ESC Cooling Concept- Theory and Practice

[questtek]

After reading so much this morning, I will now always use copper to make my ESC cooling plates. Also, there are many cooling fluids that can be used and much better than water. Now I might be able to use the little ice cubes Steven had so not to taint the fluid yet chill it.

PC fluid cooling CPUs is easy to find and good stuff.

Glad all the stuff I wrote on thermal coefficients and heat conduction made a bit of sense to you. Since I mentioned CPU coolers let me give you a practical example. If you go to buy a normal CPU cooler with aluminum heatsink and 12 VDC brushless fan you will pay about $12 retail. You can pick the identical CPU cooler up in many surplus outlets for about $3 NEW. I have a good friend in the surplus business and this is what he sells them for at the local electronic swap meets.

However, if you do a search you will find that a COPPER heat sink CPU cooler will cost between $39 and $49. You can see this if you look at CPU coolers at www.frys.com (our local discount computer retail store in So California). Hummmmmmmmmmm.

If you put on your thinking cap there is a better, easier and cheaper way to cool your now copper heat sink that is mounted to the FETs oy your ESC that the Blue Ice methods I proposed. I even copied one forum members idea and tried to run 1/4" copper tubing inside the 94 cent Blue Ice pac. You can see the picture attached. I put thermistors on the inlet and outlet so I could monitor the output cooling temperature changes once per second for 5 minutes. I was really disappointed since in these tests I was seeing only a 9 degree F temperature difference at best! Not good enough so I had to come up with a new idea....and I did it by accident. This new idea shows real promice but you will have to wait a bit for me to build the system, instrument it then test it. Stay tuned for more unbearable excitment in the world of FE brushless ESC cooling technologies!

[Jeff Wohlt]

Interesting because I used 3 brass tubes and it took water down from 90 to 75. But when I put cooler water it did not bring it down as much.

Certainly when I slowed the flow by adding 3' of silcone tubing the water was cooler coming out. For this set up the coils are the key and flow in the coils and not fast. Easy enough to regulate flow with a small adjuster...like a drip garden inline that can slow the water down if needed...as in a pump method.

[Meniscus]

Most fish pumps I find were air not water but I will look. I found a very small round one that weighs nothing and only $12.

I also want a 6v for slow flow and then just a small pack of 2s would run it for a long time and actually bump it up a bit with flow...

My mistake, I meant a small fountain pump. They have real cheap ones that hardly flow any water, for instance a fountain that spills water over cut bamboo sticks or a rock.

Very cool find on the liquid cooling! If you were to go with a closed system, then there are many more options.

________________________________

I can see you scale guys now, incorporating motor mockups sticking up out of the hatch, tied to copper heat sinks inside. In my case, I'd be too tempted to generate Hydrogen and Oxygen out of water for the flamethrower out the back Or for that mater, to power a steam engine to depressurize and cool the air inside. This is just talk kids

[Jeff Wohlt]

I can find those pumps all day long but they are 120v.

Oh there is a way. I did not know there was a better cooling fluid. Distilled water is great but there is better...but then you do not want to use ice with it like you can with water. You have to use the cooling cubes...or maybe have a seperate place for water and just freeze the entire deal and let the water run around it or something. A block form of ice will last much longer as well.

We are really only talking about enough to stay cold for one heat or sporting around. At this point I can't say how long the cold water will last after removing heat so that is what I will try and test...not fair now with colder weather but worth it.

[domwilson]

I have one of those fountains. It has a "wall wart" that says 1t's output is 12v AC.

[questtek]

Interesting because I used 3 brass tubes and it took water down from 90 to 75. But when I put cooler water it did not bring it down as much.

Certainly when I slowed the flow by adding 3' of silcone tubing the water was cooler coming out. For this set up the coils are the key and flow in the coils and not fast. Easy enough to regulate flow with a small adjuster...like a drip garden inline that can slow the water down if needed...as in a pump method.

You are probably correct. My flow rate was pretty high. I should have measured it. The drip system miniture black plastic valve is a good one to try to control flow. I use those fittings a lot, especially the "T" fittings to put the thermistors in for my Eagle Tree monitorig.

I am off on a new approach. SOmehow I lost sight of the main objective which is to keep the FETS cold as efficienl as you can. Use of water cooling, while pretty good, may not be the only answer. I will post more of the new system I am currently testing.

[questtek]

I can see you scale guys now, incorporating motor mockups sticking up out of the hatch, tied to copper heat sinks inside. In my case, I'd be too tempted to generate Hydrogen and Oxygen out of water for the flamethrower out the back Or for that mater, to power a steam engine to depressurize and cool the air inside. This is just talk kids

Ok, OK, so you want to go James Bond on us! I included some pictures that probably fall into this category. It involved a 5" diameter finned COPPER sink twin heat pipes. You can see it on the back of my Mean Machine in the pictues attached.

The unit weighs 5.1 oz or 146 grams. If you do the calculations and set it up correctly with all the appropriate tricks, you will lower the temperature at ESC FET's by 50 to 60 degrees F! This means if your ESC peaks out at 140 degrees ,F ( just before smoking), you could potentially reduce the surface FET temperature to 80 to 80 degrees F. This special finned copper heat sink with heat tubes and some other tricks can pull out almost 90 watts of heat.

While incredibly efficient it is not practical for small FE's obviously. But, if you did have one of those big mother FE's that are only run by guys who wear Extra Large Speedos, this may be the James Bond toy for you...........to keep your cool.

[questtek]

I can find those pumps all day long but they are 120v.

Oh there is a way. I did not know there was a better cooling fluid. Distilled water is great but there is better...but then you do not want to use ice with it like you can with water. You have to use the cooling cubes...or maybe have a seperate place for water and just freeze the entire deal and let the water run around it or something. A block form of ice will last much longer as well.

We are really only talking about enough to stay cold for one heat or sporting around. At this point I can't say how long the cold water will last after removing heat so that is what I will try and test...not fair now with colder weather but worth it.

I am now experimenting with a VERY simple idea for ESC cooling. It is no way near the finished system but it is the start of a concept that involves bringing the FET's in direct contact via copper heat sink to one of the 94 cent plastic containers of Blue Ice. The attached pictures say it all.

I machined a 3/16" rectangle of pure copper plate. I carefully mounted it direcly on the indentation in the Blue Ice pac, (on the side that has the label..after removing it). The conduction through the copper plate is incredible, as you would imagine. I now place the speed controller, in this case a 100 AMP model, directly onto the copper plate via some thermal conducting white "grease".

It is in the freezer now so hopefully I will have a chance to test it tomorrow.

This represents probably one of the simplest ideas for DIRECT cooliong with out the need or electronics or pumps. But the idea can be greatly improved. With a little more work some copper tubing can be added and thus provide a cooler water flow to the motor.

[Ub Hauled]

how about some heat transfer paste to get the heat off the FETs faster to the copper plate?

I live in Lake Forest, Ca....

I know exactly where you live!!! Kyosho is just up the street from you, I have been there in a few occasions. There is a lot of water in that complex, very nice to see from the road...
are you allowed to swim in it?

[questtek]

how about some heat transfer paste to get the heat off the FETs faster to the copper plate?



I know exactly where you live!!! Kyosho is just up the street from you, I have been there in a few occasions. There is a lot of water in that complex, very nice to see from the road...
are you allowed to swim in it?

Yes, I always use the white heat transfer paste. I think I called it "grease" in a previous post but paste is more appropriate.

In term of the lake, you are probably thinking of Lake 2 which is called the Sun and Sail Club You can see it from Lake Forest Dr. I live on Lake 1, (called 1 beacuse it was the first of the three artifical lakes built in this area. I live on the lake at Lake Forest Beach and Tennis Club. If you have a chance, stop over some time. I have never been to Kyosho ...do they sell from their shop to the public?

[Ub Hauled]

I am not 100% sure they sell stuff at Kiosho, but they have the shop with the "thinkers there and they are always up to something... Kiosho in my opinion makes the most beautiful, RC cars, they are not the best out there but in finishing they take 1st place. I went there with something I both in one of the RCX shows and was having problems with, they were friendly and fixed it... maybe it's the kinda place you can walk in and talk to the people making the things happen... besides, straight across the street is a "little" place called "RED" one of the best HD cameras in the market (it's the pet project of the owner of Oakley), never been in there... don't have the money to afford their cameras.
Yes, I was thinking of Lake 2, I didn't wanna mention the name of the street going up since I dind't know if you wanted to give away your location, but yes, it is Lake Forest Dr. that takes me up to Kiosho.
If I ever come down there again, I'll get in touch with you and see if we can meet.

[questtek]

Blue Ice / Coppersink Cooling Tests

The purpose of this bench test is to monitor the temperature of the FET side of the copper plate integrated into the Blue Ice pack as described in a previous post.

Themistors were mounted on FET contact side of copper plate integrated into the 94 cent Blue Ice plastic container and another measuring ambient temperature.

Ambient temperature remained a constant 79 degrees F throughout the testing.

Copper cooling plate temperature decreased with time. These values represent the temps the FET’s would see, (ESC is not under load for this initial so it generated no temperature).

Time Temperature of Copper Plate Delta Temp (ambient-copper cooling plate)

1 minute 59.9 degrees F 19.1 degrees F
2 57.3 21.7
3 54.9 24.1
4 52.5 26.5
5 50.8 28.2
6 49.0 30.0
10 48.5 30.5

To me, this looks VERY promising. I was hoping for a 20 degreed drop and the tests show the actual temperature drip is closer to 30 degrees F. This impresses me.

I have been waiting 3wbeeks now for the 200 AMP Mystery ESC’s to be delivered for some real testing. Hopefully I will not have to wait much longer to publish some actual performance results.

[questtek]

Looks like a new, very promising het transfer materal was just developed and released. One of the technical magazines I receive monthly just had an article on it. I ordered some samples so hopefully it can live up to its claims. It is apeel and place thermal solutions that will even withstnd the heat of solder reflow. It will permit the adhearing of the FET sde of the ESC directly onto the copper cooling sink.

See:

http://www.bergquistcompany.com/thermal_materials/bond_ply/bond-ply-450.htm

[Drag Boat Bob]

Did you put the whole thing in the freezer after attaching the ESC via 'grease'?

If so, did you seal the ESC with anything?

Just wondering about the condensation as everything starts to thaw.

This is great stuff, keep up the good work!

[Jeff Wohlt]

There are virtually 1000's of small boxes with lids that could be filled with blue and frozen to actually fit on top of the ESC fets with the termo tape with very little weight if that is the idea. No need for larger. A small copper plate mounted on the box, even alum plate would be fine for that cold, then freeze it and you can even hold it on the ESC fets with two rubber bands. It would be very cold.

I do worry about condensation but at what point that would happen is undetermined yet. No matter if the gel begins to met as it would still have plenty of cold....maybe not frozen yet only left in the cooler of ice so it gel is still very cold....which should do plenty.