RC ESCs in a real size motorbike? How stupid would that be?

[Dui ni shuo de dui]

Hi guys,

I came across this thing, which is a ESC for hobby remote controlled race boats or surfboards:
https://www.aliexpress.com/item/32693130599.html?storeId=513485&spm=2114.12010612.8148356.53.51fd57c3ih4PQ3

800A per phase, apparently a 1000A/phase version exists too
integrated BMS
Programmable
Up to 22S of any chemistry
Watercooled

The power figures, features and prices seem fairly good to me, so I was wondering what is the actual catch?

Could these thing be able to power a big hub motor for, let's say, my motorbike?
Is that a stupid idea (I guess yeah), but if so, why?

 

[amberwolf]

If it's like the rest of the RC controllers people have used on here, big or small, they don't like high startup loads for very long, and don't usually have any way to actually monitor what's going on and react to it, so they blow up on startup. Not always the frist time, but eventually.

Most of the motors these things control in their "normal life" would spin up very quickly, so BEMF would quickly rise and lower the phase currents.

A hubmotor has HUGE startup currents for a long long time, relative to what these controllers were designed for, so they often end up as a box that no longer has any magic smoke to run on.

Even a middrive with an RC motor powered via RC controller tends to blow up for the same reason, because it takes so very long to get a bike moving and start raising the BEMF and lowering the phase currents.

Of course, I don't know anything about this specific controller, so it could be better than that.



Ifyou look up threads from 2010-2011, especially around the deathrace / spookytooth / tucson , undead race, etc., you'll find a number of attempts to do stuff like that. Technology has improved, but stuff I've seen posted since then doesn't seem to show much better results.

An exception has been with friction drives like Kepler's, probably because they can slip, and/ or because they have quicker startups to reach RPMs that let the BEMF rise, and multistage reductions like Recumpence's drive, wiht enough power to get the bike moving very quickly from a stop so the load isn't there very long. .

 

[Dui ni shuo de dui]

If it's like the rest of the RC controllers people have used on here, big or small, they don't like high startup loads for very long, and don't usually have any way to actually monitor what's going on and react to it, so they blow up on startup. Not always the frist time, but eventually.

That's what I was a bit afraid of.
I'm not sure they are monitoring phase currents... But if they would, then it would just come to a simple matter of cooling, right? In which case with a good water cooling flow and possibly a bigger exchanger for added thermal mass, wouldn't it solve the problem?

Anyway, thanks a lot for your answer

I did browse a bit through the forum and came on to this thread, which is also quite interesting and gives an idea of what to realistically expect:
https://endless-sphere.com/forums/viewtopic.php?f=2&t=44897&start=650

 

[flat tire]

It doesn't have hall sensors. Sensorless startup on a hub with high power will probably blow it quickly.

It won't have good power handling capability. Look at the physical size. Not all controller regimes are that efficient. Even at 95% efficiency it still has "too much" heat to eject continuously with a few KW.

RC ESCs hate low speed and part throttle which is where you will be spending too much time operating the hub.

Since that ESC is very expensive, why not just get one of the numerous controllers that are known to work really well with hub motors?

This might be a cool choice for some project, but it is the worst choice for yours.

 

[amberwolf]

I'm not sure they are monitoring phase currents... But if they would, then it would just come to a simple matter of cooling, right? In which case with a good water cooling flow and possibly a bigger exchanger for added thermal mass, wouldn't it solve the problem?

Probably not, because the heat spike happens very fast inside the die of the FET in cases like this, and there may not be time to remove the heat before it causes a runaway resistance change and thermal failure event.

But you could try it anyway; just document it all thoroughly so anyone else looking to do it will know what to expect.

Personally, I'd rather go with something known to work on big hubmotors, but I don't have money to spend on expensive experiments. (if I did have a lot of spare money, I probably *would* do a lot of experiments though :lol: ).

 

[Dui ni shuo de dui]

It doesn't have hall sensors. Sensorless startup on a hub with high power will probably blow it quickly.

Good point, I didn't pay attention to that.

It won't have good power handling capability. Look at the physical size. Not all controller regimes are that efficient. Even at 95% efficiency it still has "too much" heat to eject continuously with a few KW.

Well it's not as small as the pictures would suggest. It's actually close to the size of a sabvoton controller, without the casing.
Maybe a tad smaller but the Sabby is air cooled, this one is water cooled so it makes removing the heat a lot easier.

Since that ESC is very expensive, why not just get one of the numerous controllers that are known to work really well with hub motors?

This might be a cool choice for some project, but it is the worst choice for yours.

Yeah sorry, I should have been clearer on my first post. I don't really intend to buy it, I was mostly asking out of curiosity. My current controller is already close to maxing out my motor's capacity so changing it wouldn't make a real difference. But I was curious about those ESCs to see if they could be an alternative, especially for something else like a high power electric mountain bike. Guess they aren't.

Probably not, because the heat spike happens very fast inside the die of the FET in cases like this, and there may not be time to remove the heat before it causes a runaway resistance change and thermal failure event.

I'm not sure to understand why that would be different from a regular controller. They seem to have the same amount of Fets. Is that because they are SMTs so the hot side is on the PCB, which is far less thermally conductive than an aluminum block?


But you could try it anyway; just document it all thoroughly so anyone else looking to do it will know what to expect.

Personally, I'd rather go with something known to work on big hubmotors, but I don't have money to spend on expensive experiments. (if I did have a lot of spare money, I probably *would* do a lot of experiments though :lol: ).

Yeah, unfortunately I don't have that kind of cash to spend on experimenting
Really wish I did though, I love fireworks.

Thanks for your input guys, appreciate it.

 

[spinningmagnets]

The open-source Vedder-ESC (VESC) has been a huge improvement over past ESCs. It uses Field oriented Control (FOC) and that requires a shunt for each motor phase (three shunts).

As far as the physical size, the Adaptto and Phaserunner use very efficient FETs instead of "affordable" FETs, so they were able to run high amps with only six FETs. I suspect that this one is the same in that regard.

 

[DogDipstick]

I wonder if it has Active freewheeling for low loads ( sycronious rectivication)....
For 550$ it better lol.