My PCB is mounted on 10mm or 3/8 inch 4-40 metal standoffs directly near the top of the heatsinks. The FET's are on the bottom side of the PCB pointing downwards.
That way my P1 P2 and P3 are easily accessible and PCBs are further away from PSU components.
That way my P1 P2 and P3 are easily accessible and PCBs are further away from PSU components.
Why? I'd imagine the opposite since heat tends to rise.
My concern was if (I doubt) the PCB would get hotter when the FETS are below....
My concern was if (I doubt) the PCB would get hotter when the FETS are below....
mfrimu, you have it arranged for the best efficiency of the heat sinks. The little bit of heating that the board gets from being above the output devices is insignificant.
As of curiosity, why devices at top would be better for heat distribution?
( PS: My devices are a bit lower than half way, not at the bottom... )
( PS: My devices are a bit lower than half way, not at the bottom... )
Last edited:
I think I've read that somewhere in Papa's papers, long ago. 😀
As this was counterintuitive to me as well I assumed it was tested and tried to explain it to myself by the better airflow you get if the center of heat is higher and the hottest air leaves the heatsink as soon as possible.
Right?
As this was counterintuitive to me as well I assumed it was tested and tried to explain it to myself by the better airflow you get if the center of heat is higher and the hottest air leaves the heatsink as soon as possible.
Right?
if devices are situated in one row on heatsink , they need to be horizontally on lower third
warum?
darum!
warum?
darum!
that's plain logic and physics , which I don't need to prove to anyone
plenty of examples all over the board
just think of temperature interaction of vertically aligned heatsink and airflow and you'll get it
plenty of examples all over the board
just think of temperature interaction of vertically aligned heatsink and airflow and you'll get it
both no explanations
convection also depends on height difference of cold and warm area
talking of physics it should be clear that airflow is not so very simple physics
Last edited:
well , I'm lazy to do your homework - chasing nice and illustrative graphs , and posting them here
believe me or not , your choice
go to any proper heatsink mnfctrr site and you'll find informative reading
for instance - heatsink design tools
believe me or not , your choice
go to any proper heatsink mnfctrr site and you'll find informative reading
for instance - heatsink design tools
Mounting output devices in the top region of the sinks will render the the lower part of the sinks less to non-effective. Wich leaves the upper region to run hotter. Hotter sinks=hotter outputs.
The baseplate thickness also plays a role here. Effective heatspreading distance around the outputs is determined by the baseplate thickness.
The baseplate thickness also plays a role here. Effective heatspreading distance around the outputs is determined by the baseplate thickness.
I've built the F5T with the boards mounted Fet side down and with Fet side up. I haven't seen much difference in operation and device heat.
Logic says to put your hot things lower on the heatsink, as heat tends to rise, and the cooling air will be at it's lowest temperature there. But, in what little I have done, there seems to be almost no difference in practice.
Logic says to put your hot things lower on the heatsink, as heat tends to rise, and the cooling air will be at it's lowest temperature there. But, in what little I have done, there seems to be almost no difference in practice.
Why? Heatspread in the sink is omnidirectional.
So it' all about airflow. Probably there is better airflow by setting the hottest point higher and it outweighs the effect of cooler air reaching the hottest point first.
When heating at the bottom you will heat up the upper sink area by the heated air coming from the bottom, or at least you have less cooling in the upper area.
How to decide? By trying, not by "simple logic" which can be just wrong intuition.
Zen Mod,
Maybe I can assist.
dieringe,
R-Theta has a nice modeling tool on their web site. Follow the link below and you can interactively simulate different locations for the devices and see the results.
Heatsinks and Cooling of Power Electronics Solutions
no you did not. go back and read the last line. there is something about baseplate thickness there.
Yes sorry I did not. But I don't understand how the baseplate thickness could decide about the problem. If you had instantaneous heat spread it wouldn't matter at all where to mount.