| Created: 26/12/06 | Last update: 17:02 27/12/06
<<< OzValveAmps

Hot Tranny

Possible causes of tranformer overheating.

New: 26/12/06

Evan wrote...

I got a little old 5w AWA PA amp going last night (model 828) ... seems to go alright, apart from a considerable level of hum, ...
[6AV6 pre, 12AX7 post, 6V6 single-ended o/p, 6X4 rec]

What concerned me was that there was a strong smell when I ran the amp (I only ran it for a total of about 10 mins) which I finally traced back to the power transformer which was incredibly hot!

It was too hot to hold my finger against even after a couple of minutes of being turned off. I should be glad if I haven't done any permanent damage to it. The amp operated fine (with the exception of the hum) and there was no sign of damage done, except for the smell during operation.

Do you have any info on your pages about this issue (overheating power transformer)?


This is a clear case of a transformer seriously overheating, smell and too hot to touch in only a few minutes.

What's “normal”?

Generally speaking a power transformer may get just a bit too hot to hold your finger after an hour of hard drive, and still be healthy. For various reasons (mainly economy of materials v running costs) transformers of American manufacture (and thus from most Asian suppliers) tend to be specified for greater losses and a higher maximum temperature, say 65°C, than Australian designs (e.g. A+R), say 50°C max.

But rapid heating to this extent after only a few minutes at high power can be a sign of trouble. On idle, without drive, transformers should not get more than warm to the touch - you should be able to put your hand on it with no discomfort even after an hour.

The Rule of Finger here is if you can't hold your finger on the core or winding for more than a few seconds, then it's too hot. (It's the same rule for silicon transistors too).

Excessive mains hum can be a symptom of an electrical overload, and transformers often “growl” in proportion to the load. It is normal for a power transformer to bump or thump at the moment of switch-on (the discharged electrolytic caps in the pover supply look like a short circuit for a few moments while they initially charge up), but a sustained throaty growl like a V-8 is a sign of trouble, blowing fuses or impending smoke.

In amps of all types this is commonly caused by the main solid-state rectifier failing short across the tranny secondary, so that is a “must check” (but there may be other faults that caused this).

There are two basic reasons why a power (or any) tranny might overheat like this, and we need to localise the problem.

* The first (hopefully) is that there is some fault within the amp drawing too much current through the tranny.

* The second is that the tranny itself has a shorted turn dissipating power within the tranny.

When a turn or turns short together they form a shorted turn which acts much as an external load or short-circuited winding would behave. This is bad news.

So how can you tell?

Here is a circuit fragment of the power supply and output stage in question.

The gizmo on the left, VIB-1, is a non-sync vibrator
for operation from 6 volt accumulators or car battery of the day.

Possible causes

* Heater circuit

If the valve heaters all light to the reasonable brightness then you can assume the heater wiring isn't overloaded.

Heater circuit faults are fairly uncommon, however filament pilot lights fail and often get fiddled with so occasionally you may find a short from one side of the pilot light holder to ground.

The other power supply output is the main HT rail, and this may be excessively loaded by something. Here we will start furthest from the power supply and work back towards the power transformer.

The tranny is getting hot because too much current is flowing somewhere, and in many cases (but not always) that “somewhere” will also be getting hot due to the excessive current. So we are on the lookout for other hot-spots.


Generally the preamp stages can't draw excess HT current, even when faulty, to overheat the power tranny (but a reverb driver may be an exception). Symptoms of excessive preamp current will be heating of the bypass resistors in the HT line (R23 10k, R17 47k).

Around the output stage

* Input coupling cap C6 0.05uF leaky
* (local feedback cap C7 22pF leaky)

With either of the above some current will be injected into the output stage grid circuit, pulling it positive, reducing the bias, and causing the output valve to draw excessive current. With the waxed-paper caps in use here this is quite possible. The 22pF is also possible but less likely as these low values tend to have good insulation such as silvered mica.

Symptoms: hum, distressed and overheating 6V6 and 6V4, and perhaps output tranny as well, hot cathode resistor R20 330r, cathode volts high.

* Cathode bypass cap C34 25uF leaky or short (note in this case the heater line is also biassed by the cathode voltage).

This will have a similar effect because the by-pass cap is shorting the bias voltage. This by-pass cap is an electrolytic type and therefore will age and fail similarly to one in the HT, but generally more slowly. While cathode by-passes last longer however, they don't last forever and at the age of this chassis all components must must be suspect.

Symptoms: hum, distressed and overheating 6V6 and 6V4, and perhaps output tranny as well, cathode volts low, poor sound.

Screen bypass C36 8uF leaky or shorted.

Symptoms: cool 6V6, but bypass R23 10k and 6V4 hot, poor output.

Short in the output tranny (to frame, or to grounded secondary).

Symptoms: no HT, dead, 6V4 very hot (red plates).

Around the power supply

Main filter cap C8 8uF leaky

Symptoms: hum, 6V4 hot

6V4 rectifier gassy.

This can be very hard on the power transformer. A gassy valve can be recognised by a cloudy blue glow within the envelope or valve structure. This should not be confused with the healthy blue blush in the glass of output valves.

Valves that are only mildly gassy may only have the glow faintly inside the elements, and may self-cure if run under limiting conditions for a few hours.

But any significant blue glow is generally an end-of-life sign. In severe cases the whole valve may glow bright blue-purple, eventually offset by a cheery red glow from the anodes and smoke effects from the tranny.

Symptoms: blue glow inside rectifier, running hot (red plates)

Suppressor caps C9, 10, 12 and 13 directly on the transformer windings, leaky or short.

In the 1960's these sticky waxed-paper type caps were considered to be replace-on-sight, and now most are over 50 years old they must be assumed to be faulty on sight. These ones are really only required for 6 volt vibrator operation to suppress contact hash. In this case these are my prime suspects.

Vibrator-type inverters were found in radios from the 30's and 40's before mains electricity was common in the bush. These range from 6 volt which were common car batteries, to 32 volt used in wind-generator homelighting systems. They were also normal in the valve car radio's of the era. Solid state vibrator replacements also became available, so it can be done.

And that leaves the transformer.

Danger! sign Shock Risk sign

* Disconnect all of the secondary windings, heater and HT, taking careful note of where they are connected. Only leave the mains input connected. (it is best to try and disconnect the power supply from the tranny, leaving its leads firmly terminated, rather than the tranny from the supply).

* Power the transformer via a light globe, say 40 watt.

If the tranny is okay the lamp won't light and you should be able to measure reasonable voltages on the disconnected windings. The fault is elsewhere.

If you short any of the windings (carefully!) the lamp should light brightly.

If the lamp lights and you are sure there is nothing connected to any secondary, then the tranny has a shorted turn.

If this is the case then the tranny is stuffed and can only be recovered by replacement or rewinding.

I dips me lid to Evan Lowden

Valid HTML 4.01 Transitional <<<OzValveAmps