Inexpensive AF Output Transformers

[Bazz4CQJ]

Has anyone else picked up on the Bogden T725 transformer which sells in the US for ~$6? It is intended for use in inter-office PA systems, but looks like it would work as an output transformer in low power (<4W) valve circuits.

There is a fairly lengthy discussion on them here; http://makearadio.com/misc-stuff/t-725.php

They show up on ebay but usually attached to a speaker. I contacted the one seller who supplies them separately, but he won't ship overseas.

Anyone got friends in the US ?

[cmjones01]

Looking at that page, it's just a matching transformer for 70V or 100V line speaker systems. That means its core is unlikely to have an air gap, so it won't be very suitable for use as a single-ended valve output transformer - the standing DC current may bring the core close to saturation, so you're likely to end up with poor audio quality. It's no different to using a little mains transformer as an output transformer - the same compromises apply.

Chris

[Nickthedentist]

...and you can buy 100V line transformers here very easily too.

There's an RS audio OP transformer available "cheaply" (£13.21 + VAT) which is fine for cheap 'n' cheerful sets: http://www.vintage-radio.com/images/...ransformer.gif
http://uk.rs-online.com/mobile/p/products/2106475/


Nick

[Bazz4CQJ]

In terms of loss of quality, how specific is it possible to be about how saturation affects output; I'm guessing it will limit low frequency response? In terms of comparison with mains transformers, don't the numerous tappings on the Bogen enable better matching with particular valves?

B

[cmjones01]

When a transformer saturates, its windings stop being inductors and start being resistors. That means that it doesn't 'transform' as well. The result in the audio would be horrible distortion, especially at low frequencies. I suspect that the output valve anode dissipation would go up because, as the audio signal on the grid increases the anode current, the voltage across the transformer primary wouldn't increase much, so all the power would be dissipated in the output valve. That's not good for it.

I spend some of my professional time designing switch-mode power supplies, which have some things in common with valve output stages, surprisingly. If I let any of the inductors or transformers saturate in those, the result is usually smoke, or at least things getting hot enough to desolder themselves from the PCB!

Chris

[kalee20]

Originally Posted by cmjones01

I suspect that the output valve anode dissipation would go up because, as the audio signal on the grid increases the anode current, the voltage across the transformer primary wouldn't increase much, so all the power would be dissipated in the output valve.

Actually the power dissipation would stay the same, that is, the same as the no-signal condition.

For a Class A output stage, maximum dissipation occurs at zero signal. When you start to drive the stage, you get some output power, the dissipation in the output valve decreases, and the total power drawn from the supply stays the same.

If the load is short-circuit, then the anode voltage stays constant, the anode current varies above and below the mean. So you get zero output, and the valve continues to dissipate the same power.

It is not good, but neither is it any worse than leaving it on with the volume at minimum.

[cmjones01]

Originally Posted by kalee20

Actually the power dissipation would stay the same, that is, the same as the no-signal condition.

Good point - I didn't think it through, considering what would happen on the other half-cycle! So it won't work well but nothing will actually melt or smoke...

Chris

[Radio Wrangler]

Saturation caused by DC bias is going to flat-top the output waveform on one polarity of the cycle only. It sounds awful.

Push-pull operation would circumvent the need for a gapped transformer and there are double valves that the phoolish fraternity haven't latched onto yet... like the little 6J6 or the bigger QQV0x-xx tetrodes, though the common screen and cathode could cramp the designer's style. Still you could have two bottles with both halves strapped in parallel and the screen strapped to the anode. Pair of QQV03-10 maybe?

David

You can find 'push pull' tappings on most line transformers, a bit low on impeadance though. I did make an amp with an ECC81 (or was it an '82?) using an RS 100V line transformer once. It worked up to a whole watt I guess.

[kalee20]

Impending saturation in output transformers is bad news, but it does not sound quaite as awful as you might think. Most cheap transformers tend to go into saturation in a rather soggy manner thus you get bass reduction, mild distortion, but don't lose intelligibility.

Toroidal transformers, or transformers made with more exotic core materials, go into saturation rather more abruptly. But such items are used where best performance in a small space or weight is needed.

[JHGibson]

There is some good information on uses of the T-725 on

www.makearadio.com/misc-stuff/t-725.php

John.

[Bazz4CQJ]

My curiosity about this transformer centred on the idea that it is only rated at 3.5W, but for a table-top radio it could probably be run well within that limit and still produce enough output. At those power levels, we will probably have to miss out on the excitement of glowing anodes and melting solder . Its very small size might make it attractive for a portable, perhaps using D valves?

The RS transformer is ~4 times the price and I note that there is absolutely no mention of that having an air gap. If I can obtain one, I'd be happy to pay the ~£3 to find out what it can/cannot do. If anyone else is interested, please PM me.

[JHGibson]

If core saturation becomes a problem, you could run some DC into the other half of the primary using a push-pull configuration but with a single output valve.
John.

[Bazz4CQJ]

Nice idea! As you point out,the makearadio site is well worth a visit.

B

[Radio Wrangler]

"If core saturation becomes a problem, you could run some DC into the other half of the primary using a push-pull configuration but with a single output valve.
John."

I'm afraid this is one of those one step forwards, one step back situations.

Whatever applies the DC to the other winding has to allow it to have the AC signal on it. If you use a choke, you're back to needing gapped magnetics. If you use another valve, you nedd all the bits that would have made a push-pull amp anyway.

Sorry,

David

[JHGibson]

If you just use a resistor to earth to supply the current you will have a simple solution.
However you will loose half your audio power in the resistor but there should be enough left over.
John.

[JHGibson]

The technique of running a DC current through the other half of a output transformer's tapped primary was often used by Philips. One example of which was model B3G63A (trader sheet 1374).

The purpose of the current was to reduce the static magnetic field in the core to nearly zero so that the primary inductance could be made large as there was now no need for an air gap in the laminations. This increased inductance extended the bass response.

In this particular radio the 1.2K resistor towards earth is connected to a 50 UF electrolytic, and this junction becomes the HT+ to the rest of the radio.

The trade off was that for some loss of audio power in the 1.2K, you obtained a better low frequency response.

John.

[Herald1360]

Is that a different reason for doing this than hum cancellation, then? Would be a typically Philips sort of thing to do, if so......

[kalee20]

Well not really... the hum neutralisation was nothing to do with DC cancellation. In fact the DC flowing through the auxiliary section of winding (output valve screen-grid + rest of radio's load) was only a fraction of the DC flowing down the majority of the winding (ouput valve anode current). Fewer mA and fewer turns does not lead to much DC cancellation.

You do get some audio power loss, in the 1.2k resistor, but it's not very much.

The idea of the idea is trhat the AC ripple down one auxiliary section of the winding (defined by the 1.2k resistor) cancels the AC ripple down the main section of the winding (defined by the pentode output valve's ra, which is quite high). In theory complete cancellation of this AC ripp[le can be achieved; in practice it is limited by component tolerances.

Yes, it's Philips, and yes, it's very elegant!