how warm should a transformer be?

[AltecDarpa]

I have a 12.5v transformer at 3.5A powering a TDA2005 amp, using the test circuit provided by the chip maker, the transformer gets warm after a while at high volume, not hot but just warm. is this normal?

I am driving 8ohm speakers with it, would this make it warmer then using 4ohm speakers?

[WhiteD3vil]

Assuming that the transformer is rated for the increased current through the 4-ohm load, it will get warmer (higher resistance = less current = less heat).

[SGregory]

Warm is not usually a problem. Some trannies get so warm you can barely hold you hand on them. The key is to determine if it gets warms and stays at that temp or if it continues to get hotter and hotter. Not knowing your circuit and values it is hard to determine if it is in an overload condition.

Just make sure it is fused properly. I am not familiar with the TDA2005 chip but typically for a given rail voltage, the higher the load impedance the less power can be driven and the less heating in the supply. Not a perfect rule but generally valid.

[rjb]

If the transformer heats very quickly, you have a problem. If the transformer heats up slowly and reaches an even temperature, then that is not necessarily bad.
Basically the temperature inside any transformer is much hotter than the outside, and this heat is allowed for in the design, assuming it is not so great as to break down the insulation and cause it to fail catastrophically. The designer will rate his transformer for a given ambient temperature, and a given life, as the insulation life will be shorter as the temperature rises. A cool transformer lasts longer than a hot one. So a designer will user a cooler running transformer for something run continuously, such as a clock radio, and a hotter (ie cheaper) one for something seldom used, such as a battery charger.
For audio equipment my rule of thumb is mildly warm is best, able to hold your hand permanently on it is just OK, not able to hold hand on it is bad.
The losses generated within the transformer cause the heat. To greatly simplify, as voltage is basically fixed, it is the current drawn that has most effect. So you might expect, (depending on the amplifier) 4 ohm speakers to draw more and hence be worse than 8 ohm . Roughly the amplifier available power may be doubled, hence the current doubled. But actual drain depends on the "class" of amplifier, the relative efficiency of the speakers and the volume you run them at. In fact, given equal noise output, the transformer load may not be that different. Also many transformers are designed so the voltage output sags, so decreasing the available current at lower impedance loads. So not a precise answer I'm afraid.

[AltecDarpa]

Thanks for the responses, I am learning much on this forum.

It just gets warm and stays warm, not hot at all. I was just unsure if this was ok.

The amp is a class B, it is rated to 1.6ohms, but I would not try that on the current Transformer.

http://kitsrus.com/pdf/tda2005.pdf

it lists the output peak repetitive currant at 3.5A, so that does not leave me much of a safety net, I will try to find a 4.5A-5A trafo for this project before I try any lower ohm speakers.


on another note I was quite impressed with the amps performance, I had always heard class B amps had high distortion, this one is actually one of the better sounding amps I have heard, out of rotel, mac, nikko, yamaha, ect.

[Th3 uN1Qu3]

The TDA2005 isn't "pure" class B. But even if it were, with enough negative feedback you can get rid of the distortion. I experimented with this a while back... An OPA2134 and two transistors, a BD139/BD140 pair. Giving out a respectable 3.5W @ 8ohm. The distortion was only noticeable on low sine waves (200Hz and under). With music it sounded quite nice for only 3 active components.

The bad thing about class B distortion (crossover distortion) is that it is of equal power at all harmonics, decreasing only when the transistors run out of bandwidth (almost in the MHz range...). This is why it's so objectionable.

[Ashley Ace]

Usually something about 40C above the environment is normal. Assuming the room air temp to be about 22C, the tranny`s 62C is normal. And perceptively, 62C is really hot so you can barely touch it.