View Full Version : Is it normal for SS bias to vary with temperature?

Blue Shadow
04-12-2013, 03:37 PM
I opened a new to me Onkyo M-501 power amp. Today I thought that while it was open and I could see a couple VRs on the amp board, I would look at the SM and see what adjustment needed to be done. I turned the amp on and by the time I got back to it a few minutes later it was very hot, no load, no signal just on.

I knew the bias, which is what the VRs were for needed to be adjusted. First side had about 73mV other side a bit less. 205 is spec. I made some adjustments, waited, adjusted, waited and then after doing this a while I realized that the heat sink temp had a major effect on the bias voltage. Just blowing on the heat sinks could change the number a mV or two.

So I let the amp cool and started over, much closer to the correct number. Again some variation, but after making adjustments it would drift up or down I think depending on the direction of the correction. So I let it cool a couple times and adjusted the bias per instructions, after about 5 min. Bias rose from near 0 to operating range over those 5 minutes.

My question is if this is normal operation for some gear. I see how it can be but I have not run into it before...having only done a dozen or so devices. The bias seems to have stabilized in the correct range and the temp of the heat sink has dropped from 125+ to about 108F.

04-12-2013, 03:55 PM
I'm not an amp repair expert, in fact I'm more of an amp repair moron, but I have noticed service manuals with a detailed bias procedure usually involve X signal for Y minutes into Z dummy load before doing the measurement and adjustment, which I take to mean the answer to your question is "yes".

04-12-2013, 04:33 PM
It is very hard to get exact tracking of bias with temperature, so some variation is the norm. My only concern would be to make sure there is a reasonably good connection of the temperature sensitive device - transistor, diode or thermistor - which is mounted on the heat sink. Don't remember which onkyo's, but some had the bias diode mounted (glued) to the center lead (collector) of one of the output transistors - and would blow the fuse if it wasn't affixed. YMMV

04-12-2013, 05:00 PM
I'd say "yes", but it should stabilize at some point.

My last experience adjusting bias in a very very old amp, no service manual available was this:

Adjusted some XX value, let it warm up for 15 minutes, heat sink was warm, Bias current went higher, let another 15 minuter, the amp was hotter, the bias continued increasing, and so, more heat, more current in an infinite loop. I adjusted to a lower value, until get an stable reading with stable temperature.

04-12-2013, 09:15 PM
yeah, it'll change with temperature. Some circuits will actually mount a diode to the heat sink and wire it into the bias circuit as a zener diode. The diode's purpose is to change voltage drop with temperature and attempt to maintain a more stable bias level. If its got those, try replacing the heat sink compound to make sure its getting good thermal transfer. Probably wouldn't hurt to replace the compound on the output transistors as well if its dried out. The hotter they get, the more wonky the bias will get. It turns into a chain reaction too since the bias usually drifts in a way that makes the transistor get hotter.

Blue Shadow
04-12-2013, 09:54 PM
I doubt this amp has any temp sensing control of bias but what do I know. It was interesting chasing this as the adjustments to the correct level lead to a significant deviation a few minutes later as the unit changed temps. Fortunately once the correct bias was close to set the numbers did not change much as the final adjustments were made. I'm guessing it is good to go since it is not getting hot now. Maybe I stuff it into a system tomorrow and give it a listen.

04-12-2013, 10:03 PM
Just to add a bit more info-

Instructions to one of my amps via the manufacturer requests you let the warm up 15 or 20 minutes before testing, in order to get a correct reading. I just read this mere days ago, filing my important hifi papers and manuals.

04-12-2013, 10:45 PM
It's just a Myth (, evidently.

It's the exact same science that debunks these tweaks that also debunks.... amplifier warm up (solid state), and other sonic "differences" that don't hold up to scrutiny.

04-13-2013, 08:46 AM
If you have any small transistor attached to the heat-sink, near the output transistors, that's the temperature sensing control. An schematic will confirm that.

04-13-2013, 08:58 AM
After doing this for close to 40 years, I find the best method is to turn the power on, wait a couple minutes, adjust the bias, monitor for 20 minutes and continue to adjust the bias until it stabilizes. Transistors pass more current as the case temp goes up, so it is normal for the bias to be affected by heat sink temp or ambient temp. On some designs you have to run wires out and leave the cover on to get the proper adjustment, otherwise the amp will warm up with the cover on and the bias will go up. Most amps have a diode or small transistor attached to the heat sink to track the bias. Some attach it to the driver transistors, and those typically don't track as well.

04-14-2013, 06:27 AM
It's just a Myth (, evidently.
That is not what Blue Shadow was asking about. Besides, SS amplifier warm-up can happen, but it is usually a design flaw, a malfunction, or a measurement amplifier needing time to achieve its nominal accuracy.

04-14-2013, 06:39 AM
I'm not an amp repair expert, in fact I'm more of an amp repair moron, but I have noticed service manuals with a detailed bias procedure usually involve X signal for Y minutes into Z dummy load before doing the measurement and adjustment, which I take to mean the answer to your question is "yes".

Yes, that's just an instruction to "pre-condition" or "warm up" the amp prior to bias setting. You don't have to do it with a signal generator and a dummy load, :headscrat if you want to do it this way just play the amplifier for hour with some medium volume program material while checking the heat sink temperature now and again :scratch2:

OR just turn it on and adopt the procedure dr*audio uses :yes: :yes:

Blue Shadow
04-14-2013, 10:44 AM
I did what Warren said, but after I did the cold (72F heat sink temps) a couple times to get closer to the right number. With the high bias causing high temps, adjustments were needed for 10mV changes over a few minutes so turning the amp off and letting it come up to temp slowly smoothed the adjustments to 3mV.

Then let it sit idling for a while and verify the 205mV, well less variation than that.

05-17-2013, 10:50 AM
I had the same question and started another thread when I couldn't find this one. I'm listing the link here so others can find both threads in the future.....