Troubles with Max battery current on Kelly Controller

Tummy000

100 µW
Joined
Aug 11, 2022
Messages
8
Hi,

This is my first post here. I'm currently making an electric moped. Sadly I've encountered some problems. Hopefully some insightful people can give me advice.

The setup I'm using is as follows:
One 60v (17S) 41ah battery with a 100A fuse. It is connected to a Kelly KEB72451UX controller. These are built into an Benda scooter (better known as the Ebretti 518) . The bike is still using the stock motor, but with a thermistor added. I'm planning on adding Ferrofluid in the near future.

The main issue:
In the settings of the Kelly Controller you can set two limits. Max battery current and Max motor current. The battery current setting can be changed from 20-70%. This setting is supposed to keep battery amps low, but is does absolutely nothing. I've done multiple runs back to back while changing this setting. (Of course I've reset the controller in between runs) The power of the bike remains the same. Meanwhile the the Max motor current setting does work. I can't set it any higher than 40% or else the fuse in the battery pops at higher speeds. The motor current setting can be set from 30-100%.

While accelerating from a standstill the acceleration remains constant up to near top-speed using the 40% setting. This means power increases with speed. This also means I'm leaving some power at lower speeds. The controller is supposed to deliver high motor current while starting and gradually decrease motor current as motor speed (and thus voltage) rise as to not draw too much power from the battery. Regardless of the max battery current setting, my controller doesn't do this and always tries to deliver the set max motor current even while voltage is high. The controller uses too much power and trips the battery fuse.

This can't be right. The controller is supposed to limit battery current to the set maximum. Has anyone else experienced this before?

Kind regards,
Tummy
 
I don't have this problem with the KLS series, maybe it's a fixed battery to phase amps ratio on yours.
With the KBS I had a similar issue where the first 30 seconds it pulled more amps than what was programmed, after that it was fine.


You'd need a more powerful battery or different controller if you're not satisfied with how things are right now.
 
:flame: I have the same controller too and I am having some overcurrent issues. I have tried to reduce the battery amps and motor amps and my controller acts the same way... batt amps does not make a significant effect. I have not tried to turn it all the way down yet but every now and then I get a jolt and an overcurrent error. If I try to apply power too fast it will go into overcurrent situation. Yes I run with a 200 amp fuse.
 
This is an absolute bummer. Especially for a controller with a total cost of over 500 eur. A 50 eur aliexpress controller has beter riding charactaristics than this junk.

I wanted this controller for the current (torque) multiplication. But I literally can't even use it. It also is a square wave so a horrible vibration at low speed.

The amp ratio shouldn't be fixed. What's the point in using it in the first place in that situation. This controller is rated at 4,5KW , my battery can supply 6KW so it shouldn't ever be to weak.

Software is clunky overall. No more Kelly controllers for me. That's houndreds of euro's down the drain..... I've heard Sabvoton controllers are better in every aspect and also cheaper. Might give them a try.....
 
This is the absolute worst f***ing controller i've ever seen. I've now severely limited both the amps and top speed to that of the standard cheap-a$$ controller that came with the bike. Even at these power settings the motor is becomming red-hot. Heat has neven been an issue with the old controller.

The Kelly brand has been nothing but disappointing up until now.
 
I can't find the thread now, but there is one where someone explains that the kelly current limits are all percentages of the "rated" max phase current. Meaning, even the battery current percentage would be of the phase current rating.

If true (which I don't know), then let's say the controller's phase current limit is 150A. The battery limit of 20%-70% would then be 20% to 70% of that 150A, or 30A to 105A. The phase limit of 30%-100% would then be 45A to 150A.


Why the controller isn't limiting battery current I don't know--I would guess that it is a firmware bug. If Kelly has other firmware they can send you (if you have the necessary programming hardware; you'd have to ask them what that is for your model unit), maybe it would fix the problem. But it could actually be a design problem with the firmware for that model, instead (which I'd consider a bug, but design problems never get fixed while bugs might because the programmers don't consider design problems bugs (or problems)).


The other problems reported also lead me to suspect design problems in the firmware.

The heat problem usually means that the phase/hall combination or "angle" (timing) is incorrect, so that the current pulses are happening out of sync with where the magnets in the rotor are. If it's sufficiently out of sync, it'll cause stutter or noise or even fail to spin, and will behave differently at different speeds and loads. (usually worse at lower loads and speeds)


Kelly doesn't seem to "know" how to design their setup software either, because not much in it makes sense from a user / usefullness perspective. A user should not need to know to type in specific numbers to turn things on or off, there should be buttons with "on" and "off" in them. For anything variable, there should be a slider, knob, etc., and a value readout next to it, that shows the user the actual usable range, and the presently set value. It should be that simple.

But theirs is definitely not, and the UIs of other manufacturers are just as messed up or even moreso in some cases, based on various threads around here (not even counting bugs that cause undesired behavior--just talking about the UI design itself). Not even going to the very poor translation of technical terms in the UIs (for the few that have any non-Chinese in the UI).
 
Hey Amberwolf,

Thanks for your reply. I've heard before that they use the way you mention for calculating current. The rated current of this controller should be 220A. With Phase current at 30% it should be limited to 66A. And the 20% should make that a 44A battery current.

I've also seen a different way they might use. For battery current it's: Max_Phase_Current * Phase% * Battery%. That would make the max battery current 220 * 0.3 * 0.2 = 13.2.

I know for sure the current drawn from the battery is more than 13.2A and also more than 44A and 66A.

I've send Kelly an e-mail stating my issue and the things i've tried. I'm waiting for their response.

Regarding the hall angle. It's new information for me. Didn't know it also had an effect on heat generation. It should be a 120deg motor, but i've tried both 60 and 120. In both cases it seems to heat up extremely fast. But I feel like it's a bit slower with the 60deg setting.

Hopefully it's a software issue. If so, Kelly should send a new firmware update in response to my e-mail.
 
The battery current at full throttle, no load (wheel off ground), for a typical system is around a couple of amps or so, maybe up to a few for big motors. Much higher, and it can indicate the wrong phase/hall combination.


The "hall angle" I'm referring to is their "hall angle idenfication test" (which is not likely the correct term for what that test does) that has to be run in their setup software (at least on the kellys I have any experience with; don't know about yours) in order to make any motor work with the controller. Without running this, the controller will just run the motor however it happens to by default, which may or may not be correct.

The other way to deal with this should be to perform a manual phase/hall combo / combination check, for most controllers. There are 36 possible combinations, but it can take only (or less than) 6 changes to find the right one. I don't know if there is additional stuff the Kelly checks for and sets itself up with during it's HAIT, so no guarantee this would completely fix the problem without running that.

If the motor is running in the correct direction, you can leave the phases as they are, and swap just the three hall signal wires around until the battery current at full throttle, no load, is around a couple of amps. If the motor direction is wrong after this, you can either use the reverse function of the controller (in the setup software or with the reverse wire signal), or swap the phase wires around until it's in the right direction *and* at the lowest possible battery current at full throttle with no load. If it is easier to swap phase wires than hall wires, then do those first instead of hall--it makes no difference to the test.

If it's a 120 degree motor you need to use that setting in the controller, or it won't correctly decode the halls and the timing of pulses to the motor to run it won't match with where the rotor really is, and it wont' run right.

Tummy000 said:
Regarding the hall angle. It's new information for me. Didn't know it also had an effect on heat generation. It should be a 120deg motor, but i've tried both 60 and 120. In both cases it seems to heat up extremely fast. But I feel like it's a bit slower with the 60deg setting.
 
Their response:

The battery current limiting is based on the peak current setting from motor current also.
It can not limit the continuous current,just for peak current limiting.

You can set up the motor current at a lower value such as 60% and set up 20% for battery current limit.

Thanks,
Fany


Sounds pretty weird to me. This is a 4,5kw rated controller, how can 6kw not be enough. Even at the lowest current setting? And what the hell does peak even mean? This is a 220a peak and 85a continuous controller. 220a peak is said to be for 10seconds. After that the power is 85a???? This doesn't seem right. It doesn't cut power at all, not even afer minutes.

Who in their right minds would make the current limit a % relative to the "peak" power? At Kelly they can't even tell you what peak power is.

Their explanaition changes nothing about the fact that the battery current slider just doesn't work. At all. 20% is still equal to 70% as far as the controller is concernd.

I'm really getting pissed of right now with this piece of junk. If I wanted a controller without any changable setting I would have bought a Aliexpress controller for €50. The Aliexxpress controller would also have been way smoother.
 
amberwolf said:
If the motor is running in the correct direction, you can leave the phases as they are, and swap just the three hall signal wires around until the battery current at full throttle, no load, is around a couple of amps. If the motor direction is wrong after this, you can either use the reverse function of the controller (in the setup software or with the reverse wire signal), or swap the phase wires around until it's in the right direction *and* at the lowest possible battery current at full throttle with no load. If it is easier to swap phase wires than hall wires, then do those first instead of hall--it makes no difference to the test.
I've verified the hall wiring. Even stripped a new bit of wire on all cables to ensure good contact. Nothing changed.

I measured the amps with the wheel of the ground @ full throttle. Then changed the setting from 120 degrees to 60 degrees. The power consumption is exactly the same (3.46A). Even after the dot. Like EXACTLY the same. Motor also doesn't run colder/hotter while riding. Changed the setting bach and forth a few times to be sure. No change at all.
 
Tummy000 said:
Their response:

... It doesn't cut power at all, not even afer minutes.

Who in their right minds would make the current limit a % relative to the "peak" power? At Kelly they can't even tell you what peak power is.

I mean dont get me wrong, mine works ok, gets reliable 60mph.. but.. yes, I have overcurrent, resets, and the slider does nothing to prevent this. I have tried to clean up the wires too, a good return to ground. I have been told to try to reduce currents, but I still get same performance and jolt action time to time.

...it rides well, only when I hit the throttle some times I have a problem, a jolt, and the bike sets an error.

I think I have an older model, the model with 150 phase amp, not 220.. but I do see over 180A from the battery input line from time to time. When I measure.

I dont know.


Here is a quote form Kelly FAQ: I do not know how it applies to my controller. the sticker is worn from mine, I do not know if I have a X model or a older one... but it does have the new style label with the website name on it. Mostly worn away. I do have same problem, my good powerful battery causes over current events, and slider changes very little performance.

The controller current rating is the motor current. Basically motor current=max motor current *percentage, battery current=motor current*percentage.
For example, KEB72450, we set Max Motor Current and Max Battery Current to 80% and 60% respectively. The motor current will be 150A*80%=120A peak. The battery current will be 120A*60%=72A.

ref: https://kellycontroller.com/faqs/
 
Hi DogDipStick,

The controller they use as an example is the same controller I have, only without the ultra high-speed swiching. I found their example earlier. But the math doesn't work out. In short, nothing works.

Good to know I'm not the only one having problems.

Next time I'll pass on Kelly and go with Sabvoton. Still don't like how they charge €500+ for this guesswork.
 
Today i've continued troubleshooting. I've bought equipment to measure the current and voltage over time.

First off, I burnt myself by touching a phase connector on the controller. My finger now has a burn blister. It should probably not get this hot where it's able to burn a person in a fraction of a second. Does anyone know if this is normal? It was after only 3 runs (0-42kmh).

Second, I found out that torque (amperage) is not constant with speed, but it increases. (Yes, not only power but also torque increases). I can feel the bike starting to pull harder as it speeds up. As discussed before, this should not be the case. If anything, torque should decease with speed to keep battery power constant. Attached is a graph with torque vs time. At t=0 Throttle is increased to 100% and it is only released at t=12. You can see two dips in the graph, the fist just after throttle is given, this dip is NOT noticable while riding, maybe because of the low vehicle speed or because of the short duration (the polling rate of the measuring device is not that fast). The second dip is very much noticable, it occurs when the vehicle hits 42kmh, that is the speed where it's limited to. My guess is that the controller slightly overshoots the set motor voltage and corrects this by limiting current to 0. After this dip, current is lower, this is logical given that the vehicle is no longer accelerating but rather keeping it's speed.

I would like to know how to stop current from increasing with speed. And when possible even how to decrease it with speed as to not pull too much power from the battery.
 

Attachments

  • explained graph.jpg
    explained graph.jpg
    269.5 KB · Views: 522
Tummy000 said:
First off, I burnt myself by touching a phase connector on the controller. My finger now has a burn blister. It should probably not get this hot where it's able to burn a person in a fraction of a second.
If just the connection itself is hot, it means it is high resistance, not making a good connection, so power is wasted at that point as heat.

If these are ring terminals, it can be a bad crimp from the terminal to the phase wire, or it can be a bolt not quite tightened down enough on the controller itself (or incorrect order of terminal/washer stack, with washers between rings and controller instead of on top of rings between them and bolt head).


If it's the controller that is that hot, it means it is seeing a lot of load from the motor (which may be normal depending on the power needed to do the job you're asking of it, and just more load than it is able to easily handle so more power is wasted as heat than there should be), or it is not yet setup to correctly drive the specific motor connected to it, so more power is wasted as heat than there should be....


I don't know what the settings/etc need to be for the rest of the issues you're seeing.
 
Back
Top