100V line transformer as valve o/p transformer?

[Kat Manton]

Hi,

I've wondered if it's possible to use common 100V (or 70V) line transformers to replace output transformers, at least in smaller valve sets which are some way from 'high fidelity' and aren't producing much power.

One which I have to hand has a secondary intended for an 8-ohm speaker, with primary impedances marked 1.7k, 3.3k, 6.7k and 13k. It'd be possible to obtain other primary impedances with a 4-ohm or 15-ohm speaker connected; so it would appear to be possible to get a reasonably good match for various output valves.

The one thing which concerns me is that clearly these things aren't designed with DC flowing through the primary in mind.

But could that be easily fixed? I had cause to partially dismantle a 'proper' output transformer recently, and noticed that all the 'E' and 'I' laminations were respectively stacked together (which made dismantling easier) rather than interleaved. That results in a 'gapped' core.

If I can dismantle the interleaved 'E' and 'I' laminations from a 100V line transformer, then re-arrange them so I then have all the 'E' laminations together and all the 'I' laminations together; would that be all that's necessary to make the transformer more suitable for a single-ended class-A valve output stage?

I don't claim anything beyond a basic knowledge where transformers are concerned; I can interrelate turns ratios, voltage ratios, impedance ratios and manage to measure/calculate these for unknown transformers. I'm aware that a gapped core is needed if there's DC flowing through any winding (unless it's a push-pull one where the currents cancel.) But that's just the basics. "The devil's in the details", hence I figured I'd ask here; I know we have several resident transformer experts who know more about them than I ever will!

Regards, Kat

[G8HQP Dave]

Adding a gap in the core will reduce the inductance, and so impair the low frequency handling. However, as radio loudspeakers often can't handle much bass anyway you might get away with it.

[kalee20]

I'd say exactly the same. Single-ended transformers are gapped (sometimes just a butt-joint instead of interleaving may be sufficient), and turns increased to give sufficient inductance for the lowest frequency desired.

However, non-interleaved transformers may start out by having more inductance than needed, as the turns have to be at least a certain number to ensure absence of saturation (and consequent distortion) at the lowest AC frequency and amplitude that they'll see. So, after you do as you suggest, you may still be OK.

I would say that almost certainly you won't hear any problems. If you wanted to go to the nth degree, you could make a test winding by threading thin wire around the laminations on top of the existing coil; apply AC to a winding and measure the induced voltage on your test winding; hence calculate the actual turns which exist. Then you bcan measure the core area, assume a saturation flux density of 1.3T, and, also knowing what your DC standing anode current is, calculate the magnetic circuit reluctance to give 0.65T with this standing current. Then you calculate what inductance this will give. Next, pull all the laminations, insert them non-interleaved, connect the primary to an inductance bridge, and add paper shims until the inductance drops to this value. Finally, you can calculate lowest frequency to be handled (assuming that this causes a flux swing from zero to 1.3T).

Or, you could think, life's a bit too short... suck it and see!

[Ed_Dinning]

Hi Kat, Peter has but it precisely; 1 layer of thin paper in the gap, suck it and see!
My DAC 90 output transformers use 1 layer of 2 thou Melinex tape as the gap.

Ed