Being kind to your output transformer

[Mike Phelan]

When I were but a lad, and we feasted on Hovis and listened to Dvorak's 9th Symphony, I was told that never under any circumstances should I run a radio with no speaker, as it would damage the output transformer.

Being of an enquiring mind, I asked my colleagues 'why?' They said that the AC component of the audio would damage the insulation of the primary.

I have always adhered to the precaution, but am questioning it again, as I saw plenty of o/c speakers and transformers that reproduced faint audio, but none that were damaged.

So, is this just an old wives' tale? Or is there a reason for it?
What does the team think?

On a similar note, assuming it is all true, presumably there would be no damage with no speaker, providing the VC was turned right off; say, if you just wanted to check some DC conditions.

[Duke_Nukem]

I always thought that under these circumstances it was the output valve that was given more of thrashing than the o/p transformer.

TTFN,
Jon

[YC-156]

Hi Mike,

Originally Posted by Mike Phelan

So, is this just an old wives' tale? Or is there a reason for it?
What does the team think?

Hmm, my answer probably belongs in the 'it depends' department.

In a triode amp changing the operating conditions to a vertical loadline, Ie. infinite load impedance, might actually alter the operating conditions in such a way that the valve no longer operates in pure class A (for a single ended output stage). If it starts to run in class AB or worse, the anode current will be definitely discontinuous, and the back-EMF of the output transformer primary is free to run amok as the anode current suddently stop during each audio cycle. This, I would suspect, could be bad for it.

Edit: This would probably only happen at or near maximum volume, which is probably not encountered that often. Unless, of course, the speaker is faulty and the user tries to adjust the volume to get some audio out of the set. At low or no volume the damage potential would be nil methinks.

In a single ended tetrode/pentode output amp, it would require some serious misfortune for the anode current to become discontinuous, so the danger of transformer damage would be much less.

The worst case scenario is probably a push-pull class AB or B stage. During normal operation the current in each valve is precisely balanced, such that the current just accurately flows continuously during at least one of the valves for the whole audio cycle. It wouldn't take much to offset this balance, causing serious back-EMF problems twice every audio cycle.

So for older sets with triode output stages the danger is there, at least in principle, but the more modern sets with class A pentode finals I wouldn't worry about it.

Just my £0.02.

Frank N.

[ppppenguin]

As Franks states, the problem is back EMF causing insulation failure in the transformer. May not be too good for the valve either. Turning up the volume too high to cause clipping would give discontinuities in the waveform that induce high back EMF. Also motorboating, clicks and pops etc. You can't hear them so you don't know about the problem. I would prefer not to take the risk. If the speaker is disconnected it's no trouble to use a dummy load. Doesn't have to be the same impedance as the speaker. Even a 47R resistor in place of a 3R speaker should keep everything safe.

[Station X]

I had an AR88 output transformer go open circuit primary when the speaker became disconnected due to a dodgy plug. I understand that this transformer is particularly susceptible to going o/c if the speaker is disconnected or has the wrong impedance value. A set of contacts on the headphone jack connects a 5 ohm resistor across the secondary when headphones are plugged in. The output valve is a 6K6GT pentode.

I suspect that most o/p transformers burn out when the o/p valve passes excess current due to faulty biasing.

Graham.

[Mike Phelan]

Thanks for all your comments. I did suspect that the problem was something to do with the spikes of audio appearing at the anode.

I have never run with a transformer unloaded and in view of this discussion, I will always load it.

[Biggles]

I've always gone by the general rule "never run a valve amp into an open circuit, never run a transistor amp into a short circuit" but there are probably exceptions to this? Sometimes you can spot trouble with an output transformer loading because the windings will "sing" in sympathy with the audio. I think it would be safe enough to run an output transformer with no load as long as there was no audio passing through it i.e. under d.c. conditions only, but it is better to run it into a dummy load.

Biggles.

[Station X]

I think it would be safe enough to run an output transformer with no load as long as there was no audio passing through it i.e. under d.c. conditions only, but it is better to run it into a dummy load.

I've been doing a similar thing this morning, checking out the DC conditions on a push-pull AF modulation amplifier of a valve AM transmitter. The mod amp can produce about 40 watts of audio. There was no load on the secondary of the modulation transformer, as I hadn't applied power to the RF stages.

One thing to be very careful of when doing this is to make sure that the amplifier doesn't start oscillating due to feedback, as the amplifier will then start amplifying and there will be AC in the transformer.

Graham.

[Paul Stenning]

I had a tone correction capacitor (I can't remember the type, but it was new) across an o/p tx go short-circuit when the temporary speaker connection fell off on a set I was working on a few years ago. It was playing at a reasonable listening level at the time.

[Mike Phelan]

Originally Posted by Paul Stenning

I had a tone correction capacitor (I can't remember the type, but it was new) across an o/p tx go short-circuit when the temporary speaker connection fell off on a set I was working on a few years ago. It was playing at a reasonable listening level at the time.

I suppose it would not take too much for the self-inductance of an undamped primary driven by a few tens of volts at the anode to do that - hence the working voltages used on these components.

In summary to me throwing the question around, the answer seems to be 'don't do it'

[Heatercathodeshort]

Hello,
To put it in simple terms, an unloaded audio output transformer acts rather like the line output stage of a television receiver. Very high peak voltages appear on the anode especially if it is being driven at high audio frequency.
In practise valves are very rugged devices and can take incredible punishment with little complaint. I have had the odd case where frame coils have gone o/c on television sets causing arcing in the output valve holder but no actual damage to our faithful vacuum friends.
Regards. JOHN.

[zak]

The output transformer is a load matching device,to get max power into the speaker it has to be matched to the source.What the high impedance of the anode see's is the 3ohm speaker multipied by the impedance ratio,which is the square of the turns ratio,if the speaker is o/c (or disconnected) the anode impedance will be very high,and since the transformer is current driven,you are going to have high ac volts,you could have several kv,esp on sharp spikes,at high volume levels,this is enough to damage insulation or cause flashover.There is often a tone corrector capacitor/resistor combination across the primary,this loads the transformer,esp hf and high volts and certainly holds down the spikes,this may help to save the transformer if you were foolish enough to run the transformer without the speaker.Another thing,is if the transformer is unloaded,even for a fraction of a second,some damage may be caused,and although the transformer may appear to operate as normal afterwards,the degraded insulation may give rise to crackles,and pops,intermittantly,very difficult to trace.

[Biggles]

I once read an article years ago suggesting connecting small neon lamps across the primary of an output transformer to suppress any high voltages if a speaker became disconnected in a PA set up. I suppose that there is more chance with PA equipment of a speaker being disconnected inadvertantly? The neons operate like a zener and normally dont strike due to lack of voltage, but if a high voltage developed then they would strike and suppress it. You could maybe use "varisters" to do the job nowadays instead of neons.

Biggles.

[zak]

Originally Posted by Biggles

I once read an article years ago suggesting connecting small neon lamps across the primary of an output transformer to suppress any high voltages if a speaker became disconnected in a PA set up. I suppose that there is more chance with PA equipment of a speaker being disconnected inadvertantly? The neons operate like a zener and normally dont strike due to lack of voltage, but if a high voltage developed then they would strike and suppress it. You could maybe use "varisters" to do the job nowadays instead of neons.

Biggles.

This seems a good idea,but you need a neon with a fairly high strike voltage,most strike at 90 or so volts,which I think may be a bit low,it certainly is worth trying though,anything that may save the trany is worth a try.

[mikelect]

Could you use two neons in series to give 180v


Mike

[mickjjo]

I would have thought that using neons would be undesirable, surely they would light on volume peaks and cause horrendous distortion . Must try it one day..... .

Regards, Mick. .

[Mike Phelan]

Originally Posted by mickjjo

I would have thought that using neons would be undesirable, surely they would light on volume peaks and cause horrendous distortion . Must try it one day..... .

Regards, Mick. .

I'll bet Adi will before you do, Mick

[jay_oldstuff]

I have a little practise amp home made circa 1960's acording to the person who gave it to me (actuely think its an old tape recorder amp its got a magic eye on it) anyway i digress with the speaker disconected and my gitar lead pluged in it actuely receves a radio station (no coil or tuning condencer in there) and the sound seems to come from the output tx. i discovered this before lerning it was a bad idea to run a valve amp into no load has'nt seemed to do it any harm though

Jay