What will heat up a controller and what parts get hot?

[macribs]

With DD hubs going slow speed up hills they will produce heat on both motor and controller it seems like. High speed on normal roads not so much of a problem. How are RC motors and controllers suitable for RC motors different from those hubs regarding heat?

What will cause a RC motor controller to heat up? And what parts do get hot?

Will there be different load on controller or motors if hall sensors are added to RC motors and the controller run sine wave?
Or what about field weakening? From what I've read it seems heat is the single largest reason controllers fail. If one could remove heat more efficient then air cooling of the controller housing hopefully heat failure will not be an issue.

I guess the Fets get heated, hence the mount in a heat sink. But we often see what I think are capacitors on the inside pics of controllers - do they overheat as well?

It seems to be doable to get a modified PC water cooling set to cool down the fet's and the large heat sink. But will that be sufficient to keep the controller from over heating? What other part of controller might need a separate water loop?

And what about those tiny RC motors. At the moment there are not a lot of water cooled high powered options out there it seems.
But I do see several water cooling sleeves for retro fitting. Have anyone tried to water cool RC motor?

 

[arkmundi]

As to the what: any electro-chemical medium through which current is running. That can be as simple as a wire, or as you say, things like FETs, capacitors, transformers, batteries, etc. Note that a battery, for instance, when at rest, generates no heat. As soon as current flows, however, the whole battery will heat. Marginally, if the battery is new and in good condition. But ageing will cause micro-plating of the terminals and internal dendrite formations as the active electro-chemical composition degrades. That's what increases the Internal Resistance (IR) of a battery and the more resistance, the greater the heating.

Its resistance in the classic Ohmchart that is the "cause" of heating.

But the natural resistance in sophisticated electronics of any kind means heat is generated. All consumer devices have built in heat protection circuitry. Common methods of heat control are fans and heat sinks.

Water cooling of any electronic device is not a good idea for the obvious reason that it uses water. Spring a leak and your equipment is gonzo. Heat sinks are the best method. A Meanwell charger, for instance, does not use fans, only heat sinks, so are famously quiet and immune to water, so can charge outdoors in the rain (some of them). A fav of ES'ers on long distance travel.

If you're trying to control heat, explore the heat sink approach to its fullest. If it proves insufficient (doubtfull), then there are some other things that might be tried.

 

[spinningmagnets]

Have anyone tried to water cool RC motor?

There are RC motors designed for model boats, and some of them have a much thicker end-plate (the end that the internal stator is attached to), and these have a passageway to route water through it. There is a poor selection of them, but they exist. Plus, you could have a thick DIY aluminum base-plate brazed onto an RC motor of your choice, to replicate the style in a more available RC motor.

I'm fairly sure the most common heat-related failures on an E-bike controller would be the FET's (from reading, I've been lucky so far). If you increase the number of FETs without changing the amps you use (going from a 6-FET to an 18-FET?), then the controller will run cooler.

You can add two fans, one pushing in, and the other pulling out (located at opposite ends, of course). You should be able to find fans that run fairly quiet, and since you would have fans, you could encase the whole mess to muffle the fan sound. Centrifugal style fans move more air than axial. If the fan drain worries you, install a thermal switch, so the fans only come on when the heat is nearing a dangerous level. You could even add an LED to show the fans are energized as a warning to pull back on the amps?...

Low heat is one of the major benefits of going to the hassle of a mid drive. By downshifting and keeping the motor RPMs high, the motor and controller will run cooler...compared to keeping it in one gear and bogging the motor down.

Guys who like DD hubs have sometimes moved to a smaller rim (20-inch?) and raised the volts to keep the same top speed. Maybe the same watts, but at higher volts and lower amps (amps = heat).

When it comes to high-power RC ESCs, copy a successful build exactly, one that matches your needs, or...be prepared to throw money at the build on a regular basis.

 

[macribs]

About the leak I am not overly worried. I ran for several years various rigs (pc and servers) that relied solely on water cooling.
And those where running 24/7 with all kinds of loads. In fact even the pumps where more reliable then the fans that cooled the radiators. The main thing for water cooling is to keep different types of metal separated to avoid corrosion. My dictionary does not have the correct therm for the English language.(maybe term is galvanic corrosion?) But using copper and aluminum in the same loop of water cooling will lead to metal corrosion. Using one or the other and there is no problem.
Many of you will have seen this on boats. It is common to use a sink or even magnesium anode so the anode goes before anything else.

I hope to us water cooling for motor as well as controller. I guess one does not mean leave out the other. If heat sink for Fets are water cooled a small fan could cool the rest of controller. I guess the FET's are what will produce the most heat?

Will sine wave produce more heat then pwm? It seems sine wave will turn on/off current more rapidly.
I am not dead set on this but I would like to see if it is doable.

4motus uses water cooled hub motor. Works great. Others have also cooled hubs with water loop here at ES.
But so far I have not seen anyone do that to a controller. I mean if you got a radiator set up why not try to put that to the most use?

You know Tesla goes even more radical. They water cool even the batteries. Not that I plan to go into that as it would require even bigger radiator and most likely fans as well.

But as for break downs going for water cooling come at a price. That cooling loop will be a single point of failure. Which for many is not an option. If that pump stop working one day only option to get home is "unplugged" by pedals. I could go for dual pumps with a mechanical cut off for one pump - so in case of failure I could do a manual switch engaging the second pump. But that will be too much PIA with extra wiring, extra tubing etc. So I think single pump will have to do. After all, when doing bigger services one could just swap the pump for a new one ahead of time - those things costs next to nothing anyway.

 

[Thud]

How are RC motors and controllers suitable for RC motors different from those hubs regarding heat?
fundamentally, they are exactly the same...its the scale of components & their application in-situ that favors one set up over the others

What will cause a RC motor controller to heat up? And what parts do get hot? the 1st cause of heat is resistance in the wires, traces,& components that make the controller. Unless I am corrected, the #1 heat generating aspect of motor control is switching the power transistors on & off in rapid succession...each switching event creates a micro second of increased resistance (and heat) & depending on the inductance of the motor, the flyback effect (nearly infinite amperage) will be hammering the back door of the fet's & diodes there to absorb the spikes. Now let's compound the equation with commutation events. Example 1. Typical DD hub motor = 23 magnetic pole prs.(46 magnets) lets say you need 23 commutation events to complete 1 rpm of the motor (actual # is actually more given the tooth count, but close enough for this example) multiply the switching by motor rpm, lets say 300 rpms x 23 = 6900 switching events per minute. For a direct comparison, we'll take an 80mm outrunner that needs to spin 6000 rpm's to deliver its power...it has 7 pole pairs. 7 x 6000 = 42k switching events. Thats 6x more switching to generate heat for the high rpm motor.

Will there be different load on controller or motors if hall sensors are added to RC motors and the controller run sine wave?
Or what about field weakening? From what I've read it seems heat is the single largest reason controllers fail. thats too many questions over too broad a range....pass :lol: If one could remove heat more efficient then air cooling of the controller housing hopefully heat failure will not be an issue.I asked these questions looking for the limits on the XieChang controllers...far more learned fellows explained to me that the limit is in the fet's themselves...I could run the fet's fully submerged in cold oil & they will still fail since the heat generated cannot dissipate through the part fast enough to realize the benefits of liquid cooling in a high rpm application...this is also compounded by paralleling transistors to build current ratings...unless they are closely matched by their switching thresh-holds & remain matched as they heat up...their performance drops dramatically.

I guess the Fets get heated, hence the mount in a heat sink. But we often see what I think are capacitors on the inside pics of controllers - do they overheat as well?

It seems to be doable to get a modified PC water cooling set to cool down the fet's and the large heat sink. But will that be sufficient to keep the controller from over heating? What other part of controller might need a separate water loop?read above

And what about those tiny RC motors. At the moment, there are not a lot of water cooled high powered options out there it seems.
But I do see several water cooling sleeves for retro fitting. Have anyone tried to water cool an RC motor?

member Etard built the best active water cooling for a race bike set up...he had a water atomizer aimed at the windings in his motor...it would spray water directly onto the end turns & turn to steam, then blow out the motor taking the heat with it. All fine for some fun tinkering...but if you want anything resembling reliability, design & use components within their ratings & you will have no problems with heat, overheating or failure...but what fun is that.

 

[Punx0r]

My dictionary does not have the correct therm for the English language.(maybe term is galvanic corrosion?)

Correct

 

[amberwolf]

With DD hubs going slow speed up hills they will produce heat on both motor and controller it seems like. High speed on normal roads not so much of a problem. How are RC motors and controllers suitable for RC motors different from those hubs regarding heat?

Mostly, they're not, under the same conditions assuming the same capability, and no extra cooling setup.

Keep in mind htat if you use a motor (any kind) setup to run at the slower speed, there will be less heat in it than one setup to be more efficient at the higher speed that's then run at the slower one (be it thru gearing or thru winding it differently, and/or controlling it differently).

What will cause a RC motor controller to heat up? And what parts do get hot?

Same thing as any motor controller (or other power-handling device)--it's the losses inside all of the parts in the current path. The heat that's generated in those will also migrate to everything else in the controller, so if it's run at/near/beyond it's limits it'll eventually heat everything in it up just as hot as the parts in the current path, assuming it's sealed up with no extra cooling.

FETs are the primary heat gnerators, but the capacitors themselves also have some internal resistance, and if there is current flowing thru them (if voltage is spiking or dipping enough to cause them to flow current in or out to buffer the voltage elsewhere in the circuits they are in parallel with), they'll heat up too. There are a lot of things that can be done during design to optimize things for the least heat generation even at higher loads, but most controllers don't have much, if any, of that stuff. There've been more than a few discussions about htat sort of thing in the various custom-controller-design and power-stage-design threads around ES, as well as controller-modification threads, though thy're probably not al that easy to find iwth a quick search.



Voltage regulators will also generate some heat, though minimal compared to the above.

Will there be different load on controller or motors if hall sensors are added to RC motors and the controller run sine wave?

It probalby makes some difference, htough I don't know how much, for sine vs trapezoid and hte resulting heat, in the motor itself.

Others that know better, please correct me if I'm wrong here: AFAIK, in order to run a sinewave out of a controller, it's gonna be using the linear region fo the FETs themselves, which means greater heat generated in them, as they are in their higher-resistance region longer than they would if used to simply switch on and off. I guess if they are still PWM'd to create the sinewave, that's not really true, but otherwise I would expect it to still be true.

Or what about field weakening?

I've no idea if field weakening would make any difference to the heat; it makes a motor wnat to run faster, so I guess maybe if you wound a motor so that with FW at max, it'd run at the higher speed you want, and FW at min it'd run at the slower speed you want, it might be more efficent as a motor. :?

From what I've read it seems heat is the single largest reason controllers fail. If one could remove heat more efficient then air cooling of the controller housing hopefully heat failure will not be an issue.

The best way is to make your controller and motor setup so it's being run within the limits of the system to start with, and so that it's all efficient enough to not gnerate the heat in the first place--any heat you generate is simply wasted power you could've been using to move yourself along.

You can remove heat from things in a lot of ways, some better than others depending on yoru environmental concerns, but if you can prevent making tha theat in the frst place it's a way better scheme.

 

[friendly1uk]

With a pwm drive, the resistance of the fets while in the on state, is a good indication of how much heat will be generated. RDS I believe, resistance on drain from source.

Peeps talk about swapping the fets for one's with lower rds values. The threads generally concern higher power use. While a cheap 6 fet box might be rated 12-15A the best one's are rated at 25A. These better one's generally use fets with lower rds on values to improve thermal loading.

I think with an rds of zero, you have a superconductor that basically won't heat up.

Perhaps a fet swap will alleviate your heat concerns.


I don't know what to say about the sine stuff. I have not seen one on a scope. Paraphrasing what others have said, It's more like a square with rounded corners.

 

[Punx0r]

As detailed above, RDSon isn't the whole story, as there are also switching losses to consider. Sometimes controllers seem to blow at part-throttle and I imagine it's the switching losses that killed them.

 

[friendly1uk]

As detailed above, RDSon isn't the whole story, as there are also switching losses to consider. Sometimes controllers seem to blow at part-throttle and I imagine it's the switching losses that killed them.

Arr, interesting. I meant to say my post was reaching out of my depth a bit. Makes perfect sense, though I haven't heard the term since I took electronics at school in the 80s.