Mid-drive high RPM, overheating, and max controller amp rating

richj8990

100 W
Joined
Jan 21, 2020
Messages
277
I've been noticing a lot of people complaining about all kinds of mid-drives where they have to keep the rpm's up in a low gear or there either will be inadequate power or the motor/controller will start to overheat. So they have to put it in such a low gear that the optimal cadence for pedaling is over 100, and they have to ghost pedal to get the desired amount of power from a low gear.

I've been running a lot of simulations on the Grin webpage...it really looks like the limiting factor (besides the motor itself overheating) is the max amps of the controller. A controller that has more max A rating is going to handle the increased current better --- so why don't the manufacturers simply provide a stock controller that can handle more amps coming through? Are they afraid the motor will burn out so they put on a mediocre controller so that dies first?

Example: let's say you have a 500W mid-drive with 100 Nm torque; in order to get the correct torque you have to run the cassette cog as 1:1 with the chainring, so let's say 42T chainring and 42T in this case lowest cog. And let's say 500W motor power, 40 volts going through, 12.5 amps at 400 rpm. But you may only be going 3-4 mph up a 10% grade and a non-ebike can do almost that mph without any electric power! So it becomes ridiculous. You want to go faster. You have the power, you just don't have the right controller to handle the gearing. 22A max controller.

So you upshift into 21T. Now the torque may fall to 50 nm but the speed will increase. To get the same 500W with only this time around 200 rpm motor speed, the lower rpm limits the volts going to the motor to 20. That means you need 25 amps and the controller will only allow 22A. So it either kicks down to a lower power (which means your climb either gets much harder or is over), it shuts off, or it overheats and may get damaged. But what if you had 30+A controller on there. Then the 25A would not be so bad. The question then becomes: if a mid-drive motor is rated 500W, 48V (and therefore nominal 10.4A), how many more amps can it realistically take on a climb if the controller can handle 30+ amps? Can it take 20 continuous amps for 10 minutes or more? 25A? 30A? The larger controller can take it, but can the motor. If yes, then just get a more heavy-duty controller and stop ghost pedaling.
 
It depends entirely on the mid drive. Some can indeed handle considerably more current than the factory controller and some cannot, and in fact controller mods are pretty common on some drives for that very reason. Also you have to remember the gears and clutches in a mid drive may not always be happy at increase currents (increased torque) and again some can handle considerably more and some cannot. Also of course some mid drives are geared more reasonably than others which is really the idea solution for a mid-drive you want to pedal since once you start bumping up the current you start entering power levels where you don't even need to pedal, better gearing and matching the motor to the drive yields better efficiency so smaller drive and battery which is often desirably in a pedal assist bike.
 
LOL.... I must be doing it wrong because I have never had this "problem" with my 3 mid drives. Only time I need to exceed 100 RPM is when I am in my 42-11 and still want to go faster and have no more gears.
 
It depends entirely on the mid drive. Some can indeed handle considerably more current than the factory controller and some cannot, and in fact controller mods are pretty common on some drives for that very reason. Also you have to remember the gears and clutches in a mid drive may not always be happy at increase currents (increased torque) and again some can handle considerably more and some cannot. Also of course some mid drives are geared more reasonably than others which is really the idea solution for a mid-drive you want to pedal since once you start bumping up the current you start entering power levels where you don't even need to pedal, better gearing and matching the motor to the drive yields better efficiency so smaller drive and battery which is often desirably in a pedal assist bike.

I really don't understand why mid-drives would have so many problems climbing hills, that's what they are built for. If they can't do that at a reasonable speed, like more than 5 mph, what is the point in having one. If you put on the right tires and have the right wattage for a geared hub drive, it can climb fairly well without overheating. Anyway, I'm going to have extra controllers and displays soon to play around with and see which is best.
 
WTF, Rich. Where do you get this stuff? I climb steep hills at 15 MPH or more when I want to, in normal cadence, with all 3 of my mid drives.

I really don't understand why mid-drives would have so many problems climbing hills, that's what they are built for. If they can't do that at a reasonable speed, like more than 5 mph, what is the point in having one. If you put on the right tires and have the right wattage for a geared hub drive, it can climb fairly well without overheating. Anyway, I'm going to have extra controllers and displays soon to play around with and see which is best.
 
Yeah I mean as I said it's not an issue on some mid-drives, it sounds like you have one of the worse ones, you never mentioned which. On my TSDZ2 bike which is known for the opposite problem actually the issue is really being in a low enough gear to climb the hill only because it doesn't have enough power otherwise. The power band is very peaky and narrow due to the low controller current. This may be part of the problem you are facing, low current controllers and the KV and voltage in some stock configs make them feel like they have a narrow power band so you feel like you have to be in just the right gear to get a lot of assistance. There are advantages and reasons why they do this though, efficiency being one.
 
Yeah I mean as I said it's not an issue on some mid-drives, it sounds like you have one of the worse ones, you never mentioned which. On my TSDZ2 bike which is known for the opposite problem actually the issue is really being in a low enough gear to climb the hill only because it doesn't have enough power otherwise. The power band is very peaky and narrow due to the low controller current. This may be part of the problem you are facing, low current controllers and the KV and voltage in some stock configs make them feel like they have a narrow power band so you feel like you have to be in just the right gear to get a lot of assistance. There are advantages and reasons why they do this though, efficiency being one.

I'm really posting this question for/about other mid-drives besides my own. I don't have a huge issue. I can do a 10-20 mile ride and the controller will not ever cut out. And I have the luxury of actually being able to solve the problem by testing out other controllers. Other people have proprietary motors with internal controllers and other 'secure' protected hardware/software that they may not be able to hack into or otherwise solve, and they supposedly are screwed. Either they pedal as fast as they can or they get a different motor or entirely different bike. That just plain sucks. Especially if they paid a lot more for their setup than I did.



OK this is what is happening with the KT, at least I think it is. It's not happening in (throttle) PAS 5 but it is in 4 and 3. What I think is going on is that if level 5 throttle is 1000W max, then level 4 is probably 450W. Which means 1/2 the voltage is allowed to get to 450W...yes or no? Either voltage or current is limiting for PAS max watt output and I'm assuming it's voltage.

If PAS 5 is 1000W (max allowed, remember this is throttle and not pedaling) and the bike is in a middle cog at 300 RPM, my KV is roughly 9.0 so that's 33V. Optimal watts for this are around 600. So that's 18A, which is within the 22A max for the controller.

But now I go to PAS 4 up the hill, and it's only rated for a max 450W. Volts then are limited to 16.5 instead of 22 because 450W is 75% of 600W. Which means amps have to go to 27.2 in order to maintain the 450W maximum. 27.2 Amps are over the 22A controller limit, which then starts to shut off the power, and power will fade or default down to PAS2. Again this is what I 'think' is happening. Because it doesn't happen in PAS 5 with 3/4 throttle, and doesn't happen in PAS 4 with 1/2 or even 3/4 throttle, but DOES happen with PAS 4 full throttle. Even if I don't feel any additional power in PAS 4 between 3/4 and full throttle (the throttle feels like a dead zone at the end of the travel), full throttle will for sure shut down or PAS default down the power within 10-30 seconds, every single time. To me that means the lower PAS levels are voltage-restricted. This does not happen with real PAS pedaling on a hub drive, same controller and another very similar KT display, but different outcome because it's motor RPM is not indirectly tied to the cassette cogs.

Solution: more heavy-duty controller. 30A just came in the mail, 45A one on the way. Better to have too many controllers than not enough IMO.

You may ask: why not just throttle in PAS 5 all the time, or uncouple the throttle from the PAS levels. Because on a steep hill, especially if there are off-camber turns and uphill switchbacks, PAS 5 throttle tends to often wheelie the front tire, which means either immediately stopping or slow crashing. PAS 4 eliminates that overtorqued wheelie issue but then has it's own overcurrent issue above. That will hopefully be solved with a stronger controller.

The motor has many names lol. Many. L-Faster, Dome Motor, and I just found out that it's originally a GNG 450W from all way back in 2013. Do not get me wrong, it's a good motor, especially for $100-150 by itself. This is a controller issue.
 
Last edited:
OK, What dynamic ( over time) heat transfer temperature measurements have you recorded to confirm your computer simulations?
That is: Proof of your theory your percieved power issue is actually NOT a poor heat transfer problem that a properly cooled unit would solve?
Not my problem, just wondering as an old thermo dynamics guy.
Good Luck to you

Ed: Here is my related and unexact reference to the one aspect of a throuough analysis including heat transfer by Justin Le:
Perhaps you have read Justin's work and completed a similar analysis or model.
A complete test bed.
BBS in Wind Tunnel.jpg
 
Last edited:
OK, What dynamic ( over time) heat transfer temperature measurements have you recorded to confirm your computer simulations?
That is: Proof of your theory your percieved power issue is actually NOT a poor heat transfer problem that a properly cooled unit would solve?
Not my problem, just wondering as an old thermo dynamics guy.
Good Luck to you

Ed: Here is my related and unexact reference to the one aspect of a throuough analysis including heat transfer by Justin Le:
Perhaps you have read Justin's work and completed a similar analysis or model.
A complete test bed.
View attachment 338861

Hello...um...I don't really think your answer was to any of the questions asked above.

Are you sure you are replying to the correct post. This is about too much current going into a controller that's not rated for that amount of current, and there is a quick stepdown or shutdown of the controller. It's not an overheating problem, it's that the controller is cutting power specifically to prevent an overheating problem. The problem can happen when both the motor and controller are cool and not warm or hot.
 
Yeah I mean as I said it's not an issue on some mid-drives, it sounds like you have one of the worse ones, you never mentioned which. On my TSDZ2 bike which is known for the opposite problem actually the issue is really being in a low enough gear to climb the hill only because it doesn't have enough power otherwise. The power band is very peaky and narrow due to the low controller current. This may be part of the problem you are facing, low current controllers and the KV and voltage in some stock configs make them feel like they have a narrow power band so you feel like you have to be in just the right gear to get a lot of assistance. There are advantages and reasons why they do this though, efficiency being one.

I think I may have figured out the real problem: the connection to the throttle. Not that this specific controller/throttle issue could be 'helped' by me personally; I'm not going to splice and solder all day with a resistor wire trying to fix this controller/throttle connection.
Last night I hooked up a different generic controller to the same throttle I'm using and the starting response was terrible. Took 2-3 seconds to even engage but then the power was smooth after that. That tells me it's a throttle to controller to motor issue. Maybe that would be OK for a hub drive but not a mid-drive where you have to lay off the power while shifting. Especially up a hill, and especially if a lot of watts go into the sudden engagement that could wheelie the front tire.

So my current setup with the KT controller and this same throttle is not great but compared with last night it's acceptable. Much rather have a deadband in the first 40% of the throttle travel than feel and hear the motor desperately trying to drive the chain until it finally is able to, and then there is a sudden lurch ahead every time after I shift. BTW watts are 'shown' on the display going into the motor, and a loud motor grumble while attempting to drive the chain, but there is no chain engagement for 2-3 seconds. Bad controller? Probably. Will return the POS. But also what is happening on that controller/throttle combination can also happen on the current KT controller. As in the motor sounds are not as bad and later engagement is better but that's after I take thumb off throttle and reset everything (give the system a rest for a few seconds). Happens once for every say 30-40 times I engage the throttle. Other controller it happened every time.

So there are probably two different problems overlapping each other: one is inconsistent throttle to motor engagement, one is too high of a gear with limiting current. Two separate problems. I can try out different throttles and controllers and see how improved it gets.

Buying 6 different throttles to test out, all returnable. Including one that worked well with a different (discontinued) controller. Will rank the throttle performance later in a separate post. I think that may fix the issue. Will certainly not make it any worse.
 
I figured it out finally. Put on a KT LCD3 that reads watts. Realized that because it was not reading the speed properly, it was lowering the watts based on incorrectly high speed. VERY IRRITATING! Why can't they have an option to simply ignore the bike speed, period.

P01 says to set any geared motor to 1. That's what I did. Then in PAS 1-4 I accelerate slowly from a stop and the MPH shoots up to like 45 mph because the speed calculation is so far off. So then power is cut. STUPID. Also happened when I guessed 14 magnets and set it to 35. The default came as 90 for P01. I set it back to 90. Problem went away. Really, really irritating. Don't say to set a geared motor to 1 if that throws off other parameters. The MPH is still off a little which means the real setting should be somewhere between 80-100. Which means I have a lot more magnets than I thought (have not taken apart the motor; someone else who did showed 12 coils but did not show the # of magnets on the other side of the casing).

The other thing: with a previous LCD3, I knew that imitation torque sucked because it would cut power like this. Switched over to speed control (P03 = 0 and not 1). Better. Now looking at the watts I can see the efficiency for pedaling in PAS 2-4 much better. No cut out of power for the 1.5 mile ride back home on this test run. Hopefully it will stay fine. So there is nothing wrong with the controller, it's the damn programming. Will still try out other throttles however.
 
You’ve put a solid effort in to customising a mid-drive setup, and done well keeping a lid on the frustrations involved. Your build thread’s great.

This thread’s too convoluted for anyone to be able to provide practical help though. I’d recommend continuing in your build thread, or starting a fresh one seeking help with optimising your controller setup. Details matter. The more you provide, the more detailed the responses you can expect. Same for clarity.

If it’s KT controller understanding and/or expertise you seek, the participants with the most of it can be found in the huge kunteng open source firmware thread. With the degree of customisation that you’re striving for, you should seriously consider the open source firmware in any case.
 
You’ve put a solid effort in to customising a mid-drive setup, and done well keeping a lid on the frustrations involved. Your build thread’s great.

This thread’s too convoluted for anyone to be able to provide practical help though. I’d recommend continuing in your build thread, or starting a fresh one seeking help with optimising your controller setup. Details matter. The more you provide, the more detailed the responses you can expect. Same for clarity.

If it’s KT controller understanding and/or expertise you seek, the participants with the most of it can be found in the huge kunteng open source firmware thread. With the degree of customisation that you’re striving for, you should seriously consider the open source firmware in any case.

That's the whole problem with many of these posts: they are too detailed. I'm not an engineer. I'm not a programmer. I just want to know why the controller cuts power when it should be adding power. It does it very intermittently and I'm just going to live with it. I've noticed that a lot of mechanically-oriented people simply cannot speak in lay terms. Or they are very secretive, or both.

I could list all of the display parameters, but there are things about the motor that I simply don't know, so if someone asks, I will not know the answer.
 
I see what you mean.

Still, you obviously wanted a challenge, or something unique, or better, or you would’ve gone the easy route with an integrated system like Bafang.

Controllers are just a necessary evil. Don’t think for a minute that anyone enjoys playing with them. But unfortunately we have to, so we do the hard yards until they perform satisfactorily. I don’t think you’re there yet, I’d be putting more effort in to a controller setup.

The answers you seek will be only one or two posts away. You just need to be systematic in describing:
1. What you have
2. What you want it to do
3. Why you think it can do what you want
4. What your first attempt was
5. What resulted from the first attempt
6. How many times you repeated the attempt
7. Why you think the attempt failed
8. What subsequent tweaked attempts you made
9. Why you think these also failed
10. What you think the gaps in your knowledge are that are preventing you from solving the problem

But like I said already, this thread is a write-off, you’ll want to start afresh.
 
"This is about too much current going into a controller that's not rated for that amount of current, and there is a quick stepdown or shutdown of the controller"

OK go ahead and Good Luck , I'm not researching that electrical failure theory
 
This is the answer to my problem from another thread, it has nothing to do with amps or overheating:

"You have to have an external speed sensor on the wheel with a mid drive. Set P2 = 1 on the lCD and the controller will use the external speed sensor input. Then the value of P1 will not be used and it can be anything. When users don't have an external speed sensor, then the controller will try to calculate speed knowing the P1 value. On a geared motor with clutch, the controllers lose the speed info on coasting because the motor no longer turns."


This is exactly what is happening.

1. Mid-drive motor.

2. No speed sensor on wheel, cannot put one on unless I change to a waterproof controller and then put on a Bafang BBSxx speed sensor and hope the speed sensor is compatible with the 3-pin PAS in a non-Bafang external controller. Guessing no, it will not be compatible.

3. Cannot use PAS as a speed sensor because cranks are fixed.



So this is why the watts are OK most of the time and then fall down to 11-25W after coasting. MPH on display also shoots up far more than actual speed. So then power is cut. This happens both in P3 settings --- 0 speed control and 1 imitation torque.

Coasting = controller tries to recalibrate speed and cannot = MPH too high, regardless of P1 or P2 setting = power is cut. Setting (throttle-based) PAS to 5 will regain power. Can ride in throttle PAS 3 or 4 on a flat surface or up inclines as long as I know that I need to go to PAS 5 after any coasting at all or any drop in elevation no matter how slight. If the 'flat' surface falls even 0.5% of a grade in elevation, the speed on the display will incorrectly reset to high and power is cut. Then PAS needs to go to 5 to recalibrate the system. Only other thing I could try is on C4, set to 2 and then raise the speed up to the maximum level, but then I lose the PAS-level based throttle power from "4".

Complete power cutout (not to 11-25, to 0) was because of a different problem: on a normal bike (or hub drive w/o brake cutoffs) I'm used to riding with 1-2 fingers resting on the brake lever. On mid-drive, right brake has a power cutoff magnet. Inadvertently touching the brake lever or magnet too much = power completely cut. Need to ride with fingers completely on handlebar grips and not touching brake levers.
 
Back
Top