Toroids as O/P trannies

[boiss]

Has anyone had any experience of using mains toroids as PP audio O/P transformers. I found some info on Audio asylum and something else from Oz. It looks like they work rather well if balanced reasonably and they are significantly cheaper than audio O/P trans.?
Mike

[Denis G4DWC]

I have succesfuly used a small Toroidal Mains transformer as an Output Transformer in a Telefunken Tape recorder, Magnetephon KL85 I think or something similar.

The original transformer had a tertiary winding for negative feedback so it wasn't straightforward to find a replacement. I think I used a 9V + 9V type (Might have been 6V + 6V) from Farnell which gave a reasonable impedance match (Impedance ratio = Voltage ratio squared). One secondary was used for the Loudspeaker and the other for Feedback. It worked fine. Sadly some months later the motor burnt out and as yet haven't got a replacement. Anyone got one??

These transformers are unlikely to work in a single ended output stage as the quiescent standing current will almost certainly saturate the core. It would be wise to make sure that both valves in a push pull stage draw similar quiescent currents as a significant difference may still saturate the core.

It should also be noted that if the primary is wound in a bifilar fashion (Both primary sections wound together side by side), the insulation may not be up to the job if the HT is too high. In my small 6W push pull stage ( 2 x EL95) it worked fine. At least it got it working again.

Denis

[quantum]

I've used mains transformers as emergency p-p output transformers on one or two occasions, and they work, but be aware that the construction of these transformers will probably give you a restricted frequency response, so I don't think that hi-fi applications are in order! However, for a replacement in a set working on the AM bands where treble is restricted, you would probably find such a replacement adequate.

Even with single ended stages, you could probably get away using a mains transformer in a stage using a battery valve or a small mains output valve - I've experimented with such and found no problem with core saturation.

[Biggles]

I've used a normal mains transformer as an single ended output transformer in a home built set when I had nothing else suitable. The output stage was for a DL96 so the standing quiescent current was very low and probably wasn't enough to saturate the core. It worked ok but the sound was a bit "woolly" compared to the proper replacement transformer I bought from Isoplethics.
Worth a bit of experimenting though. Adi?

Biggles.

[boiss]

That's an interesting point about the bifilar windings regarding the insulation and something I hadn't considered, especially as I could have 1KV across the windings. As far as the frequency response goes, they appear to be excellent from what I've managed to find on the net, with measured -3dB points of <20Hz and >150KHz for toroids up to 250VA. Opinion differs on Audio Assylum from working better than any EI core transformer but not as good as some of the best C core transformers to not being worth considering. Because Audio 100W O/P transformers cost well over £100 each if they are supposed to be any good, Financially out of the equation, some thought needs to be given to rewinding a toroidal transformer. It appears rather daunting but if it has only about 4 turns per volt, It may not be all that difficult after all to put on 2 separate windings.
Mike

[Denis G4DWC]

Mike
I also measured the frequency response of the Toroid before I used it in the Telefunken. It was was pretty good to well over 20kHz. The owner of the Tape recorder said it sounded just like it had always done so.

The difference between a mains transformer and an output transformer is the method of winding to keep the leakage inductance down. Have a look at how the Williamson Output Transformer was wound in the following link and that will give you an idea why a decent output transformer costs a fair bit..

http://www.xs4all.nl/~ideas/amps/chapt7.html

I spoke to the people who took over Avel Lindberg today and they confirmed the primary of their transformers is bifilar wound so when you connect the two primaries in series there is only the enamel insulation between centre tap and what would be the Valve Anodes.

Denis

[Aub]

Hi,
I've used mains HT transformers as modulation transformers in AM amateur transmitters. The results are usually quite acceptable for the comparitively narrow frequency response required. However, I've never tried using a toroidal type, probably because you don't tend to find HT torroids. I supose the closest thing would be a torroidal mains isolation transformer.

Good luck!

Aub

[Heatercathodeshort]

Hello,
Mains transformers of all types usually make surprisingly good audio output transformers. When I was a kid [47 years ago] I used old television frame output transformers as well...The construction is very similar if not to a higher spec. Regards John.

[Ed_Dinning]

Hi Gents, most mains transformers run at about 1.5Tesla flux density. A quality O/P trans will run at half that and usually use Unisil grain oriented lams which go into saturation quite gently. This all helps to reduce distortion. Mains toroidals usually have quite a thick "lam" and will be lossier for the HR spectrum. They still work quite well and give close coupling. Thet are only really suitable for push-pull as it is not easy (but has been done ) to gap a toroid for single ended use.
The Oz designs use Unisil and thin strip. They also have screening fiol over the various windings to reduce capacitive coupling which would affect the feedback circuits.

Ed

[boiss]

The bifilar winding of the primary would appear to stop any thoughts of high power levels with these toroids unless the valve impedances are very low.
I take note of the Williamson design althogh it is not for a toroid. I must admit I forget or maybe I never knew what Inductance leakage is.
However because surplus toroids are very cheap around £10 for a 300VA it's worth a try at rewinding one. The biggest problem would appear to be getting the wire onto the core, just what does a hand wiring shuttle look like?
Mike.

[Denis G4DWC]

I have only wound secondaries on a Toroidal Mains Transformer for my first computer (Tangerine, I still have it!)
I had the XYL holding the toroid, me doing the threading and keeping the windings neat and my young daughter, about 6 at the time, running up and down the lawn with the end of the wire.
You could make a shuttle from a piece of wood but I suspect you would need at least two pairs of hands to do the job. A primary would be pretty time consuming...

Denis

[Ed_Dinning]

Hi Gents, Toroid winding, use a saddle and stirup.
Saddle is a place on the bench in front of you with a wood block and felt pad, hole in toroid faces towards operator.
Stirup is a loop of belting or similar, it is anchored behind toroid and passes through the central hole and one foot is placed in it to anchor the toroid to the bench.
Shuttle is an "H" shaped piece of wood with sufficient space ob the centre limb of the H for all the wire. Continusly pass shuttle through toroid and move toroid on saddle to keep winding at the top. Quite simple really.

Ed

[boiss]

Hi Ed,
That sounds terrible at first but when the idea sinks in it makes some sense. I've got a 300VA toroid so on Sunday I'll wind some wire on it and see what the turns per volt ratio is. I guess that the H section should be long in the "Z" axis in order to hold as much wire as possible.
Mike

[Ed_Dinning]

Hi Mike, thats it.
It is usually possible to have joins in the layers if there is not enough space on the shuttle. From you deacription I would guess about 4 to 6 TPV.
Double this figure if you can for better fidelity and linearity.

Ed

[boiss]

Ed, One thing that I've noticed is that they are not all wound in the same way. On a 100VA one the primary is on the outside while on the 300VA one the secondary is on the outside. on the "300" the secondary is 12V quadrifilar. Im not going to try and rewind this one because I intend to use it for the heater supply (10V at 20A), it was to see what the tpv was. When I get another I suppose that I would need to rewind the whole thing anyway but not using multiple wires, after all 100W into 8 ohms is only about 6A pk. If I put the secondary on the outside I can play with the number of turns and matching quite easily. Do I have to use all of the toroid for each half of the primary or can I wind one winding per 180 degs of the toroid which would appear quicker and easier?
Mike.

[boiss]

OK, I have looked at this 300VA tranny and the tpv is approx 3, ie. with ten turns wound on the output was 3.24V. I was wrong about the secondary, it's not quadrifilar but sextofilar with wire around 0.8mm dia. The DC resistance of the primary is 4 Ohms. After doing some tests with a series resistor on the primary, I found that the inductance of the transformer off load is of the order of 73H, this would appear to be a pretty good ballpark figure for an Audio O/P transformer. The resistance is also very low at 4 Ohms.
Each turn on the toroid would be a tad under 8" so if a 12" long shuttle was used it would only need a third of the total number of turns on it.
As for the windings, if two separate primary windings are put on the toroid with each occupying only 180 degrees of the toroid they will be more matched than by putting one on top of the other.
Any suggestions?
Mike.

[Ed_Dinning]

Hi Mike, the tricks you describe are equally valid on all types of transformer.
Multifilar windings are often used to ease handling of thick wires or to "fill" a layer without too much "build up"
Becaus eof the closed magnetic circuit on a toroid the 1/2 primaries can be wound opposite each other instead of on top. This may ease insulation requirements.
Mains primaries were often wound this way for 110/220 volt use by series parallel connection.

On thing to try is to plot the V/A curve of the core with a given number of turns. It should be a straight line up to the turns per volt figure you will use in practice. For Hi Fi audio use you should aim to work at about 1/2 the start of saturation level.

Ed

[boiss]

OK Ed,
How do I get the V/A curve for the core?
Mike.

[Ed_Dinning]

Hi Mike, you need a variac or a tapped transformer to give a variable source of AC voltage. This should go from near zero to 10/20% above the nominal rating of the transformer in question. Measure volts applied to winding and current flowing into it with no load on the secondaries.
Plot results and you have it.
There may some slight curling at the start of the plot and it should be just starting to form a "knee" at nominal voltage on a good transformer. You can then work out your turns/volt figure for that tranny.

Note that a variable power resistor will not work properly due to phase shifts

This is a good way to do it if you can't remember the formula:

N= V/ 4.44 B F Ae (or you do not know the B value for the lams used)

Ed

[boiss]

Hi Ed,
Unfortunately I dont have a variac, the only voltages that I could apply via a transformer are 6V and 110V, not a great deal of use. Another thing is the accuracy of the current reading I would obtain because unloaded the transformer only draws approx 10mA at 220V. I am unable to measure directly AC current, this value was obtained by measuring the voltage across a 68 Ohm resistor in series with the transformer (0.67V). Also I've no Idea what the value of "B" is.
I think that I will get hold of another one if I can and try and rewind it with 345 turns on each primary and a tapped secondary winding. I have seen one on ebay but there's still 7 days to go. As I said before the one that I have already will be used for the heaters. I intend to run the heaters on AC and play with the phase and balance the voltage across the heaters to reduce the hum. If it's not acceptable then it will have to be a stabilised DC supply with mega heatsinks. I guess it will keep me warm in the winter.
Mike.