Hello,
I am planning on building a pair of ZV9 amps. Please suggest possible ways to dissipate all that heat (>100W/channel?).
I do have 4 pieces of Conrad MF30-75 which is rated at 0.37 C/W, which if memory serves me Nelson Pass had written somewhere is too low, I think his recommendation was less than 0.28 C/W
Steenoe, has used heatsinks which are substantially bigger and indicated that those were running hotter than he wanted it to. Reading that has me concerned.
What do you all think of maybe using a PC fan running at a low RPM on top of the Conrad heat sinks?
I considered using water / liquid cooled systems used for PCs, but at $300 / piece, it will be very costly.
Please suggest ways for cooling these amps.
Thanks,
Dinesh
I am planning on building a pair of ZV9 amps. Please suggest possible ways to dissipate all that heat (>100W/channel?).
I do have 4 pieces of Conrad MF30-75 which is rated at 0.37 C/W, which if memory serves me Nelson Pass had written somewhere is too low, I think his recommendation was less than 0.28 C/W
Steenoe, has used heatsinks which are substantially bigger and indicated that those were running hotter than he wanted it to. Reading that has me concerned.
What do you all think of maybe using a PC fan running at a low RPM on top of the Conrad heat sinks?
I considered using water / liquid cooled systems used for PCs, but at $300 / piece, it will be very costly.
Please suggest ways for cooling these amps.
Thanks,
Dinesh
Hello Dinesh,
Besure to take the Conrad C/W rating and multiply it by 1.4 to correct it for room temperature.
I have used several www.conradheatsinks.com with great success. I really like their power coating rather than anodizing. They won't turn purple over time.
-David
Besure to take the Conrad C/W rating and multiply it by 1.4 to correct it for room temperature.
I have used several www.conradheatsinks.com with great success. I really like their power coating rather than anodizing. They won't turn purple over time.
-David
here is the url to a thread in the solid state forum on water cooling an Alpha
http://www.diyaudio.com/forums/showthread.php?s=&threadid=190&perpage=10&pagenumber=1
http://www.diyaudio.com/forums/showthread.php?s=&threadid=190&perpage=10&pagenumber=1
AndrewT said:
Andrew,
Thank you for the reply, given that the ambient temparature is usually 24C (75F), I would like to keep it under 40C (100F). Definitely under 60C (140F). Is that a good target?
How does one figure out how much the fan can help?
dw8083 said:
David,
Thanks for pointing that out, it means that I need to shoot for even better C/W rating.
Nice to know, I have good underrated heatsinks
Hi,
for Tc=40degC the temp difference from ambient to sink is less than 40-25~=12Cdeg. You cannot afford heatsinks this big. the temp correction factor is off the scale, probably more than two times.
For Tc=60degC the temp difference from ambient to sink is less than 60-25~=32Cdeg. the correction factor is ~=1.3times. This is becoming manageable. Note that you need to calculate the case to sink differential using Rth c-s for the power and surface area and isolator you are using.
BTW it is not a room temperature correction factor, it is a ambient to sink temperature differential correction factor. Conrad use 80Cdeg as their standard whereas most manufacturers use 70Cdeg as the standard. It makes the Conrad sink look a little bit better in the specmanship stakes.
To make some sense of the temperatures and sink size, I suggest you should look at three senarios.
1. quiescent conditions in mid summer.
2. loud, continuous music with average output 10db below peak output again in mid summer.
2. maximum power dissipation (about -6db to -3db) giving peak dissipation for very short term peak loading. This one is quite difficult for us amateurs to calculate due to thermal inertia preventing stable equilibrium conditions to be achieved. Most would ignore this third senario, I do not rely on it.
for Tc=40degC the temp difference from ambient to sink is less than 40-25~=12Cdeg. You cannot afford heatsinks this big. the temp correction factor is off the scale, probably more than two times.
For Tc=60degC the temp difference from ambient to sink is less than 60-25~=32Cdeg. the correction factor is ~=1.3times. This is becoming manageable. Note that you need to calculate the case to sink differential using Rth c-s for the power and surface area and isolator you are using.
BTW it is not a room temperature correction factor, it is a ambient to sink temperature differential correction factor. Conrad use 80Cdeg as their standard whereas most manufacturers use 70Cdeg as the standard. It makes the Conrad sink look a little bit better in the specmanship stakes.
To make some sense of the temperatures and sink size, I suggest you should look at three senarios.
1. quiescent conditions in mid summer.
2. loud, continuous music with average output 10db below peak output again in mid summer.
2. maximum power dissipation (about -6db to -3db) giving peak dissipation for very short term peak loading. This one is quite difficult for us amateurs to calculate due to thermal inertia preventing stable equilibrium conditions to be achieved. Most would ignore this third senario, I do not rely on it.
Thanks BlueMartini for pointing out that thread. If I were to try water cooling it would be something that does not rely on any moving parts, like some of the PC liquid coolers. The problem with products designed for PCs is that the surface area available to mount devices is very small since they are usually designed to cool one chip, the CPU.
Andrew,
That is very informative. I know my Conrads are too underrated for this application, but can you suggest any heatsinks that might be usable. Digikey has this option but at $120 it is expensive
http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?Ref=15011&Row=126221&Site=US
Regards,
Dinesh
Andrew,
That is very informative. I know my Conrads are too underrated for this application, but can you suggest any heatsinks that might be usable. Digikey has this option but at $120 it is expensive
http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?Ref=15011&Row=126221&Site=US
Regards,
Dinesh
Hi Dinesh,
Since you have 4 of these heatsinks, perhaps you can
build monoblocks with them? That just might be enough.
Then there's always fan-cooling. This is interesting:
http://www.aavidthermalloy.com/cgi-...ngth=75&LengthUnits=mm&SButton1=Change+Length
This shows a fairly generic flatback heatsink not too dissimilar
to yours. If you look at the two graphs at the bottom,
you can see the dramatic change in simulated thermal
resistence as a function of airflow rate.
Hope this helps.
Dennis
Since you have 4 of these heatsinks, perhaps you can
build monoblocks with them? That just might be enough.
Then there's always fan-cooling. This is interesting:
http://www.aavidthermalloy.com/cgi-...ngth=75&LengthUnits=mm&SButton1=Change+Length
This shows a fairly generic flatback heatsink not too dissimilar
to yours. If you look at the two graphs at the bottom,
you can see the dramatic change in simulated thermal
resistence as a function of airflow rate.
Hope this helps.
Dennis
Dennis,
Yes I always wanted to build mono blocks I think you have convinced me. I will start with 1 heat sink /channel. Look at how hot they get and then distribute the FETs among 2 as needed.
Thanks for pointing me to that site. What's cool is that those graphs are interactive.
Regards,
Dinesh
Yes I always wanted to build mono blocks I think you have convinced me. I will start with 1 heat sink /channel. Look at how hot they get and then distribute the FETs among 2 as needed.
Thanks for pointing me to that site. What's cool is that those graphs are interactive.
Regards,
Dinesh
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.