and you would need to establish a multi-dimensional model for your perfect transformer and factor in many variables. In the end you'd still end up with a subjective comparison and be A-B-ing because you'd always wonder what the others sounded like. At least I would.

[Helmholtz]

OT performance depends on a number of parameters. The most influential ones being:

- Primary inductance, determining bass response
- Leakage inductance(s), mainly determining high frequeny response
- Primary self-capacitance, determining OT self-resonance(s) together with leakage inductance(s)
- Primary impedance, influencing output tube operation (loadline), max output power and power amp distortion characteristics
- Primary DCR, only affecting OT losses.

Depending on the amp (and speaker) a smaller bandwidth OT is sometimes the better choice as it filters out undesirable frequencies.

[g1]

I don't think resistance numbers will tell you anything. Some marshall primary halves can be around 50 ohms where are similar spec'd fender might measure around 200 ohms.
AC voltage measurements will tell you turns ratio so you can figure out the primary Z, but you probably know that from the specs anyway.

[Helmholtz]

I don't think resistance numbers will tell you anything. Some marshall primary halves can be around 50 ohms where are similar spec'd fender might measure around 200 ohms.

By a very rough estimation a 100R higher DCR per primary half might reduce available output power by 0.5W in a 20W amp.

[Chuck H]

By a very rough estimation a 100R higher DCR per primary half might reduce available output power by 0.5W in a 20W amp.

Bet you wouldn't notice

But what does that extra resistance due to either more windings or smaller wire do to the capacitance? Inductance? I think there must be audible differences. Even if they don't detriment spec or function all that much. That's why you try all three

[Helmholtz]

As said above, primary DCR only affects the losses.

Inductance depends on turns number squared, coil loop area and most important the core, which increases the air coil inductance by a factor of several hundred. So core material and construction (effective air gap) strongly influence inductance and saturation.

Core size has no direct influence on inductance.

DCR depends on wire gauge and total length of wire. Smaller sized OTs will have more DCR as the smaller bobbin requires smaller wire and more turns to achieve required inductance.

More primary turns increase saturation headroom. As does a larger iron cross section.


Measuring transformer parameters requires special measuring equipment and set-ups - and takes time. Especially inductance varies with test voltage and frequency.

I measured the relevant parameters of around 40 OTs. But I am not able to predict which set of parameters will sound best in an amp.

So trying the candidates is by far the most efficient procedure.

[Bob M.]

Well, the second half of the posts on this thread was quite interesting and the kind of discussion I was after.
Thanks to all who participated.
Bob M.

[HaroldBrooks]

As said above, primary DCR only affects the losses.

Inductance depends on turns number squared, coil loop area and most important the core, which increases the air coil inductance by a factor of several hundred. So core material and construction (effective air gap) strongly influence inductance and saturation.

Core size has no direct influence on inductance.

DCR depends on wire gauge and total length of wire. Smaller sized OTs will have more DCR as the smaller bobbin requires smaller wire and more turns to achieve required inductance.

More primary turns increase saturation headroom. As does a larger iron cross section.


Measuring transformer parameters requires special measuring equipment and set-ups - and takes time. Especially inductance varies with test voltage and frequency.

I measured the relevant parameters of around 40 OTs. But I am not able to predict which set of parameters will sound best in an amp.

So trying the candidates is by far the most efficient procedure.

Good advice Helmholtz and everyone else who chimed in to just try the different OTs to see what works the best.

As long as the impedance mismatch is not too high, the transformer not really far too small, the rest I feel is in the subjective ear testing. I too would like to have an easy way of determining what transformer will do what sonically, but now that I've tested several OTs (both new and old) on a few different amps, I believe you have to listen and play the amp with the said OT installed, that's the only way you will know for sure.

I've had my "Perfect" sound in one amp down to a science in my practice room, and then enter a dive bar with a poor layout and little acoustic absorption, only to find one of my lesser amps (the kind I ignore generally) sounds much better !

It's Math, Science, and Art with guitar amps and musical instrument audio. All three are inseparable, particularly with musical instrument amps, as you are creating a sound, not reproducing it faithfully.