So you warm up your amp for an hour, you get 50 degrees Celsius on the heat sink, assuming this is the ideal temperature for the amp to work at its best, Now, you turn on your fans and the heat sink drops to to 35 degrees Celsius, Does your amp still retain the same sound quality when it was at 50 degrees Celsius?
We have some OCD members...
Reptile tank turntable heaters, floppy drive motors for turntables, heat sink color and polish and now .....
Seriously, stay about max. 10 to 15 degrees above ambient.
Capacitors are mostly 85 or 105 degrees, and semiconductors you can go to above 100 degrees.
LCD will fail above 70 degrees.
So stay below that range.
And you can always buy loads of test equipment to try out your theories.
I would suggest listening to the same music at different heat sink temperatures, then tell is all your results.
Buy a PT100 probe, use it on a good DMM, or buy a nice thermometer to measure the temperature.
Also, decide how you will control the heat sink temperature while testing.
Have fun...
Reptile tank turntable heaters, floppy drive motors for turntables, heat sink color and polish and now .....
Seriously, stay about max. 10 to 15 degrees above ambient.
Capacitors are mostly 85 or 105 degrees, and semiconductors you can go to above 100 degrees.
LCD will fail above 70 degrees.
So stay below that range.
And you can always buy loads of test equipment to try out your theories.
I would suggest listening to the same music at different heat sink temperatures, then tell is all your results.
Buy a PT100 probe, use it on a good DMM, or buy a nice thermometer to measure the temperature.
Also, decide how you will control the heat sink temperature while testing.
Have fun...
Last edited:
Very early amps, say about 1966, often had inadequate idle current on the push pull class AB silicon output stage. That made crossover distortion, which didn't matter in germanium and really did in silicon. These amps could grow more idle current when warmed up for an hour and sound a lot better at low wattage. Thermal runaway was a real problem in 1966; read about SWTC a market leader.
Proper output transistor bias arrangements have been understood since about 1972, and now the patents are expired. Some designs still didn't pay attention. Some PA amps or instrument amps (guitar, bass) were never marketed to be used at less than 80% of full wattage, and sounded bad below that. These also could use a warmup at lower volumes to grow adequate output transistor idle current.
Last edited:
those PA amps had spec barely acceptable at max power (lowest distortion), but terrible crossover distortion, hence distortion was rising with decreasing power
no amount of warm up would change that
no amount of warm up would change that
You’ll get lower noise if you cool the input stage. 10K is doable, but requires a reasonable amount of work.
Actually, some of the modifications done to gaming computers, like Peltier effect heat pumps, liquid or refrigerant cooling for heat sinks and so on may have some effect on the performance of audio amplifiers.
These are more doable than the ones suggested above.
Worth investigating ?
Eye Roll from me at least...
These are more doable than the ones suggested above.
Worth investigating ?
Eye Roll from me at least...
Last edited:
The electron speeds might go down as you go towards absolute zero, interesting topic.
If really interested, look on the net for more data, particularly as used for deep space probes.
If really interested, look on the net for more data, particularly as used for deep space probes.
Strange for me, that nobody remembers that the BJT's hfe is increasing with the temperature. So higher temperature means more current gain, what can change lot of parameters in any amplifier.
Thermal management of an audio amplifier is an important thing. The best way to do it is to avoid too much heat. That's why I don't like class A power amplifiers so much. In home audio you usually need not much more than an output power of 10 to 20 W per channel. Theoretically your class AB amplifier can achieve an efficiency or COP of 75%.
In reality, depending on your circuit design, you will have reject half (COP 67%) to full (COP 50%) output power as heat. I have found, the best way of heat rejection is to avoid bringing it into your amplifier housing. So I used to attach the heat sinks of the amplifier at the back side of the amplifier housing with the cooling area in vertical direction.
In reality, depending on your circuit design, you will have reject half (COP 67%) to full (COP 50%) output power as heat. I have found, the best way of heat rejection is to avoid bringing it into your amplifier housing. So I used to attach the heat sinks of the amplifier at the back side of the amplifier housing with the cooling area in vertical direction.
But it also means less Vbe. So you have to limit the temperature in order to keep it in the regulation interval for your quiescent current.