How hot should a resistor get?

[parkerbeder]

I have a power supply I just fixed, and now it likes to kick off every few minutes. I tore it back open and saw that ther is a 3W 36K resistor placed pretty close to a thermistor. It's getting to... I just checked it and it's at 220F. is this a big issue, or do these guys just get hot? It's pretty much certain that I bumped the thermistor closer to the resistor than it was originally, and it's the thermal protection cerciut that is kicking it off, my question is if it's a big deal to just bend the thermistor away from the resistor a bit, and have a go at it? or do people usually design thermal protection circuits with a resistor next to a thermistor like that to get a feel of the load under the circuit?

Thanks again for any response!

-Parker

 

[Vizier87]

did you say you modified the circuit and made the resistor go closer to he thermistor? Then that must've been the problem. Route the resistor somewhere else then. If space is an issue... well perhaps you can put a 5V desktop cooler fan to blow the resistor's heat away.

 

[tvtech]

For the design to work as intended, I would say that the thermal coupling between the thermistor and resistor should be as good as possible ie: they should at least be touching each other.

I reckon there is something else still wrong with your power supply.

 

[cowana]

Why not replace the resistor with a 5W version?

Or two 3W ones (of twice the resistance) in parallel.

 

[tvtech]

You folks are missing the point here. Original design used a 3W Resistor. For a reason.

A 5W Resistor defeats the object of the PSU protection.
Sure, it will run cooler....and then what??

Just asking.

 

[Vizier87]

I'll have to say that makes sense. but what if the problem's not with the resistor?

 

[Sceadwian]

Are you sure the 3W resistor has anything to do with the thermal protection system? It's probably just a static load placed on the supply for stability, and should be mounted to some kind of heat sink, probably directly to the case, I've noticed them in a few PSUs. For a thermal protection system sanity would suggested that the thermistor be attached directly to or very near the output transistors or the output transistor heatsink. Attaching a thermistor to a power resistor doesn't really make any kind of logical sense. It also doesn't make sense that a thermistor would be free floating around the inside of the PSU, or the resistor in question. Perhaps you nocked the thermistor off of whatever it was attached to? If It's lose in the case the only thing I can think of is that it's there to measure air temperature in the case..

 

[Vizier87]

a picture'll helpalot.

 

[parkerbeder]

**broken link removed**

if you can see, the thermistor is just hanging out inside the case, probably looking for ambient. that resistor is getting pretty darn warm. if I leave it all apart like that, it'll run indefinitely, but if I bolt it back where it belongs, it lasts 2-3 minutes. I'm thinking definitely a heat issue, but I can't figure out what could have changed that would make it run hotter. Originally the power supply was blinking, so I replaced all the caps with panasonics of at least the same rating. Is there something else I should change also?

 

[tvtech]

Try and take a photo for us. It will help a lot.

 

[parkerbeder]

I had put a pic in the last post, is it not showing up?

 

[tvtech]

I had put a pic in the last post, is it not showing up?

Cannot see a photo...

 

[parkerbeder]

**broken link removed** Hopefully facebook's hosting is more universal than sprint... Does this work?

 

[tvtech]

OK. Thats way better. From what I can see, the thermistor looks to be close to the PSU heatsink. Makes a little more sense.

One more photo needed. Close up of thermistor/heatsink area. And that 39K 3W included in the photo too.

 

[parkerbeder]

**broken link removed** Here ya go, hopefully that will shed more light on the subject...

Also, those big 100uf caps are a bit longer than the OE ones, making them touch the case on the top, do they carry a charge across the top plate? I have been tucking a piece of thin cardboard in there to theoretically insulate them, but might a spark be jumping the gap when the caps are charged?

 

[Vizier87]

probably you can just bend the resistor to the other side towards those diodes and see if it helps?

 

[parkerbeder]

I had already tried that, though i didn't bend them too far apart, and it does, kindof. If I keep the supply unbolted like that, it does, then, run indefinitely. But, as soon as I bolt it back to that panel that it's sitting next to in the pic, it seems to get hot, or for some reason, kick off every few minutes. It will, after this fault, reset and function again, after a minute or two.

 

[Birdman Adam]

Honestly, if the supply works, then leave it. Every power resistor I've come across has been almost/too hot to touch, when everything was working properly.

Or just bend the thermistor & resistor away from each other...?

 

[parkerbeder]

The problem is that it works, unassembled, but you can't really stick a sound board in a recording studio with the back panel pulled off, power supply jerked out to stay cool...

 

[tvtech]

Wow. I can see clearly now. At last.

OK. The thermistor is an NTC. It's job is to limit inrush current when the Power Supply is first fired up. The 36 or 39K resistor has nothing to do with anything really.

You have another fault on the PSU. Check the main smoothing capacitor first. Thats the big one in your close up photo.

If you are on a 110V supply you should read around 150VDC across it when the PSU is on.
On a 220V supply you should read around 300VDC across it when the PSU is on.

Start with that and I will walk you through the rest.