An externally hosted image should be here but it was not working when we last tested it.
During an Aleph 4 build it was necessary to reduce the number of heatsinks on the rear side to one instead of the two on the other sides - 7 in total.
Question is, why is the heat not transferred from the back panel to the heatsinks on the sides ?
All alluminium plate and angles are polished on their adjoining surfaces and smeared with a very light coating of heat transfer compound.
The heatsink at the back gets warm but the one on the left remains stubbornly cool.
How could I improve the sharing of the heatsinks ?
One would assume - Obviously incorrectly.
That the MOS-FETs on the rear plate would heat up the 10mm alluminium rear plate. SO FAR SO GOOD, the rear plate gets HOT.
That the LEFT HAND alluminium bracket would act like a normal heatsink bracket.
Even though this is tightly secured against the back panel and is thermally tied to it, as it is to the left hand panel, it doesnt seem to transfer any heat to the Left Hand panel.
??????????????
Would a heat pipe help in any way ??
That the MOS-FETs on the rear plate would heat up the 10mm alluminium rear plate. SO FAR SO GOOD, the rear plate gets HOT.
That the LEFT HAND alluminium bracket would act like a normal heatsink bracket.
Even though this is tightly secured against the back panel and is thermally tied to it, as it is to the left hand panel, it doesnt seem to transfer any heat to the Left Hand panel.
??????????????
Would a heat pipe help in any way ??
The standard rule I have applied is that the backplate of the heatsink should be ~ 1/10 the maximum radius from hot device to extreme corner of effective heatsinking area.
Your single heatsink could be 200mm wide by 120mm high by 8mm thick backplate with 40mm by 4mm thick fins.
The fins have a 10:1 ratio for good heatspreading ability.
The 8mm thick backplate will have good heatspreading ability if the largest radius from hot device to extreme corner is <=80mm.
If you add on, via a perfect interface, but increase the spreading distance to 160mm then the ratio has moved to 20:1. If the add on heatsink is 200mm wide then the total spreading distance from the nearest hot device is ~250mm. The ratio is now ~30:1.
The distant heatsink can never be an effective radiating area. It is simply too cold due to too much thermal resistance.
Your single heatsink could be 200mm wide by 120mm high by 8mm thick backplate with 40mm by 4mm thick fins.
The fins have a 10:1 ratio for good heatspreading ability.
The 8mm thick backplate will have good heatspreading ability if the largest radius from hot device to extreme corner is <=80mm.
If you add on, via a perfect interface, but increase the spreading distance to 160mm then the ratio has moved to 20:1. If the add on heatsink is 200mm wide then the total spreading distance from the nearest hot device is ~250mm. The ratio is now ~30:1.
The distant heatsink can never be an effective radiating area. It is simply too cold due to too much thermal resistance.
So my total backplane thickness, Heatsink + Alluminium Plate is 20mm thick. I would have expected some heat transfer through the corner bracket to the adjoining heatsink.
Again my question. Would a heatpipe help to conduct the heat away from the poorer heatsink on the rear panel. If YES, how are they best employed ?
Again my question. Would a heatpipe help to conduct the heat away from the poorer heatsink on the rear panel. If YES, how are they best employed ?
Attachments
Last edited:
Thermogen does not flow good through narrow objects...
It's highly ulikely to get proper mating between the parts. Somewhat 3-5um roughness and 30um per 100mm straightness/warpage. If you insist on perpendicularity of parts, try T connection, i.e. shortening the path for thermogen
You may search for hockey puck diodes/triac heatsinks which are probably the best shot without going beserk
Sure heatpipe will tranfer the heat if (again) properly implemented. Have not seen them on avail, though... besides proc. coolers
Attachments
If you have an access to milling shop, I would suggest to make an additional heatsink of yours u-shaped to increase mating surface. I mean bend the plate and mill the mating surfaces.
You could make some heat pipes yourself. All you need is some copper tubing to make the heat pipes ans access to machine shop to machine some channels in the heat-sinks and plates.
Hers is guy who made some of his own heat pipes.
Benchtest.Com - Heat Pipe 1
Let us know how it turns out.
Hers is guy who made some of his own heat pipes.
Benchtest.Com - Heat Pipe 1
Let us know how it turns out.
Oh, well...
Decent ebm papst fans' prices are on par with heatsink ones (brand new). Set of two capsule heatsinks is around 200$. Even more for fan(s). So what seems to be a solution?
I suggested used ones from flea market...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.