Thermal time constants are very long compared to audio frequencies. And by orders of magnitude. Whatever magazine published an article claiming that this is something significant in audio must have either had a very stupid editor or a real need to fill space.
VCR is still specified. For example, the Yageo metal oxide resistors that I use a lot are spec'd at 0.001%/V. The corresponding Panasonic resistors come in at 25ppm/V. Carbon films from Yageo do better, with a 5ppm/V. And so on. All these numbers are out of the Digikey catalog. I am suspicious of resistors that don't have this specified.
VCR is still specified. For example, the Yageo metal oxide resistors that I use a lot are spec'd at 0.001%/V. The corresponding Panasonic resistors come in at 25ppm/V. Carbon films from Yageo do better, with a 5ppm/V. And so on. All these numbers are out of the Digikey catalog. I am suspicious of resistors that don't have this specified.
Hi SY,
Carver Lightstars can burn out the feedback resistor. I have a board with a hole where this part used to be.
Modulated it right out of existance!
Carbon comps are bad for voltage coefficients. I definitely want to see this figure before I buy a part (I use carbon comps sparingly, RF stuff). Ceramic caps may do a similar thing.
-Chris
Carver Lightstars can burn out the feedback resistor. I have a board with a hole where this part used to be.
Modulated it right out of existance!Carbon comps are bad for voltage coefficients. I definitely want to see this figure before I buy a part (I use carbon comps sparingly, RF stuff). Ceramic caps may do a similar thing.
-Chris
I'm astonished a fb resistor was burned out by modulation. Would this not have occurred because the amp locked to the rail and the input stage protecting diodes took the fb node end to ground? If you had a 75V rail, this would be close to 0.6W continuous dissipation, on a 1/4W resistor it would undoubtedly burn it out......
Otherwise I find thermal issues hard to credit with resistors in the fb loop. A 0.2g die on an input transistor or VAS could be a different story; that would heat up and down with very low frequencies.
Cheers,
Hugh
Otherwise I find thermal issues hard to credit with resistors in the fb loop. A 0.2g die on an input transistor or VAS could be a different story; that would heat up and down with very low frequencies.
Cheers,
Hugh
Hi Hugh,
As the rails on high power amplifiers get up there, odd stuff starts to happen. Failure modes begin to remind one of tube circuits. The voltage rating of a resistor starts to become a concern for instance. A surface mount version would tend to keep me up at night!
-Chris
Nope. + / - 125 V rails and a reviewer being an idiot. True, the resistor was 1/2 W (metal film of course) and should have been larger. Cost accounting strikes again. Therefore we had to do it in the field at a much greater cost. Only lost a couple that way. I personally think two 1 W metal films in parallel ought to have been designed in.
As the rails on high power amplifiers get up there, odd stuff starts to happen. Failure modes begin to remind one of tube circuits. The voltage rating of a resistor starts to become a concern for instance. A surface mount version would tend to keep me up at night!
-Chris
Thanks Chris,
Ah, I get it!! If we have +/-125V rails, 90% rail efficiency, we have almost 225Vpp across the feedback resistor, which is 632mW in a half watt 10K. This would eventually lead to failure, particularly if the resistor was not well ventilated and the music was very busy for long periods.....
I always rate to no more than 250mW in these cases, just under half. In this situation the half watt resistors last for decades.
That's 1 serious amplifier, BTW.
Cheers,
Hugh
Ah, I get it!! If we have +/-125V rails, 90% rail efficiency, we have almost 225Vpp across the feedback resistor, which is 632mW in a half watt 10K. This would eventually lead to failure, particularly if the resistor was not well ventilated and the music was very busy for long periods.....
I always rate to no more than 250mW in these cases, just under half. In this situation the half watt resistors last for decades.
That's 1 serious amplifier, BTW.
Cheers,
Hugh
Hi Hugh,
The resistor in question was 9K1.
This amplifier used high frequency "down convertors" to modulate the supply rails. I had one for a while. The ratings were : 300W /ch 8R, 600 W /ch 4R and 1,200 W /ch into 2R. It required two 120V 15 A circuits and was a dual mono design. One of the very best sounding Carver amplifiers. Still, not as good sounding as many others but it would dump current!
The resting voltage on the outputs was only about 13 VDC + and - . Scary energy hanging around waiting to be used. When something went wrong, it could get very dramatic.
If you are curious about anything, you could drop me a line via PM. I don't want to stay OT here.
-Chris
The resistor in question was 9K1.
This amplifier used high frequency "down convertors" to modulate the supply rails. I had one for a while. The ratings were : 300W /ch 8R, 600 W /ch 4R and 1,200 W /ch into 2R. It required two 120V 15 A circuits and was a dual mono design. One of the very best sounding Carver amplifiers. Still, not as good sounding as many others but it would dump current!
The resting voltage on the outputs was only about 13 VDC + and - . Scary energy hanging around waiting to be used. When something went wrong, it could get very dramatic.
If you are curious about anything, you could drop me a line via PM. I don't want to stay OT here.
-Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.