I have not read the posts from 2 to 19.
A heatsink is usually rated (Rth s-a) for a deltaT around 70 to 80Cdegrees.
If you want to use the heatsink at a deltaT of 30dgrees (or any value different from the manufacturer's specified figure) then you must apply the manufacturer's de-rating for the new deltaT.
For a normal 75C deltaT being used at 30C then expect a De-rating Factor around 1.2 to 1.4.
i.e. if you need 0.3C/W, then divide by the DF to get an effective Rth s-a(30C) of 0.3/1.3 = 0.23C/W.
Now look up the manufacturer's datasheets to find a sink with 0.23C/W
Rth s-a ONLY works at the manufacturer's specified DeltaT.
The user must de-rate for other temperatures.
A heatsink is usually rated (Rth s-a) for a deltaT around 70 to 80Cdegrees.
If you want to use the heatsink at a deltaT of 30dgrees (or any value different from the manufacturer's specified figure) then you must apply the manufacturer's de-rating for the new deltaT.
For a normal 75C deltaT being used at 30C then expect a De-rating Factor around 1.2 to 1.4.
i.e. if you need 0.3C/W, then divide by the DF to get an effective Rth s-a(30C) of 0.3/1.3 = 0.23C/W.
Now look up the manufacturer's datasheets to find a sink with 0.23C/W
Rth s-a ONLY works at the manufacturer's specified DeltaT.
The user must de-rate for other temperatures.
I can give one more real world example. A class A amp I use has 4 output devices on each 8" piece of heatsink usa 10.08" heatsink. Bias is 1 amp per device at 23 volts. Heatsinks get hot but not untouchable.
You could build with the sink you "feel" is OK, start with the bias a bit low and adjust up untill the temp of the devices is as hot as you can stand.
If fans are an option here is a simple thermostat that works well. (sorry for the crude drawing).
Evan
You could build with the sink you "feel" is OK, start with the bias a bit low and adjust up untill the temp of the devices is as hot as you can stand.
If fans are an option here is a simple thermostat that works well. (sorry for the crude drawing).
Evan
Hi evanc, thank you very much for the suggestions.
I'm a little tired of calculating c/w (more I do, more I confused...), so I decided to order one 11" x 10.08. I'll make 6 channels of Aleph J (multi amp configuration), so after I confirm it's OK, Ill order the rest.
I'm in 2 hors form NYC, 2 min from the beach (really). It's very cold today...
I'm a little tired of calculating c/w (more I do, more I confused...), so I decided to order one 11" x 10.08. I'll make 6 channels of Aleph J (multi amp configuration), so after I confirm it's OK, Ill order the rest.
I'm in 2 hors form NYC, 2 min from the beach (really). It's very cold today...
Last edited:
OK, now I think I can answer my own question.
I ended up buying 'unknown spec' surplus heat sinks on Ebay for cheap, and did calculation with AAVID online calculator. I built Aleph J based on that calculation, and the actual temp is not far from the number calculated by AADID. I can say AAVID calculator is pretty reliable.
I ended up buying 'unknown spec' surplus heat sinks on Ebay for cheap, and did calculation with AAVID online calculator. I built Aleph J based on that calculation, and the actual temp is not far from the number calculated by AADID. I can say AAVID calculator is pretty reliable.
Is this the 6channel amp you posted?
600W of total dissipation from those four heatsinks?
Yes, actually 6 heat sinks.
Each one is larger dimension than 4U chassis and twice as deep as 4U (80mm) and it has serrated fins. 4U chassis is actually longer width and shorter height, so it can dissipate more for the same dimension of higher heat sink like mine according to AAVID calculator.
Taller is slightly better.A heatsinks' height dimension increase is allegedly 4X more efficient/effective at dissipation than a similar dimension increase in heatsink width.
Taller = Gooder
the usual guide is that Rth s-a decreases in proportion to the square root of the height increase.
i.e. double the height and Rth s-a goes down by the factor of sqrt(2).
Whereas if you double the width the Rth s-a halves, if the backplate can be at the same, or similar temperature.
The wide is much more efficient at dissipating heat than the tall.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.