Transformer Failure - why?

[Top Cap]

Having built myself a stereo valve amplifier, this worked fine for many hours but suddenly switched off one evening. The power fuse had blown and continued to do so even with all transformer output lines disconnected and isolated. A replacement has arrived, it has a slightly different input circuit but its outputs are the same. All readings seem OK except for the 320-0-320 windings are which are significantly lower on the faulty transformer. I suspect a shorted turn somewhere, maybe Ed would be willing to investigate for me? But I did do a modification to the original circuit, using fixed bias and protection relay as opposed to the original cathode bias arrangement.
Is my circuit modification/addition responsible for the failure?
Would like to invite comments on the circuit before I install the replacement tranny.

320V windings are 180mA, output valves are biased to draw 50mA each Class A SET

[Bazz4CQJ]

If the "power" (mains?) fuse is failing when all the secondaries are isolated, it seems to me that the primary side is in question. I'm not greatly influenced in any direction by the resistance measurements.

I'm actually wondering about the 2amp fuses you have; is there any chance they could be dodgy, or at least, they don't cope with the switch on surge? Is the mains switch totally sound?

B

[Top Cap]

Definitely something wrong with the transformer, it causes the lamp on my lamp limiter to shine full brilliance whilst all secondaries are disconnected. If it is the primary then it must be a shorted turn?
I was keen to compare the DC resistance with the replacement and cannot understand the differences on the HT windings. As long as my circuit has not caused the problem by taking more current on one winding compared to the other then I suppose you are correct.

[kalee20]

If it has run for many hours, then there's no design fault (I'm assuming that the transformer has not been running stinking hot all this time, but at a comfortable temperature).

If now it is blowing fuses, then something has failed.

It is almost certainly not a 'shorted turn' (unless in the LT windings), but quite a significant portion of a winding which has got shorted out. A single turn would be responsible for maybe 0.2V. In the primary, or the HT secondary, wound with fine wire, a turn might have a resistance of 0.1 ohm, thus shorting it would generate 2A of circulating current and 0.4W of heat - enough to be a nuisance and maybe produce some wisps of smoke, but not enough to blow a fuse. Whereas shorting a portion comprising 100 turns or so - maybe one layer to the next - would sap correspondingly more power, and cause a big increase in current drawn.

The significant difference between the secondary section resistances is a likely pointer towards the lower of them having an internal short.

I think you are quite safe in connecting your replacement transformer, but just keep a watch on its temperature. With larger winding resistances, it's not going to be quite as efficient as the old one, so until it has run continually some hours, take care!

[Ed_Dinning]

Another possible fault would be if the transformer was badly laminated with the lams not correctly butted together. This will cause the no load current to rise and burnout can follow.
I assume it was not a "Chinese" type with a 220v primary.

Ed

[Herald1360]

Not helpful, but what's an FE50 valve? Can't find any references.

[Refugee]

Originally Posted by Herald1360

Not helpful, but what's an FE50 valve? Can't find any references.

It is shown as having a 12.6V heater too

[kalee20]

Originally Posted by Ed_Dinning

Another possible fault would be if the transformer was badly laminated with the lams not correctly butted together. This will cause the no load current to rise and burnout can follow.
I assume it was not a "Chinese" type with a 220v primary.

True, but in that case the transformer would have been slowly cooking since Day 1 till the enamel degraded to failure point - and the OP seems to indicate things were fine in the beginning (maybe he can confirm?).

Whereas a random insulation failure could be caused by a mains transient finding a weak spot between two points of significant voltage.

[turretslug]

I've a hunch that "FE50" may be the outcome of successive misguided phonetic transliterations....,

http://www.r-type.org/exhib/aaa1539.htm

[kevinaston1]

What is your mains voltage?

I ask this because many years ago, I worked in for a big company in Birmingham. Our failure rates on mains transformers was far higher then other service departments.

On investigation, we found our local supply was often around 255V, and would even go higher than this; we had a constant stream of HiFi and radio products on our awaiting repair shelves, all having shorted turns on their primaries.

Last year, whilst helping out a local company (I am retired), I arrived one morning to find 12 of the same model "Disco" amplifiers; each with its massive toroidal transformer virtually melted. I only had one replacement transformer, but fitting that to one of the units brought it back to perfect health - 650 watts per channel.

Kevin

[joebog1]

I have had a few failures like this over the years. I have never been able to discern why, but some time spent thinking about it, I formed this opinion.
A lightning strike, or massive inductive spike on the mains has blown a hole in the insulation. This causes the massive short of many turns as previously mentioned.
I now fit GEMOVS, or voltage dependent resistors to EVERY transformer I use.
At least one across the primary can do no harm and will only rob the pocket of about a quid. They are available up to about 500 volts very commonly, and higher voltages are available on order. They come in various sizes with capabilities ranging from a few amps to thousands of amps, and still only occupy the space of a 20p coin.

What started me thinking about this type of failure arose after blowing a toroidal transformer TWICE in the same amplifier, which I had designed with more than enough safety in hand. I painstakingly unwound the toroid turn by turn, down through the heater windings first, then the secondary and finally the primary, to discover a hole blown clean through about two layers of windings ( it looked like it had been spot welded).

This was caused " I reckon" by the transformers own back EMF at switchoff, as, in my normal manner I overwind transformers to give low flux densities, and consequently huge inductances. What I end up with is a huge Kettering ignition system, and a few pennies renders it useless, and protects the transformer to a large extent from any nasties that come in on the mains line.

Joe

[PJL]

I would have thought the choke or the series resistors in the bias supply would be the first to burn out on excessive DC current. This tends to imply a transformer manufacturing fault although if you are using RF power valves I suppose there are all sorts of possibilities if it all went haywire.

It's alot of bias voltage and current. What does the relay do?

[Top Cap]

Transformer has a 230V or 240V input, I was using the 240V winding. There was no sign of sparks or smoke, the amp just turned off and fuse blew instantly on replacement. The FU50 is the Chinese GU50, I guess they obtained the design from Russia, who copied it from the captured German Tank transmitters. After I set the feedback to the correct level it sounded very nice indeed. The relay interrupts HT if the bias voltage goes down, just an anti-red plate device really, though looking at the valves I don't think they would worry too much . Well I shall fit the replacement and see how it all goes, £85 with postage so I don't want to lose another.

[trobbins]

I would always recommend a double protection scheme for secondary side valve rectified HV supplies:
1. insert 1N4007 (or 2 in series) in series with each valve diode anode.
2. include a fuse in the CT (caveats apply but for that circuit the CT is most appropriate).

There is definitely an arcing failure mechanism with valve diodes that will short out the HV secondary. Both those changes remove that risk. The fuse also provides much better discrimination for over-current faults on the B+, which typically occur again due to valves in the output stage. With judicious selection of the CT fuse, the fuse rating can get down quite low and so has more of a chance to blow before a power transformer secondary if the fault current is significantly limited by circuit resistances.

Some fuse selection guidelines are in:
https://www.dalmura.com.au/static/Va...p%20fusing.pdf

I'm with Joe on the use of a MOV across the power transformer primary winding, as a means to reduce risk from inductive spikes when the primary is switched.

It sounds like a cooked HT winding causing a shorted turn(s), which isn't easy to confirm without a tear down to either see a fault or to just leave the primary winding intact (and then confirm the primary winding just draws nominal excitation current).

About the only other practical test is to do a 1kV megger, to identify if inter-winding insulation has broken down, but that doesn't seem likely as it would show up as an earth leakage trip if the breakdown was primary winding related.

[broadgage]

To add to the points already made, a centre tapped HT winding with one half having a significantly higher resistance than the other half, MAY be OK.
a common construction of these transformers places one half of the winding over the other half.
The second half will have a higher resistance because each turn is longer due to the space taken up by the first half.

Some designs have a few extra turns on the second half of the winding to partially compensate for the extra voltage drop in the longer turns.

If this is done, then the outer winding will have a higher voltage off load, due to the extra turns.
The inner winding will have a higher voltage under full load due to the lesser voltage drop in the shorter turns.
At about half load these two effects cancel out and each half of the winding will have a closely matched voltage. This minimises ripple on the HT.

[John10b]

Just a small comment on looking for a transformer fault.
The resistance reading of a faulty transformer is not a good indication of a faulty winding, unless of course you have a major burn out.
I have stripped down many faulty ones and found a flash over but the resistance reading was virtually identical to a new one.
I used to ring a new transformer and look for the decay on a scope, and then do the same for the faulty one, you will see a big difference. I should add this was particularly useful with line output transformers from TV’s.
Cheers
John

[joebog1]

broadgage mentions a good point!! In EI transformers of my own making I generally make the secondaries in four sections. I wire the sections in a way similar to an output transformer to "balance" ( well it is a huge improvement ) the secondary impedances/resistances. ( See The Radiotron Designers Handbook, by Langford-Smith page 218, also page 223, of the Sixth Impression, 1963) for a better understanding). Langford-Smith was discussing leakage inductances, and winding capacitances, but equally apply to winding resistances, and balancing of the same.

I also incorporate that weak bit of wire called a fuse, as trobbins suggests. With correct ratings they will prevent more than 90% of problems.

Joe

[Diabolical Artificer]

A tfmr I'm working at present had two secondary's burnt out, DC res reading's were something like 47r 0.2r on a 166 0 166v winding. They were well cooked, all the insulation had burnt off and some of the winding's had fused in places.

Andy.

[HamishBoxer]

I am confused by C25,an electrolytic with AC on it??

[ms660]

Looks like DC from the pot divider across the HT, maybe to reduce heater cathode voltage stress in the 6SN7 if it's cathodes are well above chassis potential?

Lawrence.