Enertrac hub motor cutting out

henrynz

1 mW
Joined
Feb 1, 2019
Messages
16
Hello all,

I have put together a farm motorcycle with Enertrac MHM603 hub motor and a Kelly KLS12301-8080I controller. I have a 100 volt pack which should be capable of around 140A continuous.

The setup works very well (smooth, good power,ridden for a couple of hours) except for one very specific scenario. If I give the motor a lot of throttle (tends to be at low speeds, steep hills etc) it can cut out very sharply, repeatedly and will quickly engage again, over and over until I let the throttle off.

Before I go into details of my controller setup, voltage and current graphs during these episodes could someone please let me know what behavior one should expect from this kind of setup when a motor is overloaded/stalls etc?

For example, the first motorbike I built has a AC induction motor with a curtis controller. If I ride up to a wall and apply throttle, the motor loads up to around 50A (firmware limitation AFAIK) but doesn't cut out.

TLDR: What observable characteristics should I expect from this motor when it can't power up a hill etc... before I start theorizing on how to fix the cogging/jerking/cutting out?

Thanks
 
Outside of some intervention, the motor should heat up until it glows red and catches on fire.

The problem is almost certainly outside the motor, though possibly a Hall sensor issue but not generally with those symptoms.

My guess would be something to do with power ramp-up curves or just a general error in the controller itself. SFAIK nothing else which would produce those symptoms under those conditions with that equipment.
 
henrynz said:
I have a 100 volt pack which should be capable of around 140A continuous.
[...]
If I give the motor a lot of throttle (tends to be at low speeds, steep hills etc) it can cut out very sharply, repeatedly and will quickly engage again, over and over until I let the throttle off.

“Should be capable”.

While the behavior you describe could well be the controller intervening, it’s more characteristic of the BMS working as intended to keep you from destroying your battery. If there’s even one cell that’s weak or out-of-balance enough to trip the minimum cell voltage protection under heavy load, the result will be as you describe. When the load is released, the cell voltage recovers and the BMS turns back on. Then you beat on it again and it cuts off again, etc.

If you can check your individual cell voltages, say at the BMS plug, you can verify whether or not the pack is balanced. If it passes that test, then attach a voltmeter you can see while you ride, to determine whether the load is bringing the full pack voltage down to LVC.
 
With that kind of behavior the first thing I'd check is the throttle to make sure you don't have a loose wire or magnet. A thermal cutback or current limit trip should not behave like that, because it's downright dangerous. Either of those should require a power reset or at least a rollback of the throttle to 0v. The only time I ever experienced similar behavior is when I got caught in a downpour and some dirty water (electrically conductive) got somewhere it didn't belong inside my throttle. With 15km left of my commute home, heavy traffic and rain, on a bike with extreme torque and a throttle going from nothing to full throttle with only a slight twist made for a scary ride home.

If the issue is from a thermal trip, then just like any direct drive hubmotor even the big Enertracs would benefit from using a smaller diameter wheel, especially on an emoto used on a farm where higher loads from thick mud and/or steep hills are to be expected.
 
AngryBob said:
My guess would be something to do with power ramp-up curves or just a general error in the controller itself. SFAIK nothing else which would produce those symptoms under those conditions with that equipment.

Thanks - I have trialed quite a few tweaks but I'll recheck my controller settings and see if there are slew rates or other settings which could be causing problems.
 
Balmorhea said:
If you can check your individual cell voltages, say at the BMS plug, you can verify whether or not the pack is balanced. If it passes that test, then attach a voltmeter you can see while you ride, to determine whether the load is bringing the full pack voltage down to LVC.

Thanks for reply. I spot welded the cells together so there could be a weak cell! However the problem occurs even when the Orion BMS is not connected. When the BMS is connected I don't get any fault messages and I'm fairly certain the BMS is not cutting out the power. I did monitor the voltage drop during cut out sessions and the drop was pretty small - by memory around 10 volts but I'll check it again. Cells are all balancing fine but I'll look a bit deeper and double check the BMS isn't involved.
 
John in CR said:
With that kind of behavior the first thing I'd check is the throttle to make sure you don't have a loose wire or magnet [text removed] If the issue is from a thermal trip, then just like any direct drive hubmotor even the big Enertracs would benefit from using a smaller diameter wheel, especially on an emoto used on a farm where higher loads from thick mud and/or steep hills are to be expected.

Thanks, I'll look into the throttle. I put pretty much the smallest wheel I could on the bike. It's a 16 inch rim. Perhaps I could go smaller again but it might start looking a bit weird/ hard to lace and top speed would probably be sub- 50km/h which might be a touch tardy.
 
henrynz said:
John in CR said:
With that kind of behavior the first thing I'd check is the throttle to make sure you don't have a loose wire or magnet [text removed] If the issue is from a thermal trip, then just like any direct drive hubmotor even the big Enertracs would benefit from using a smaller diameter wheel, especially on an emoto used on a farm where higher loads from thick mud and/or steep hills are to be expected.

Thanks, I'll look into the throttle. I put pretty much the smallest wheel I could on the bike. It's a 16 inch rim. Perhaps I could go smaller again but it might start looking a bit weird/ hard to lace and top speed would probably be sub- 50km/h which might be a touch tardy.

You already did what you could. 14's are really hard to lace into and there's so little tire selection in 14 that it doesn't save diameter. I know a lot of guys were lacing those motors into really big wheels.
 
if this is happening repeatedly over a decent span of days or weeks, with otherwise normal operation otherwise, unless you are running in rivers or very rainy I wouldn't think water in the throttle would be an issue.

When these cutoffs and restarts happen, are they very rapid, with the pulses regular in length, or do the periods of off and on vary significantly in time?

I'm guessing rapid and regular, not random. Which makes me think a software process, and since it does this with no BMS, I'm still thinking controller.

Unless maybe you just broke up with a Haitian voodoo priestess?
 
I didn't mean to say look only for a water in the throttle issue. It just seemed like it may be throttle problem, which would be the best situation...easy and cheap to fix and fairly easy to diagnose. Secure the bike with wheel off the ground and start messing with the throttle and its wiring all the way to the controller to try and recreate it. Ebike testers come in handy for throttle testing too.
 
**EDIT** - I've recorded some more graphs and taken a video of the controller monitor - I''ll upload them soon. I think the graph I posted is inaccurate because the Orion BMS seems to record data at a lower resolution than it displays it (the graph here is missing some true peaks and troughs I think)

Hi again - thanks for everyone's help

I've been talking to Mark from Enertrac and Fany from Kelly about this problem and I've been trying out their suggestions (mainly around controller settings). Nothing has helped so far. I've attached screenshots of the controller settings. Bear in mind I've tinkered with "Accel Time", "Motor Nominal Curr", "Current Percent", "Battry Limit", "Max Fwd Speed %", "MidSpeed Forw Speed", "LowSpeed Forw" and probably a few others.

AngryBob -No she broke up with me. They are rapid cut outs I would say. Seem to be a direct result of the motor "giving up" at high loads. I've attached a link to a video which hopefully shows it ok

Motor cutting out video:
https://1drv.ms/v/s!AmBBbdY-PMW5nWatmJBVX_RPMUpZ?e=EDRr6N

John in CR - I haven't messed with the throttle yet but I do take your point, could be an easy fix. Easiest way to find out might be to swap throttles /put a potientiometer on for testing. You mentioned thermal rollback earlier and I'm pretty certain that's not an issue as I've been monitoring motor temp. The one time I tested thermal rollback it seemed to work very well, controller just shut off motor power and I had to wait ten min before I could ride again (I set the max temp very low as a test).

It's niggling me that my battery pack may not be capable of delivering high current because of bad spot welds, poor connections or cables/ terminals that don't have a large enough cross section. Now all of these things are possible, particularly the cables because some of these are only really rated for 70A continuous (it's still just bunged together, I will use thicker cable for the final build)....However, if the battery pack was struggling to deliver sufficient current, wouldn't we be seeing a different situation? Particularly when you look at the graph (attached) that the BMS recorded (when I had it plugged in) during a hill climb/ motor stuttering scenario. Also I've been able to draw higher current at higher speeds (around 110A) when the motor doesn't stutter.
 

Attachments

  • Control.PNG
    Control.PNG
    28 KB · Views: 1,650
  • Monitor.PNG
    Monitor.PNG
    27.7 KB · Views: 1,650
  • Motor.PNG
    Motor.PNG
    30.1 KB · Views: 1,650
  • Vehicle.PNG
    Vehicle.PNG
    34.8 KB · Views: 1,651
  • Hill above house 2.PNG
    Hill above house 2.PNG
    17.3 KB · Views: 1,650
Henry,

Motors don't "give up", instead they'll kill themselves trying and the process takes time gradually heating up until sudden failure from which it won't recover. I don't think there's any chance it's the motor, since even a phase wire connection that comes loose can't re-attach. The controller should give an error code even if that happened.

The throttle is even easier to test than you think. Simply secure the bike with the wheel off the ground and give it a go. Load or no load makes no difference to the throttle. Testing that way also gives you a way to test under load by using the brake...just disconnect the wire for brake cutoff/regen, so you can have the motor pull and brake at the same time.

The stuff you mention about battery connections and tab welds is concerning. If something like that is the problem then there will be connections getting hot and catastrophic failure including fire is possible.

Are you certain that your wiring combination (phase and halls) is correct? Strange behavior can occur with what I call a "false positive wiring combo". ie Motor spins and in the right direction, but starts are rough and usually a bit noisy. With the wheel off the ground and very gradual throttle to start it's usually easy to tell by the sonics. No-load current draw is much higher and a false positive causes very high current under load despite a lack of torque. Another video with sound of a very gradual start with the wheel off the ground is probably enough to verify a correct wiring setup.

Hopefully it's something as simple as just a loose throttle wire, because I look forward to at least one vid of your machine running right. The small piece we see above looks like it will be quite a bike.
 
John in CR -

I have tested the throttle in that manner - I have a bike jack and the throttle is very smooth all the way through the range. I don't have a back brake yet but I do test under load by just facing the front wheel against a wall.

I've pulled the bike apart and I'm putting in some thicker cables ( 35mm squared ) from the controller to the batteries, which I was going to do anyway. I may also have to improve the thickness of the interconnects between cells. I'm reasonably confident of my spot welds as I spent quite some time testing them but I won't rule it out either.

I've attached image called "BMS Stall compilation" mainly because it appears that my screen shots are much higher data resolution than the recorded graphs I posted earlier. The BMS must be averaging data or only recording data points at a low resolution...can probably change this in settings.

The video below shows a typical readout from the controller software during a short period of motor stalls with jerking starting and stopping (note there are no sustained stalls in this video such as we see on one of the graphs below with the white arrow).

https://1drv.ms/v/s!AmBBbdY-PMW5nXOetnki2jotG4dO?e=Zw97rp

Also someone at work has offered to help me hook up a oscilloscope to the motor/controller to get some accurate data which I think I'll do.

Thanks everyone for your help and interest, I'll definitely update this thread when/if this gets solved.
 

Attachments

  • BMS Stall compilation.jpg
    BMS Stall compilation.jpg
    141.3 KB · Views: 1,612
If the problem is only happening under stall conditions, then there may be nothing wrong. I would never twist the throttle to a significant % of throttle on any of my bikes for more than a split second if the wheel wasn't turning. If I'm understanding correctly you held the throttle with the motor stalled for 20 seconds. If that's true I don't know what you're trying to accomplish with such a test, but you're very lucky you didn't break something. Occasionally I may use motor torque to hold my position when stopped on an incline, but only very briefly using a low amount of throttle, and if longer I quickly switch to using the brake(s).

The reason to avoid a stall is that no work is being done, so 100% of the power drawn from the battery is turned into heat. Also, controllers need the feedback from pulses from the hall sensors as the motor starts to turn to know the rotor position relative to the stator so it can fire the phase pulses with the correct timing. Strange behavior from the controller with the motor held in a stall is not surprising at all.
 
No I wasn't holding it in stall for anything close to 20 seconds. I would apply throttle and it would almost instantly create the jolting issues and current would return to zero. The graphs shows a series of applications of the throttle for the purpose of displaying the issue.

The white arrow does show a couple of seconds which for sure isn't great to be doing but shouldn't cause issues as long as I monitor heat and current output. I have spent quite a bit of time riding and building electric motorbikes and I don't think this is normal behavior at all. As I mentioned it happens riding up hills at various slowish speeds ( sub 15km/h) as well.

What you are saying about feedback from sensors seems sensible, but it shouldn't cause these sorts of issues at low speeds. I wouldn't have thought a traction motor should do it even at stall but I've probably done enough speculating. I'll report back when I know more.
 
Henry,

If that jerking is happening at low speed then yes there's something wrong. I wouldn't do any testing in a stall though. Creeping up a steep hill is one of the hardest things on our systems, because phase currents spike to the limit even when relatively low amounts of power are drawn from the battery and there's virtually no air flow to cool anything.

Is this happening only at low rpm, or once it started happening it does it all the time under heavy load (steep hill at speed or hard acceleration)?
If it's only at very low rpm, then it must be the controller, since max power out of the battery wouldn't occur until somewhere between about 1/3 and 1/2 of top speed under hard acceleration.

If it happens at more than just very low rpm, then it could be the battery, though I don't see how it could keep repeating instead of just failing. Do you have another battery, so you can take that out of the equation?...They don't have to be on the bike as you can use the brake to replicate the load? Have you tried bypassing the BMS?

Originally you asked what should happen when overloaded, and in that case the normal process would be that the bike would just smoothly bog down to a stall.

Some of those controller settings seem strange to me, though I never did like Kelly's program options (unclear and not enough true control over important settings.) I see you have the change direction box checked, which makes me wonder if you're certain that the phase/hall wiring is correct. I'd never set up any controller so that it's essentially running in reverse to get forward rotation. Another one that sticks out is the battery current limit of 25...Is that amps or %, in either case it's not a setting you want other than while making sure wiring is correct, definitely not for riding.

What is your no-load current at WOT? What rpm is that? If it's much higher than about 10A with that motor and tire, you've probably got a bad hall/phase wiring combo.

Have you tried unplugging the wire for regen braking to take that out of the equation?

Diagnosing problems is a process of elimination, so it's good to simplify the system to start with only essential items...motor, battery, controller, throttle. In the first post you said it ran fine for a couple of hours, but did that include any heavy load conditions (steep hills, or full throttle accelerations) or was it only riding fine taking it easy? Since you verified already that it's not something wrong with the throttle, I think it's most likely a controller or controller setting issue, though it could be a BMS issue. I doubt it's the battery, as a bad connection would have fried completely.
 
Wow I've spent a lot of time on this now! I wish my tenacity made up for my lack of understanding :wink: .

These are the things I've done since I last posted:

1) I tried setting current to 60% and battery limit to 20% (Most recent suggestion from Fany @ Kelly. It didn't help. I've tried lots of different combinations of Current % and Battery Limit.

2) I tried another battery. It's a proven, reliable 90V pack capable of over 400A. It uses 50mm squared cable. This didn't solve the problem either.

3) I changed to another throttle - this made no difference.

4) I totally removed the BMS from the system to make sure it had no effect - no difference.

5) I played around with Max FWD Speed, Mid FWD Speed and Low FWD Speed - didn't help.

6) I tried reducing Max Output Frequency - didn't help

7) When I increased the Speed Err Limit from 1000 to 4000 it made the cutting out significantly worse. As I understand, this is only part of PID adjustment but the manual doesn't go in to much detail. They only really say that you can reduce the PID values if you have strong acceleration (See point 8 below)

8 ) I tried reducing all three PID adjustments to half of the default but they just made the bike go slower - it still cut out sometimes.

9) My friend ran the the original test bed for this bike and he wonders whether he ran the motor (no load) before calibrating. He can't remember but I suppose this could do damage. Although he's a pretty careful guy with an electronics degree so I tend to think it's more likely me that's done something wrong

I'm kind of stuck right now but I've emailed Kelly and Enertrac with my latest results. I could get an electrical engineer friend with an oscilloscope to help me troubleshoot once I know what to look for. I basically know that's it's the motor or controller, would be great to rule one out.


John thanks for your replies:
John in CR said:
Henry,

If that jerking is happening at low speed then yes there's something wrong. I wouldn't do any testing in a stall though. Creeping up a steep hill is one of the hardest things on our systems, because phase currents spike to the limit even when relatively low amounts of power are drawn from the battery and there's virtually no air flow to cool anything.

Is this happening only at low rpm, or once it started happening it does it all the time under heavy load (steep hill at speed or hard acceleration)?
If it's only at very low rpm, then it must be the controller, since max power out of the battery wouldn't occur until somewhere between about 1/3 and 1/2 of top speed under hard acceleration. It seems to happen anytime there is bigger load, I've had it happen from 0 to around 20km/h. Although I'm a little surprised it doesn't seem to happen when going flat tack at faster speeds, around 40km/h. It seems to be related to high current and harder acceleration, because there is a fair bit of current been drawn at higher speeds. I might take it out today and see if I can replicate at higher speeds.

If it happens at more than just very low rpm, then it could be the battery, though I don't see how it could keep repeating instead of just failing. Do you have another battery, so you can take that out of the equation?...They don't have to be on the bike as you can use the brake to replicate the load? Have you tried bypassing the BMS? Great question - yes tried swapping battery and totally removed BMS as above

Originally you asked what should happen when overloaded, and in that case the normal process would be that the bike would just smoothly bog down to a stall.

Some of those controller settings seem strange to me, though I never did like Kelly's program options (unclear and not enough true control over important settings.) I see you have the change direction box checked, which makes me wonder if you're certain that the phase/hall wiring is correct. I'd never set up any controller so that it's essentially running in reverse to get forward rotation. Another one that sticks out is the battery current limit of 25...Is that amps or %, in either case it's not a setting you want other than while making sure wiring is correct, definitely not for riding. I think the programmer could be a lot better with not much extra work by Kelly. Yes it does seem weird that I had to check the change direction box. But the wheel is mounted in standard orientation and I have checked the phase wires are right ( ie colours match letters etc as per Kelly wiring diagrams). I think I could swap two phase wires to avoid having to tick change direction. I did check all the three hall effect wires were matching - they were but I also swapped them around and re-calibrated the wheel to run in the right direction without needing to tick the "change direction" box - this didn't fix the problem so I put the sensor wires back into their correct colour matchings. Battery limit is a percentage but I'm not exactly sure what they are taking a percentage of. For Current Limit they take the percentage of max controller current, in this case 300A. In any case, I have tried dozens of settings here from minimum to max and many in between.

What is your no-load current at WOT? What rpm is that? If it's much higher than about 10A with that motor and tire, you've probably got a bad hall/phase wiring combo. Kelly readout shows 950rpm at WOT. The current reading jumps all over the place from 0 to around 25A. The current readout doesn't seem very accurate

Have you tried unplugging the wire for regen braking to take that out of the equation? Yes, originally I didn't have regen wired up and the problem didn't change AFAIK when I did hook it up later on

Diagnosing problems is a process of elimination, so it's good to simplify the system to start with only essential items...motor, battery, controller, throttle. In the first post you said it ran fine for a couple of hours, but did that include any heavy load conditions (steep hills, or full throttle accelerations) or was it only riding fine taking it easy? Since you verified already that it's not something wrong with the throttle, I think it's most likely a controller or controller setting issue, though it could be a BMS issue. I doubt it's the battery, as a bad connection would have fried completely. Yes, it ran fine unless I was pushing it hard, heavy throttle at lower speeds and/or steep hills. It was mainly a range test I was doing so I was being pretty gentle mostly

Picture below - testing with my other battery (Yes it was a quick stall test, but it would of taken forever to physically swap whole pack. What a mess!

bat swap.JPG
 
First, don't do stall tests...Seriously nothing good can come of it.

Great work on eliminating everything but the controller and motor.

I never trust supposed correct wiring configs for halls/phase, and always verify correct wiring. With wheel off the ground and slowly giving it a bit of throttle, does the motor start spinning smoothly and quietly, nearly silent or is it a bit rough as it starts to go? Roughness indicates bad wiring combo. Since you're having to check the reverse box, it's not really right anyway. Try the other 5 possible phase wire combos, using only small short pulses of throttle to test each. Since you're not running significant current you can do this with jumper wires if that's easier. Just be careful to avoid shorts. If one of those gets you a more smooth starting forward rotation, then that was your problem and the only thing that saved the controller from blowing was some auto protection. If no combos work better, then the combo is valid, and if you wanted to change direction you'd have to swap either 2 halls or 2 phases and find the right combo of the remaining 5 of the other.

Do you have an ammeter to put in line with the battery supply? That's how you measure no load current. The 10A limit of common multimeters probably isn't enough for that motor with tire to get up to full rpm slowly twisting the throttle to avoid current spikes in order to measure no-load current.

Running without load is no issue at all. Running upper rpms is no issue at all. Hubmotors and controllers are happiest and the least stressed at high speed.

I only have 1 brushless Kelly controller, a little mini controller for ebikes and don't like it, so I'm not familiar with the programming, but it certainly sounds like a controller issue. A motor issue would be a loose or broken wire, bad solder connection, etc. and would be more of a work or not kinda thing, not load dependent intermittent. I think it's gotta be the Kelly. Too bad you're not down here. I'd bring over one of my 24fet Nucs, hook it up, run the auto-setup. Swap 2 phase wires if it runs reverse and go for a ride.
 
In the programming, does setting the throttle Mid Voltage to 3.15vdc, and/or changing the throttle Mid Current to 75 amps have any effect?
 
TommyCat said:
In the programming, does setting the throttle Mid Voltage to 3.15vdc, and/or changing the throttle Mid Current to 75 amps have any effect?
Thanks - I don't think I can change throttle mid voltage but I believe I have messed around with Mid Current before. I'm not actually sure that Mid Current does anything unless you are in 3 speed mode. But I'll try setting it to 75A next time just to double check.

1) I've now tried all 6 controller phase wiring combos with the prerequisite care. All 6 combinations allow smooth running of the hub at no load which I admit surprised me ( I did calibration routine every time of course). But I did have to change direction for several of them. None of them solved issue. I've tried all the hall wire combos as well. Mark from Enertrac has just sent me a wiring diagram for his motor which is different phase connections from the generic Kelly wiring diagram. Mark's diagram matches what I have now and I don't have to tick "change direction" anymore.

2) I hooked up my 10A ammeter to the battery supply and very carefully built up to WOT under no load. Hub draws 4.5A at 950 rpm. One thing I noticed is that the rpm readout on the controller monitor doesn't quite match with my count at low speed. ie I count 27rpm and it reads out 31 or something similar. Don't know if this matters.

3)Mark suggested running the controller input at 24V instead of 12V. I was running input off an isolated DC converter. First I tried a 12V battery to see if it could be some interference problem. Didn't help. Then I ran 24V off a benchtop PSU. No luck.

4) I now have easy access to the controller terminals so I think I need to hook up an oscilloscope with the help of a friend. Will post results
 
My brother hooked up an oscilloscope to the controller.

To be precise - we wrapped a wire 8 times around each phase wire for each probe. The probes all grounded to battery negative.
We don't know if this was the best way of doing it but I was hoping the readouts would tell us something. My bro thought that directly attaching probes was risky without further research.

I got the bike to do it's stuttering thing (which was fiddly with a mains powered scope!). The attached screenshots are various views/ versions of a typical cut-out/stutter event. I don't yet really know how to interpret the results but my bro thinks that it's further proof that the motor is most likely not at fault.

I'm trying to get to the point where I can be reasonably confident to buy a new controller. Only problem is, the only recommended controller is the one I already have - but hopefully it's just faulty. It is a 2018 model.
 

Attachments

  • DS1Z_QuickPrint1.png
    DS1Z_QuickPrint1.png
    17.8 KB · Views: 1,300
  • DS1Z_QuickPrint2.png
    DS1Z_QuickPrint2.png
    18.2 KB · Views: 1,300
  • DS1Z_QuickPrint3.png
    DS1Z_QuickPrint3.png
    16.8 KB · Views: 1,300
  • DS1Z_QuickPrint4.png
    DS1Z_QuickPrint4.png
    15 KB · Views: 1,300
  • DS1Z_QuickPrint5.png
    DS1Z_QuickPrint5.png
    17.6 KB · Views: 1,300
  • DS1Z_QuickPrint6.png
    DS1Z_QuickPrint6.png
    16.1 KB · Views: 1,300
I've bought a new controller -Kelly KLS96601-8080I. Same type but higher power - 600A peak with 240A continuous.

I'll update thread if this solves the problem. I don't really love the idea of buying another Kelly but I don't believe their are many other options - programming a Sevcon sounds like a tough nut to crack.

Getting a controller with double the current capability probably wasn't necessary but I think it's better to overspec with Kelly.
 
Sorry that I missed your previous post. It was definitely the correct wiring, since the controller does the auto-config, it tells me it had to be a controller problem. You had thoroughly gone through the process of elimination to prove it. While that first controller may have been under-powered for your motor, it still shouldn't have had that kind of behavior, so you need to send it back for credit if possible or replacement if that's the only choice. Then at a minimum you'd have a backup.

I wish I'd known you were at the point of buying a new controller. I would have suggested a Nucular 24F as long as having a 95V fresh off the charger max voltage could work for you. It would have been a significant step up from the Kelly you just bought, and cheaper. The only problem is the long backorder queue. FOC control, more compact form, real torque control, and variable regen are big pluses.
 
John in CR said:
Sorry that I missed your previous post. It was definitely the correct wiring, since the controller does the auto-config, it tells me it had to be a controller problem. You had thoroughly gone through the process of elimination to prove it. While that first controller may have been under-powered for your motor, it still shouldn't have had that kind of behavior, so you need to send it back for credit if possible or replacement if that's the only choice. Then at a minimum you'd have a backup.

I wish I'd known you were at the point of buying a new controller. I would have suggested a Nucular 24F as long as having a 95V fresh off the charger max voltage could work for you. It would have been a significant step up from the Kelly you just bought, and cheaper. The only problem is the long backorder queue. FOC control, more compact form, real torque control, and variable regen are big pluses.

Hi John. I bought the original Kelly around two years ago but took way too long to get it in the bike and running. I don't think I'd have much chance of getting a refund from Kelly even if I could prove it was faulty as it's well outside the warranty period now.

I was probably a bit impulsive buying another Kelly controller. Fany Chen from Kelly very kindly cancelled my order for me while I consider my options. I do remember you mentioning the Nucular earlier and I wrongly assumed it was mainly for smaller vehicles. My max pack voltage is closer to 105V though so I'd have to remove a couple of cell groups to get down to 95 and/or drop peak charge voltage down a bit. The wait time is not optimal either. Thanks for suggestion - I'll look into it further.

One other controller I've just found:
RGBL1896 96V 1x200A Brushless DC Motor Controller
https://www.robotshop.com/en/vbl2360-96v-1x200a-brushless-dc-motor-controller.html
Never heard of these before, specs might be marginal for my application but I've skimmed the datasheet and it's a very clear and professional looking unit. Still not even sure it's suited to my motor but I'll dig deeper.

Just as a general observation to anyone reading this thread, Kelly and Enertrac have been pretty good with the time they have spent replying to my emails even though we haven't got to a solution as yet.
 
Quick update: Mark from Enertrac saw one of my video's and thought there was a possibility that the magnet ring inside the hub motor was spinning. He offered to fix it free of charge. To check this, I pulled off one of the side covers and marked the magnet ring against the hub body with a marker. I reassembled the hub motor, took it for a test ride until it started cutting out/jerking again. I then disassembled it again. The mark had not moved so I believe I've ruled out this as a problem. It has furthered my conviction that the controller is at fault.

I'm on the wait list for a Nucular controller - around 4 months. Will update when I make any progress here.
 
Back
Top