• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Does a high DC offset cause excessive heat?

PROJECT 2501

PROJECT 2501

Active Member
Marantz model 32 amplifier.
DC offset of 89mv on the right channel and noticeably higher heat radiating from the right heatsink.
Any correlation?
Red flag?
 
Register to hide this ad
High bias would/could cause output transistors to generate more heat but, from my limited point of view, DC offset is not related to cause higher heat dissipation from output trannies.

Now, Ak is a discussion forum... many more AK's will chime in to correct my theory.
 
In your speaker coils maybe, not so much for your amplifier.

Your speakers will turn DC into pure heat. If were only talking a few mV here, the coil has sufficient movement to keep it cool.
 
I have about 20mV offset on my Realistic STA-64B which uses the typical quasi-complementary output stage. It doesn't really have an easy way to adjust it as far as I can tell. Anyway, suppose you have an 8 ohm impedance speaker that has a 2 ohm DC resistance. This 20mV offset would be 20mV*20mV/2 ohms = less than 1mW burned in the speaker. If the offset was instead 100mV, then 0.1*0.1/2 = 5mW of power dissipated in the speaker.

On the other hand, after I replaced a fried transistor, the quiescent/bias current draw on one of the channels shot up to 100mA. This calculates to more than 5 watts burn! The heatsink was getting warm after being powered up. After adjusting it closer to the 20mA mark that is suggested in the manual, it dropped the quiescent drain of that channel to a bit more than 1W, which saved 4 watts worth of power burn. Unfortunately (which could be fortunate) I don't have the ear to tell any difference in audio quality.

When idle (output at 0V+offset) the dissipation in the transistors are related to the quiescent/bias current. This has no bearing whether the speaker is connected or not (well, it does, slightly, but in engineering terms it's negligible.) Thus, that 5+ watts burn will be burned whether a speaker is connected or not.

I haven't checked how warm the heatsink gets now after reducing heating from ~7W (both channels total) to ~2.5W, but I figure that it won't get as warm anymore while quiet. Sort of trying to set up a Sankey diagram for it for curiosity's sake.

17W in (quiescent)
2.5W -> final output transistor bias
??W -> remainder of power amplifier draw
5W -> two pilot lamps (two burned out, so this should be ~10W)
5W (guess) -> waste by power transformer getting warm (eddy currents or slight saturation mismatch?)
??W -> radio receiver power
??W -> preamplifiers
??mW -> speaker offset power
 
Check the pcb. Some Realistic amps, probably most have a voltage adjustment trim pot for each channel in the amp section. Maybe they need adjusting? I have several Realistic I integrated amps. None of them have bias or DC offset adjudtments. Just the two voltage adjustments. You may have to adjust the DC offset by changing out different value fixed resistors instead of having trimmers.
 
audiodummy said:
I have about 20mV offset on my Realistic STA-64B which uses the typical quasi-complementary output stage. It doesn't really have an easy way to adjust it as far as I can tell. Anyway, suppose you have an 8 ohm impedance speaker that has a 2 ohm DC resistance. This 20mV offset would be 20mV*20mV/2 ohms = less than 1mW burned in the speaker. If the offset was instead 100mV, then 0.1*0.1/2 = 5mW of power dissipated in the speaker.

On the other hand, after I replaced a fried transistor, the quiescent/bias current draw on one of the channels shot up to 100mA. This calculates to more than 5 watts burn! The heatsink was getting warm after being powered up. After adjusting it closer to the 20mA mark that is suggested in the manual, it dropped the quiescent drain of that channel to a bit more than 1W, which saved 4 watts worth of power burn. Unfortunately (which could be fortunate) I don't have the ear to tell any difference in audio quality.

When idle (output at 0V+offset) the dissipation in the transistors are related to the quiescent/bias current. This has no bearing whether the speaker is connected or not (well, it does, slightly, but in engineering terms it's negligible.) Thus, that 5+ watts burn will be burned whether a speaker is connected or not.

I haven't checked how warm the heatsink gets now after reducing heating from ~7W (both channels total) to ~2.5W, but I figure that it won't get as warm anymore while quiet. Sort of trying to set up a Sankey diagram for it for curiosity's sake.

17W in (quiescent)
2.5W -> final output transistor bias
??W -> remainder of power amplifier draw
5W -> two pilot lamps (two burned out, so this should be ~10W)
5W (guess) -> waste by power transformer getting warm (eddy currents or slight saturation mismatch?)
??W -> radio receiver power
??W -> preamplifiers
??mW -> speaker offset power
Click to expand...

Thanks for the reply but it went way over my head
 
Short answer = no. DC offset is not the source of your heat...but at 89mV, it suggests a problem in the right channel amp. As suggested earlier, check bias on that channel.
 
20mV offset is usually not enough to warrant changes. For my STA-64B it mentions in the service manual which resistor needs to be changed but, it's well within spec. But even at 100mV offset it's probably not doing your speakers any damage. However a high offset will reduces your THD window as the higher it is, the less of an operating range your transistors have before it hits rail. I'd only worry about it if it gets to the volt range where you're no longer dissipating milliwatts of power, or you're losing a significant number of music power watts because of harmonic noise increase. This is a problem for lower wattage amplifiers like headphone amplifiers and small speakers.

Voltage output offset specifically means that heat is being dissipated in the speaker versus the transistors. Quick way to tell is to disconnect the speakers, and whether the heat is still being dissipated in the transistors?

The bias current fix on my STA-64B by tweaking the switch point seems to have reduced the heat generation from the output transistors... after a few hours of use, before the vents got a little warm. Now after the fix, the vents stay cool after hours of use. Glad there was a post to tell me I should look at it! Now I wonder if there's similar for my Denon that has a considerable quiescent power draw...
 
You must log in or register to reply here.
Back
Top Bottom