Using a "power transformer" as an OT

[PeanutNore]

If the turns ratio is correct, is there any reason that a transformer intended as a power transformer will not work as an output transformer?

I am planning to build a low power head using a 12AU7 in self-split push pull. A firefly amp, basically, but I might mess around some with the preamp. I already have a transformer that has a 36.6:1 turns ratio (230v CT to 6.3v) which means that for an 8 ohm speaker hooked up to the secondary the primary impedance should be just a hair over 10k ohms. The transformer is rated for 2.5VA, and the insulation is rated for much higher voltages than the B+, so I can't think of any reason why it wouldn't work. It seems like it would be fine even with a 12BH7.

This is the transformer: FS12-200-C2 Triad Magnetics | Mouser

[R.G.]

It will work, in that it will produce sound and the nominal loading. It will have poor treble (and probably bass) response, and emphasize any crossover distortion in the drive to it.

This is because the windings in a power transformer are not as tightly coupled magnetically as the windings in a good OT. Good OT winding practice says that the primaries and secondaries should be split into sections and interleaved to minimize leakage inductance and interwinding capacitance effects. Power transformers work at a single frequency and don't need that expensive and delicate treatment.

[PeanutNore]

Cool, so what I'm hearing is I should go ahead an just order a purpose built output transformer, but if I finish building the rest of the amp before it shows up I could probably test it out with that transformer but it won't sound very good.

I've been having a hard time finding a very low wattage push pull OT with the right specs that isn't more expensive than the rest of the parts in the amp combined, though. So far this edcor one seems to be my best bet: https://www.edcorusa.com/xpp10-8-10k

[Leo_Gnardo]

I've tried the PT as OT experiment, it sounded awful and hugely accentuated crossover distortion. Edcor has a reputation as an excellent source - I'm sure you'll be happy with the results. Be aware they build to order, don't keep transformers stacked up in a warehouse, so it may take a couple weeks. From what I heard, it's a small family company with very few employees. The sooner you get your order in, the less time you'll have to wait to complete your amp.

[PeanutNore]

I actually found a replacement OT for a Blackstar HT-5 from an authorized Korg/Vox/Blackstar parts distributor for only $16. It's a 22.5k primary with 8 and 16 ohm secondaries for a 12BH7, so it should work just fine with a 12AU7 with headroom to spare. It probably doesn't come with any documentation or explanation of what the leads are, but from the HT-5 schematic I've gathered that the side with red, black, and brown is primary; and red, black, and green is secondary. I can confirm what's what with a multimeter.

[PeanutNore]

I just built an EF86 pedal that sounds awesome plugged straight into the FX return of an amp, so I'm going to build the same circuit in front of a self split 12AU7 push pull output in a little box to sit on top of a large cabinet and look ridiculous. It should be able to do some really nice clean tones as well as drive the output stage pretty hard into clipping.

[trobbins]

There have been a number of diy amps built around cheap power transformers as OTs on the Aussie Guitar Gear Heads - Index forum. They have shown quite good performance. I would suggest using your initial target PT to get the amp going and sort through all the performance limitations that arise from self-splits - and even try and identify if there are any higher-frequency trade-offs just attributable to the OT. If you get that far, then you may still be keen enough to try a normal OT.

An intermediate stage for the type of OT used, is a cheap line speaker transformer - they are quite common, and many can be used as PP OTs once you work through the turns ratios - although you may end up with a 20-30W OT to get a 10k PP winding match. The smaller 5W line transformer I've used (MM1900) provides a 20kPP match to 8 ohm.

[PeanutNore]

That MM1900 transformer is an interesting option. 20k is pretty close to ideal for a 12AU7 output. I'm not sure if I understand how it would be hooked up to the B+ for push pull though. I assume that one plate connects to the "common" tap, the other to the 20k tap, and the B+ to the 10k tap?

[R.G.]

It's worth noting that power transformers, being insensitive to the vagaries of changing frequencies, are wound in different ways. The old classic style from the 50s and 60s was to wind primary over secondary, or vice-versa on an EI core. This is a modest amount of leakage inductance, which is the most likely culprit in the amplification of crossover distortion notches and high frequency loss. Many PTs are wound on side-by-side sections of a plastic bobbin today. This is much worse for leakage inductance than the old primary-over-secondary. There will be a significant difference in the leakage inductance depending on how the primary and secondary are wound.

As a side note, if you use a toroidal power transformer, the leakage inductance is dramatically reduced, as the primary and secondaries are very tightly coupled, being spread out in a single layer. However, the capacitance between the two is maximized, and this will cause other high frequency effects.

[J M Fahey]

Don't know the original suggested wiring but in any case, the CT must sit *between* both plates, not "to a side".

[trobbins]

The 0.5W and 0 terminals are the PP end terminals, and the 2W terminal is the PP CT terminal. The DCR for each PP half winding is not too bad at 113 and 102 ohms in the unit I measured.

[trobbins]

It's likely that the inherent self-split distortion will get ugly way before the user pushes the output valves in to noticeable class B region or anything close to valve saturation.

The use of power transformers for OTs have been in the domain of diy scrap-yard or recycling mentality and for guitar use, so typically have a narrow bandwidth.

[PeanutNore]

This is my first amp build, I've been working on tube pedal designs with high voltage supplies and I have the parts sitting around to build a "firefly" amp but I must admit I'm not familiar with how the self-split push pull output stage works. What is the inherent distortion like? I was planning on making a very clean preamp with either an EF86 or a 12AX7 cascode and getting any distortion from pushing the output triodes, but I might re-think that if it's just going to sound bad.

I guess I can build it with a cascode preamp, and if it sounds bad rewire it for 2 class A triode stages. Alternatively I could wire it up with a single 12AX7 gain stage and a cathodyne PI, but then I'll have a lot less gain on tap.