noob question - heat vs. output impedance in ss amp

[g-man]

This has probably been covered somewhere but I didn't find an answer after doing a quick search.

Should the heatsink in a SS amp feel hotter or cooler running a 4 ohm load vs. 8 ohm load? My SWR Workingman 4004 can be run at either impedance (specs say 400W @ 4 Ohm or 260W @ 8 Ohm). I haven't determined this for certain but it seems like mine gets hotter when running a 8 Ohm cab vs. a 4 Ohm cab. Also, and maybe this is a dumb question, but how hot to the touch should a heat sink get before you would expect it might indicate some problem?

Thanks.

[Jazz P Bass]

You really need to quantify your question.
On a test bench at a prescribed output, into a prescribed load, the power that the transistors are dissipating should be what is going to the speakers.
In other words, are you simply feeling the heat sink.
A temperature reading would be nice.
Are you running the amplifier at full power at 4 & also at 8 ohms?

[J M Fahey]

It should be hotter running into 4 ohms, because it works harder.
Hands are very unreliable temperature testers, yesterday I killed my last batch of home made yoghourt because I *felt* it just right, while it was over 42ºC ; I needed 37ºC (human body) temperature.
That said, if you can stand grabbing the heatsink , obviously with your bare hands, for a couple of minutes, it's fine.
If you can't stand it, it's over 70ºC , start to worry.
If a drop of saliva or other suitable liquid hisses .....

[R.G.]

Things get tricky.

The amp can make an output voltage which is limited by the power supply, and the losses of the parts that conduct the heavy currents. These include the power filter caps (voltage sags under heavy loading), the wires to the power transistors, the power transistors themselves, the wires and connectors to the speakers. For a fixed output voltage, a 4 ohm resistive load will pull twice the current that an 8 would. So you often see people quoting twice the output power into 4 ohms. In reality, it won't get to 2X. The 400W and 260W are reasonable.

What the power transistors dissipate is different from the power output. The power transistors dissipate the sum of the instantaneous voltage across them and current through them. For Class AB amps, the power transistor dissipation peaks at about 40% of max output *voltage*, because bigger voltages transition them through the high voltage/high current middle of the output signal faster; they spend less time with both big voltage and big current. That's for bench testing with sine waves. The absolute worst thing you can do to a class AB amp is feed it a square wave that puts the output voltage at half the peak it can put out. That forces the output devices to have half the max currents and half the max voltage across them *all the time*. Square waves at full output voltage are actually much lower dissipation, so the heat sinks will cool off as you turn square waves up from half-voltage.

That was a back-handed way to say: the program signal you're feeding it matters. And so do the speakers themselves. Speakers are anything except a resistive load. The more reactive they are, the worse it is for output transistor heating. Then there's the issue of whether you play differently when you have a 4 vs 8 ohm cab on it (if your question is to do with actual playing, not bench testing). Not to mention whether the 4 or 8 ohm cab has more efficient speakers.

The advice you have already is good. You can test it and measure temperatures. But heat sinks will be OK at temps you don't want to keep your hands on. Average humans will not voluntarily keep their fingertips on metal surfaces which are over 130F/55C. So if you can hold your fingertips on the heat sink, it's OK, as noted. It *may* be OK up to about 70C. I can't hold my fingertip on a 70C surface. But then I'm a wimp. 8-)

Many heatsinks are protected with 70C thermal cutouts; I just bought a bunch for my stock. Over 70C, something is getting to be a problem, because the actual transistor chips are approaching 150C in most cases. And absolutely - if a damp finger tip hisses when it touches the sink, shut 'er down.

None of this is answering your actual question. That's because it's not easily answerable without a whole lot more info. Touch the sinks in both cases. If they're hot, but you don't immediately pull your finger off, play on. If it hisses and blisters your finger, you're about to lose an amp unless you do something to cool it off.

[g-man]

Thanks everyone. I haven't tried any actual measurements other than the subjective "touch the heat sink" test. Sometimes it's only warm but other times fairly hot, but I don't think it ever gets so hot you can't touch it for at least several seconds. I thought it might be hotter when using the 8 Ohm cab but it was probably under different playing conditions than when I used the 4 Ohm (I know for sure it was in a different room and probably at a different volume too). Sounds like it's not really an issue but maybe I will try using my DMM with temp probe clipped onto the middle of the heat sink during some extended playing and take some measurements.

[J M Fahey]

maybe I will try using my DMM with temp probe clipped onto the middle of the heat sink during some extended playing and take some measurements.

PLEEEEEEASSE DO and post results here.
We can often make fairly educated guesses , but nothing beats actual measurements.
The ideal way to measure temperature is to drill a small hole in the heatsink, (it will not hurt it), and insert the probe there, so it's basically surrounded by hot metal.
If just pressed against a flat surface, it will be close anyway, but somewhat on the low side.

[gbono]

25C (77F) is "room" temperature so figure that as a reference - if you compare two different temp effects due to load changes - were the "measurements" made at the same ambients?

The power transistors dissipate the sum of the instantaneous voltage across them and current through them.

Don't you mean product?

[R.G.]

Don't you mean product?

Did *I* type *that*??
Yes, product.

What I meant, not what I actually typed.

Product.

[Enzo]

RG, you need to product yourself from these lapses.

[J M Fahey]

RG, you need to product yourself from these lapses.

Maybe it should reed:
"you need to protect your shelf from those lipsticks"

________________________________________________________ ___
"Car full typing is the moist important thing in the World"

[Enzo]

I'd tell you the reel story, but you can't handle the turth.

[J M Fahey]