Water Cooled ESC & RC motor ... would this work?

I have seen several water cooled units on HK, and I got to thinking, since they are designed for RC boats, and aren't used in a pressurized system (or so I am thinking) would the way they work be similar to the old "candle powered boat) that is pushed along with heated air/water though a coil around a candle that makes a little toy boat go across a pond?

If they do work this way, what's to stop us from having an unsealed water tank (much like an automotive radiator over-flow tank) with both tubes in the bottom of the tank circulating the water to cool the motor & ESC?

Has anyone else already done something to use these water cooled motors/ESC's?

I haven't checked the details on their use in boats, but it's more likely the water is "pumped" through the system using thrust off the propeller, than convection. This is how personal watercraft are setup for cooling, where a fitting is built into the wall of the jet pump housing and it scavanges a small amount of water from there and uses the pressure generated by the impeller to force it through the motor.

A fish tank pump could be used on an e-bike, or there are also other tiny energy efficient pumps available, say from Edmund Scientific that would work great for this.

I like the idea of water convection cooling, it could make for some very cheap reliable and viable systems. It would be setup so the hot water would leave the top of the controller and travel straight up and to the top of the water tank, and the cold water would leave the bottom of the water tank and enter the bottom of the controller. So when the water starts bubbling or even if the water boils(I know motors can take temps above 100C, can controllers?) it should only push the water causing it to circulate faster.

If you are using an aluminum frame, perhaps use part of the frame tubes as the radiator? Drill two holes and have threaded bosses brazed onto them. Un-needed parts of the frame can be plugged off before the brazing. The 1934 Porsche/Auto-Union had a mid-engine, and they ran the coolant through the tube-frame to get to the radiator in the front. They were suprised at how much heat was shed by the steel tubes, and they were able to reduce the size of the radiator.



If you wanted a seal-less water-pump, encase a steel impellor inside a non-magnetic housing (Stainless, aluminum, plastic, etc) and spin a magnet on the outside. Thats how laboratory stirrers work...

Very kewl! I'm going to investigate this idea further, I might be seeing about trading up into at least a water cooled ESC! I have seen several on HK that look promising, and being longer than the nominal ESC, just means that the pipes stick out the front or back of the Kepler drive.

Here is a nice set-up that could easily be made to my specs, I just don't know if cooling the stationary side of an outrunner would be enough? Not sure if this is an outrunner or not in the pic:



I'd also need a similar coil built around the ESC to make this work.

I saw just the other day a live bait bucket bubbler, very similar to an aquarium bubbler pump, I suppose this would work as our pump too, but of course I think the convective cooling makes a lot more sense with no moving parts to fail!

Anyone ever had a water cooled ESC apart? I imagine it would have a similar coil too.

Here is a DIY version of the boat my father made for my brother, pretty much a simplified steam engine. I don't know if this will work at lower temps, but I would hope so, I will have to do some testing with my brother's old toy boat and some dye and see what minimum temperature is required for circulation.



http://sci-toys.com/scitoys/scitoys/thermo/thermo.html

Well, as some of you smart guys already know, this method doesn't circulate water, maybe if I heated just one side, but I think I will just invest in a small pump, I will test and see if the HF live bait aerator pump will do the trick, or some other small 12V pump.

I used to water cool overclocked computers and there are small 12v pumps that can move over 300gph while drawing ~24w. and some that draw as low as 10w moving 100gph. the below site also has radiators sized from 80mm square to 120mm x 360mm.
http://www.crazypc.com/products/cooling/watercoolingmain.htm

Great! Thanks!

Anyone know what kind of flow I should need to keep things cool? I see that guys running RC boats are keeping things right around 90 DEG F with water cooling, both motor and ESC!

I'd really like to recreate those results!

Anyone know if this is a good group of components? Looks like a really good price! ($52.07 shipped)



http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=380226697275&ssPageName=STRK:MEWAX:IT

I could take this and put the heat sink on the top of my ESC (or just on the outside of the Kepler Drive where there is an aluminum plate on top of the ESC, and maybe "tee" in another small coil for the ends of the outrunner housing, maybe 2 turns on each side, that combined with a heli outrunner cooling fan, should keep things nice and cool! 8)

The pump in that unit only draws .12a so its probably going to move ~30gph free flow. I would not want to go with any pump with less than 60gph free flow.

My son recommended frozen CPU for the best selection for these kinds of components. I have no experience with them, and have no idea if their prices are low/mid/or high. But, perhaps they can show whats available off-the-shelf, and then you can search for alternative suppliers.

http://www.frozencpu.com/

By the way, using heating and cooling is a well-known phenomenon (no pumps). Its called thermosiphon. Even if you add a pump, it is very useful to configure the system to be thermosiphon. You will have 'some' flow if the pump dies (to limp home) and also, the pump will not be fighting the natural flow of thermo-pumping.

The loop should be near vertical instead of horizontal. If the heat is applied at 7:00 (on the face of a circular loop) and the cooling is applied at 1:00, the natural flow will be clockwise.

How big of a overheating issue do you have ??
I am pretty sure that using the "thermosyphon" principle and the frame as a water tank/heat soak, you would pull enough heat away from the ESC to satisfy your needs.

Well, enough to cook a motor I guess. :wink:

I'm intrigued by the idea of using part of my frame for a tank, however, I'm not sure how it is constructed, and if the ends of each tube are sealed, I think perhaps I should stick with an external tank for now, since I will be able to monitor things visually, I'm just wanting to build in a bunch more redundancy so that I don't have a similar problem in the future, and I understand that some others have gotten away with much less, but I will be going over some mountain passes with a very heavy touring load, and I want to be able to run the motor pretty hard with out having to worry about heat, so if this will get me there, then if nothing else I will have more piece of mind.

I'm wanting to also have my Kepler drive covered by my back aero fairing/trunk, so normal air flow will not be available. I'm thinking the Ebay computer cooling gear will be enough to get me the basics like the heat-sink, radiator and fan if nothing else, I'm really wanting to run it with the thermo-siphon so that the need for a pump (which can fail me) will be eliminated, then just the fan will be depended on, and a basic computer fan is available at most any store, or for cheap/free from a recycle center/local dumpster. (given I can find a sympathetic ear at a store/transfer station/recycle center)

Remember ..no amount of cooling will help if you electrically overload the ESC or motor.
A current "spike" will blow electrical components before cooling will take effect.
be sure you have your systems( Motor, ESC , batterys, etc ) correctly spec'd and current limits set to prevent "magic smoke" ! :wink:

Yep, This set-up I am using has been used successfully by others, but I believe I will be putting it to additional stress with weight and lack of air flow cooling the unit directly, I plan on having my radiator exposed to the air, and with the help of a cooling fan, but I may find that I need a more robust ESC, but I have been told by someone else that is using this same motor & ESC at approximately the same weight I will be pulling, but I just want to be extra sure.

If you are going to use active methods of cooling (vs purely convection) what about the technology that is used in the small portable ice chests? But my first thought was have you taken full advantage of simple air cooling in your motor, i.e. venting (drilling holes in the motor), using ducting to divert more air into the motor, connecting a fan to the shaft, installing heat sinks to the casing of the motor in combo with these steps?

If you look on this page of cyclone you will see a heat sink made for the case of a motor: http://www.cyclone-tw.com/order-A.htm It is near the bottom of the page and to the left.

Yes, I will do my best to use some form of a outrunner cooling fan (one that attaches to the shaft and blows air onto the motor when it spins) but other than that, there really aren't much by the way of surface area that can be used to attach cooling fins on the outrunner body which is also the roller, but I guess I could attach it to part of the body, not a bad idea thanks!