Comparison between RC and ebike controllers?

[bobc]

Bobc,.. what motor are you using on the bike with that controller ?

Hi HH - the picture is not the controller (as in ESC), it's an interface board that looks at 1) throttle (hall or pot) 2) speed (spoke magnet + reed switch) and 3) battery current. The output is the pulse stream used to control RC stuff (e.g. ESC)
Using the "speed" input really does solve most of what I class as the issues in trying to use an RC ESC.
Unfortunately I jumped straight in with a design that is pretty much optimised for low cost production, which is ridiculous as it is very unlikely ever to be produced in bulk; I should have made something easy to adjust/program. Hey ho. A mass manufacturer could retail these for half a dollar or less.
The motor is an outrunner, about 50mm x 35mm with kV = 170 or 270 (forget which). It's going in 2 projects, a trike with left side drive and a folding bike with a friction drive.

 

[recumpence]

To add just a bit to this conversation regarding power output;

I have been pulling 11,700 watts from one Castle Creations Edge160 consistently (in bursts under acceleration) without any problems. I have customers pulling 4,000 watts from them in sand bikes (continuous power through the sand).

They are expensive, but they work if programmed properly and the drive is setup properly.

Matt

 

[Punx0r]

The 420A 24S Alien RC controller has been formally tested, see here:

http://endless-sphere.com/forums/viewtopic.php?f=2&t=44897&hilit=alien&start=600

Regarding running at partial throttle, IIRC the FET switching losses are much greater, resulting in more heating. They get an easier life at full throttle - just used as a switch.

Threads like these are interesting to read, but with the exception of the Alien controller test above always lack hard data or any kind of methodical testing. Why not wind up three resistive coils, throw them in a bucket of water, connect the controller and see how long it outputs its rated power before overheating?

 

[gwhy!1]

Matt... No one is doubting what these controllers can pull once rolling and sync'ed but it requires specific use of hardware and a protocol to use the throttle/pedals.. and will prob work with a much cheaper controller reasonably well also with the same hardware and use the same riding protocol.

you can not recommend a RC model controller as a standalone drop in controller for all e-bikes across the board.

but at the end of the day its still a RC model controller that was not designed for use with a EV .

 

[recumpence]

Matt... No one is doubting what these controllers can pull once rolling and sync'ed but it requires specific use of hardware and a protocol to use the throttle/pedals.. and will prob work with a much cheaper controller reasonably well also with the same hardware and use the same riding protocol.

you can not recommend a RC model controller as a standalone drop in controller for all e-bikes across the board.

but at the end of the day its still a RC model controller that was not designed for use with a EV .

Yup, nor can you use an ebike controller as a drop in for every application.

That logic applies in both cases.

Matt

 

[gwhy!1]

The only difference is that it will not stand you in $300 to try and poss lots of magic smoke and this is why people moan about rc esc . There will prob be no magic smoke or it will just not work, if a ev contoller is connected correctly.

 

[recumpence]

Exactly the same thing........

"As long as it is hooked up correctly."

 

[windtrader]

it is straightforward to make an RC controller tractable and robust by making the pulse width on the control line match motor kV x bike speed/gearing with an extra bit proportional to throttle position. When you do this your rubbish open loop RC ESC becomes a torque controlled nice thing! Honestly, it really is that easy with a wee micro board like the one I'm working on.

You are so right Bob. Someone like you with expertise in AD, DD, signal, PWM, can design a pretty simple circuit and firmware to meed the needs of these pretty simple requirements. Given this is a niche market, flexibility is desired at the expense of higher costs associated with both the flexibility and low production volume.

It sure seems with the potential market that would love to have a solid, proven, higher performance ebike power system based on RC tech at a reasonable cost uplift, someone would have produced a controller targeted for this application.

I don't have sufficient skills to do this but have worked with RTC OS and MCU and built some prototypes taking analog inputs, applying various simple algorithms, then triggering various output actions. It just seems like you need to simply input a few simple parameters then drive the PWM accordingly along with power gain via current or voltage). Again, I have very little expertise on the analog side so I may be all wet but it just seems a very basic design would be huge leap over where things are now. Just my 2 cents.

Are you the lone wolf BobC or is there a pack of you folks striving for this sort of solution or reasons to head off some other path?

 

[gwhy!1]

Exactly the same thing........

"As long as it is hooked up correctly."

No its not the same thing ...

cos even if you connect a RC esc correctly it can still end up with magic smoke showing its face.. the clear cut differences between a rc controller and a dedicated ev controller is that its a much bigger possiblilty of a RC controller WILL go pop if you get it wrong (maybe 90% ) where as a ev controller WILL NOT go pop if you get it wrong (maybe 90% ) .

I have never had a ev controller not "work" when connected up correctly or go "pop" if not correctly connected and this have been many, many types of ev controllers with price ranges from $10 upto $150 , they have all just worked out of the box . And the more expensive ev controllers ( $150) that I have used are more than capable of 10kw+ and they do start from zero rpm and have been 100% reliable after many many hours of extreme use.

 

[recumpence]

Tell that to all of the customers I have had order from me because of being sick of blowing ebike controllers.......

 

[gwhy!1]

If you have a setup that works for you and your customers then stick with it... but I do reject your claims that its the most cost effective and reliable way to go. Please remember that your complete reliable 10kw+ system is prob $500-$1000 more expensive than my complete reliable 10kw+ system. At the moment you have very few other options for controllers for your setups unless you use a sensorless ev controller or add sensors to your motors. A small piece of electronics as bobc has shown could be a turning point in turnkey reliability of using rc controllers, but until such additional electronics have been proven, RC controllers are still very much a hit or miss controller for use with EV's .

 

[recumpence]

Wow, when have I ever claimed that? Why would you lie and say I made that claim?

I have NEVER said that. Don't you EVER put words in my mouth.

This is the whole reason I did not post on this thread until the OP invited me.

When people start lying about me, I am done........

 

[gwhy!1]

No you haven't said this directly, but this appears to be the message that you are trying to spread.


now can we just leave it at that.

 

[recumpence]

I am not trying to spread that message. So, no, we cannot leave it at that. If I agree to leave it there, I would be agreeing with you that you are correct about my intent. Stop putting words in my mouth and we can leave it.

 

[bobc]

"Are you the lone wolf BobC or is there a pack of you folks striving for this sort of solution or reasons to head off some other path?"

Hah yeah - lone wolf.... whereswally on here is local & plans to do a bit on it but hasn't started coding yet

Using the little 6 legged micro, it's a very simple circuit with just 6 components - in particular no debugging aids.....

There's a thread on here about it which I'll resurrect when I make some progress - might be soonish as I heard today I'm in line for some sample shunts from isabellenhutte!

 

[yawstick]

I am doing something similar as well... I'm using arduino based servo control to limit current via the speed request adjustment. It started out as more of a home brew cycle analyst and I've not added wheel speed into the mix yet but something I was wanting to do. To limit current more at low speed and less as speed increases. I think the shunts might be a better option though as the hall effect current sensor is a bit noisy and not the simplest way.

 

[Hillhater]

Correct me if i am wrong here but...
All you guys using RC controllers are using them with RC motors ?
Even those with reliable working "RC systems" on Ebikes are using various "custom" modifications to the OEM RC gear..
such as.... extra capacitors on the controller, hall sensors in the motors, custom throttle systems, etc etc.
That normal, but it still indicates the issues with RC controllers when used on Ebikes.
I will ask again also... Is anyone using an RC controller on a hub or (non RC motor) mid drive ..CNG etc ??

 

[Dauntless]

You guys are doing some great jawboning without the insults getting slung around. All we need is a little more detail on what works and what doesn't. I remember Mr. Physics discussing pictures of Bruno's controller for what he could see was an improvement, such as the bigger heatsink. There's subject matter to get on target with.

 

[Punx0r]

Still no data or evidence. Just subjective claims and attitude...

I am not trying to spread that message. So, no, we cannot leave it at that. If I agree to leave it there, I would be agreeing with you that you are correct about my intent. Stop putting words in my mouth and we can leave it.

If it's any consolation, I have re-read gwhy!'s posts several times and I cannot for the life of me see any kind of insult or negative connotation that fits with your response.

 

[recumpence]

A couple comments ago he accused me of saying rc systems are the most reliable and cost effective way to go. I never said such a thing.......

This happens constantly on these threads. I stay away from these discussions at this point because this always happens. The only reason I posted here is because I was invited via PM by the OP.

I have said it before and I will say it again, RC systems have their place with ebikes, but they are not suited to every situation. It is an apples and oranges thing. Sensored systems tend to be less complicated and less costly, while rc systems are more compact and tend to be bit more efficient as well as light weight. Good rc equipment is costly, however.

Matt

 

[Punx0r]

I have said it before and I will say it again, RC systems have their place with ebikes, but they are not suited to every situation.

I think that point has been consistently made and comes across quite clearly in this thread.

This is the first time I've seen a retailer object to their product being described as reliable and cost-effective. Given the above caveat (there must be a similar one applicable to all products in all applications), I don't see the remark as contentious.

What would be interesting is some technical explanation of the differences in design and topology between the types of controller. RC controllers seem to make extensive used of SMD components to save space and I guess not having components to enable current limiting, ebrakes, regen, hall sensors and various other typical ebike features must save space.

The lack of significant heatsinking on RC controllers has always puzzled me. The answer must be that either they cannot delivered rated current, or the choice of FETs and the manner in which they are driven produces much less heat than in ebike controllers, EV controllers and industrial drives, which would be a remarkable achievement.

I lieu of any data, I am impressed with the performance RC controllers seem to offer in such a small/light package, but remain sceptical that their power ratings are accurate, which would be expected in the RC world. If I had the time, money and facilities I'd love to investigate the issue myself, but sadly I do not have these at my disposal.

 

[recumpence]

As for heat, I never see any appreciable heat build up in my controllers. These rc controllers are very efficient at high power levels. At low throttle settings and high current heat can build up. But I have never seen more than 150 degrees even under hard use.

Matt

 

[panurge]

The lack of significant heatsinking on RC controllers has always puzzled me. The answer must be that either they cannot delivered rated current, or the choice of FETs and the manner in which they are driven produces much less heat than in ebike controllers, EV controllers and industrial drives, which would be a remarkable achievement.

I think the airflow that a 12s 160A ESC is subjected to, in its natural use (A HUGE RC Airplane), is a reason about the sinks dimensions.
My 2 more cents for the debate: With or without current control, and a 24" wheeler geared for 50kph, I can ditch the pedals no problem, now that I'm familiar with the system......trial, static, and WOT from standstill applications is another thing, here the surge and the little delay are not good things. If you leave a little throttle residual thru the CA (resulting in something like 0.2 Kph and few watts no load) that will results in a critical improvement, anyway far from a true trial-motorbike feeling.

 

[tahustvedt]

I have a 1/8 scale F-16 on 12S lipos with an electric fan that draws 4 kW at full throttle. The ESC does not sit in the airflow, but inside the fuselage. It doesn't overheat during the three to five minute flights. You can imagine how high the average current draw is to drain a 5 Ah battery in five minutes.

 

[recumpence]

Full throttle is actually far easier on the controller than loading it up at low throttle. I bet if you flew hard at 50% throttle you would see more controller heat.

Matt