Page 1 of 1

Kelly Controller Cooling - Air, oil and water...

Posted: Jun 04 2013 12:52pm
by h0tr0d
I've done some simulations to increase the "survivability" of my Kelly @ 350A.

Bought this one: http://kellycontroller.com/kbl96351e24- ... -1258.html
350A for 30s
140A continuous

I need about 350A for around a minute, so my thermal engineering knowledge comes in handy.

Simulations (60s @ 1200W losses in the MosFET's and 250W @ caps, 30s no heat):
  • Original controller = OC
    OC + two 80mm fans and lots of holes
    OC filled with water (after dipping the controller in varnish)
    Controller with copper bars glued to the alu angle profiles inside the controller = EC
    EC + Oil
    EC+ Water
Some results in ÂșC = Y axis, "measured" at the Mosfet-aluminium surface (the Mosfet's internal temperature will significantly higher)
(please ignore the X axis, temperature peak is at 60s):
calor.jpg
calor.jpg (82.78 KiB) Viewed 973 times
kelly original.jpg
kelly original.jpg (98.64 KiB) Viewed 1047 times
kelly copper.jpg
kelly copper.jpg (92.49 KiB) Viewed 1047 times
Enjoy and tell me your thoughts...

Re: Kelly Controller Cooling - Air, oil and water...

Posted: Jun 05 2013 12:37am
by bowlofsalad
By extra copper+oil, do you mean there is oil in the extra copper (tubes) or oil within the controller?

Have you considered looking into PC cooling parts? An external heatsink with lots of fins, a fan or two, and some kind of fluid loop might be something worth considering if your current ideas don't fulfill your needs.

I am having troubling finding it, but there might be something more you can do to help deal with the heat. A more efficient mosfet is something I read about on this forum. The post was made by the user 'Jeremy Harris'.

Ah, after some searching, http://endless-sphere.com/forums/viewto ... nt#p629482 here is the post. I don't know if it would apply to your situation, but you get the idea.

Re: Kelly Controller Cooling - Air, oil and water...

Posted: Jun 05 2013 5:35am
by h0tr0d
bowlofsalad wrote:By extra copper+oil, do you mean there is oil in the extra copper (tubes) or oil within the controller?

Look at the Kelly copper pic above, just copper added inside the controller to increase thermal inertia.
Have you considered looking into PC cooling parts? An external heatsink with lots of fins, a fan or two, and some kind of fluid loop might be something worth considering if your current ideas don't fulfill your needs.
PayPal I am having troubling finding it, but there might be something more you can do to help deal with the heat. A more efficient mosfet is something I read about on this forum. The post was made by the user 'Jeremy Harris'.

Ah, after some searching, http://endless-sphere.com/forums/viewto ... nt#p629482 here is the post. I don't know if it would apply to your situation, but you get the idea.
Have thought about PC cooling stuff but no need in my case.

What Jeremy was talking is mosfets with a lower RDSon. In To220 there's nothing on the market with lower RDSon then the 075015N mosfets that the Kelly has.

Re: Kelly Controller Cooling - Air, oil and water...

Posted: Jun 05 2013 7:47am
by spinningmagnets
I'd recommend starting with air-cooling on a pre-determined loading profile, and log the results with a temp probe. If the results are good, you need go no further, but you could move on to oil cooling after and log the actually difference in the better effect to see if it's worth it?

Two fans: one pushing, one pulling, plus a dust filter on the inlet.

Re: Kelly Controller Cooling - Air, oil and water...

Posted: Jun 05 2013 7:58am
by h0tr0d
spinningmagnets wrote:I'd recommend starting with air-cooling on a pre-determined loading profile, and log the results with a temp probe. If the results are good, you need go no further, but you could move on to oil cooling after and log the actually difference in the better effect to see if it's worth it?

Two fans: one pushing, one pulling, plus a dust filter on the inlet.
I've Edited the first post so that everyone can see my results.
The difference between air cooled and water filled is SO great, I have to try it! :D

Re: Kelly Controller Cooling - Air, oil and water...

Posted: Jun 05 2013 10:07am
by JEB
My 2 cents, years ago I built a electric golf car controller, I direct coupled the mosfets to a much larger block , (10 times~ the contact area of the to220's,) which I then used a electrical insulator heat transfer material to isolate that block to a larger heat sink, then used a pressure type blower to get under the surface insulating air film for cooling.

Re: Kelly Controller Cooling - Air, oil and water...

Posted: Jun 05 2013 10:15am
by h0tr0d
JEB wrote:My 2 cents, years ago I built a electric golf car controller, I direct coupled the mosfets to a much larger block , (10 times~ the contact area of the to220's,) which I then used a electrical insulator heat transfer material to isolate that block to a larger heat sink, then used a pressure type blower to get under the surface insulating air film for cooling.
That sounds a lot more work than my idea of additional copper, dipping in varnish and filling the case with water... But a valid suggestion indeed!