Motor sputter at higher RPMs...

bchampig

1 W
Joined
Nov 19, 2016
Messages
52
Hello,

I am in the process of finishing a new build based on the LR single stage (small block) kit. I'm using a Powervelocity 12FET sin controller, and a CA3. I am running the throttle directly into the CA3 and only using the CA3 harness to the controller.

I am running the CA3 in pass thru mode. If I gun the throttle with the bike lifted off the ground, the motor spins up just fine, but if I'm riding the bike, the motor 'sputters' like a gas motor running out of gas. I am pretty sure the phase and hall wires are correct since the motor has no trouble starting up. The current draw is also fine, and the controller and motor is not running hot.

I played with lots of setting in the IOS phone app for the powervelocity controller. I disabled flux weakening, played with phase current and battery current. I am now running 40A battery/130A Phase, but have tried lower values with the same result. The data displayed by the CA3 looks ok...throttle in and out voltages, current draw, etc. Nothing looks weird and I'm not seeing any limit flags. The sputter seems to be torque related since it happens more when going uphill and when I open the throttle fast.

I realize my next step should be to eliminate the CA3 from the equation and run a throttle direct to the controller. I just wanted to know if anyone has had a similar symptom?

Thanks,
Regards,
Brian
 
Sounds like possibly running into LVC? Or maybe the BMS limiting current? The fact that it happens under high load makes me think this.
 
dustNbone said:
Sounds like possibly running into LVC? Or maybe the BMS limiting current? The fact that it happens under high load makes me think this.

I'm running 12S Lipo(Multistar packs) and I'm not using a BMS. I have the LVC setting in the powervelocity controller settings(via ther app) set to 40V. Could it be that this value in the programming software isn't accurate, and that the controller thinks the battery is at 40V when it's really at 44V?

I will give it a shot.

Thanks!

Brian
 
I tried setting the LVC on the controller to the lowest value, but the motor still behaves the same. Will try to bypass the CA3 next.
 
Yeah it's possible the CA is set to limit current as well. Or it may have it's own LVC setting, if bypassing it works then I'd think it's probably one of those two things. I would think it'd be more graceful about enforcing a current limit though.
 
It's too late to test this out today, but I remember that in testing, the controller would shut off if I set the max CA3 throttle output past 3.8v. I think it's right at 3.8v now. Maybe I'm right at the threshold of max throttle voltage to the controller.

I will turn the ca3 max throttle down to 3.7v.

thx,
Brian
 
I have Powervelocity 12FET paired with Cyclone 3k and having the same issue. No CA on mine, throttle direct to controller.
Battery is built to 120A nominal (16S12P Samsung 35E), BMS is rated for 60A cont, 160A overcurrent cut-off.

Initially had trouble with "false hall pulses" - it's Cyclone motor's inherent problem where hall sensor picks up a "stray" magnetic flux between rotor magnets. Trapezoidal controllers seems to be immune to this but all sinusoidal ones will get into a trouble (Powervelocity, Grintech, Kelly to name a few I have spotted in different topics and forums).
Here's couple of scope screenshots:
1.jpg
3.jpg
4.jpg

Did a workaround by adding additional pulldown 560ohm resistors to halls (directly to halls PCB in motor).

That fixed other issues I had with motor performance but that stuttering you're describing is still there: at top-end rpm, full throttle and high speed (which means there's considerable load on motor; at lower speeds it's unnoticeable and at mid-speed it will pass quickly).
Oh, and I replaced factory halls, no change there. Soon I will try different halls that have much lower sensitivity.
 
Thanks, this is good info. I also have a cyclone 3k and it works great with the stock cyclone controller. Never tried a sine controller with it. I will do more testing but if the throttle voltage isn't the culprit, or the CA3 current limit isn't the issue, I may toss this controller into the garbage and go with a stock cyclone controller. I dont want to use more than 40a with the small block motor anyway.

thx,
Brian
 
I lowered the max throttle output voltage to 3.6v from 3.8v and it solved the sputter issue. mSurprised the controller is so sensitive to this.

thx,
Brian
 
My stock Cyclone controller also stutters at super high RPM, while in mode #3 only tho; so I have a feeling this is not a problem with the controllers, more so of a problem with the halls not being spaced right, or placed right, so the timing signals are off by a hair and get the controller confused at high RPM... perhaps a sensorless controller would be a better option...?

I call the stutter bouncing off the rev limiter... :) as when I am riding if I don't shift it feels like hitting max RPMs allowed by ECU.

G.
 
I also have a cyclone 3k, and the stock controller only had a stutter at max rpm when the motor was running on the bench. Under load, I have never had this issue with my stock cyclone 3k, but I have only ran it up to 60V. On the bench, the motor stutter happened only in mode 3, and in the forward direction. In reverse it was fine which probably means the hall effect placement is not symmetric.

Brian
 
bchampig said:
I also have a cyclone 3k, and the stock controller only had a stutter at max rpm when the motor was running on the bench. Under load, I have never had this issue with my stock cyclone 3k, but I have only ran it up to 60V. On the bench, the motor stutter happened only in mode 3, and in the forward direction. In reverse it was fine which probably means the hall effect placement is not symmetric.

Brian

I run mine on 18S LiPo (66.6V) and I hit that stutter when I am fully charged @ 75.5, on the bench and when I am cruising on the road. Power consumption skyrockets when I hit that stuttering. So yeah, sounds like the hall sensors aren't placed properly on the motor, or the timing is off... not sure...

G.
 
minimum said:
I have Powervelocity 12FET paired with Cyclone 3k and having the same issue. No CA on mine, throttle direct to controller.
Battery is built to 120A nominal (16S12P Samsung 35E), BMS is rated for 60A cont, 160A overcurrent cut-off.

Initially had trouble with "false hall pulses" - it's Cyclone motor's inherent problem where hall sensor picks up a "stray" magnetic flux between rotor magnets. Trapezoidal controllers seems to be immune to this but all sinusoidal ones will get into a trouble (Powervelocity, Grintech, Kelly to name a few I have spotted in different topics and forums).
Here's couple of scope screenshots:
View attachment 2
View attachment 1


Did a workaround by adding additional pulldown 560ohm resistors to halls (directly to halls PCB in motor).

That fixed other issues I had with motor performance but that stuttering you're describing is still there: at top-end rpm, full throttle and high speed (which means there's considerable load on motor; at lower speeds it's unnoticeable and at mid-speed it will pass quickly).
Oh, and I replaced factory halls, no change there. Soon I will try different halls that have much lower sensitivity.

hey mate, did you add these to the hall outputs or the 5v supply line? im having the same issue you describe with a motor like lightningrods mid drive motors (bought direct from china though) where it stutters and violently jitters at mid-high rpms (and even looses sync altogether). Hoping this can save me doing a separate plate to mount magnets on that will trip the halls rather than the actual rotor magnets. chrs.
 
I thought I had resolved the motor stutter issue at high RPM by lowering the throttle voltage out from the CA3, but I think I was just masking the problem. It still happens at max RPM and under load on my LR single stage small block build. I will make up a cable to by pass the CA3 and run the throttle direct to the controller(Powervelocity 12FET).

I notice that the Current(A) and Wattage(W) limit flags on the CA3 go off when the stutter happens. I raised the limits on the CA3 to 3500W and 65A to see if it helped but they still went off, even though the mac current reported by the CA3 is around 44A.

-Brian
 
You might be exceeding the electrical RPM limitation of the controller and it's getting out of sync.
 
Fechter, that's a good point and definitely a possibility. It would also really stink because I don't know what I can do about it except replace the controller. The question is, with what controller? I wish there were more controller options out there.

THX,
Brian
 
bchampig said:
Fechter, that's a good point and definitely a possibility. It would also really stink because I don't know what I can do about it except replace the controller. The question is, with what controller? I wish there were more controller options out there.

THX,
Brian

For more RPM you probably need a higher kV motor, or rewind the motor you currently have.

G.
 
Any idea what the mechanical RPM is when it starts to sputter?

It could also possibly be a defective controller. I've seen something similar happen with one of those. It was fine at really low speeds but it would start to spike the current and run rough above a certain RPM.
 
fechter said:
Any idea what the mechanical RPM is when it starts to sputter?

It could also possibly be a defective controller. I've seen something similar happen with one of those. It was fine at really low speeds but it would start to spike the current and run rough above a certain RPM.

All the Cyclone motors I run on 18S LiPo start stuttering at around ~900 or so output shaft RPM, or around 5400 RPM internal RPM... I am using the stock Cyclone 40A controllers on both trikes, so I can't say if its a problem related to the controller or the motor, but both are pretty much equal in terms of when the stuttering happens.

G.
 
I haven't tried this yet, but I noticed that my pole count setting in the Powervelocity programming app is set to 4. I believe it should be set to 8 for the small block motor. I hope that fixes my issue.

thx,
brian
 
bchampig said:
I haven't tried this yet, but I noticed that my pole count setting in the Powervelocity programming app is set to 4. I believe it should be set to 8 for the small block motor. I hope that fixes my issue.

thx,
brian

I think it's only using the pole count to sense motor speed for the speed limiting functions. I don't think it will change the eRPM limitation. Worth a try for sure.

5400 RPM is pretty fast. If you have the same issue with two different controllers, then that would probably eliminate a defective controller. Another possible cause might be the response time of the hall sensors in the motor. Typical setup is there are pull up resistors pulling the hall signal lines high and the sensor pulls them down. If the pull up resistors are too high of a value, the signal might be responding too slow. It might be possible to just add some external pull up resistors to the hall signal lines to speed those up. With power off, motor halls disconnected, you can measure resistance from the 5v hall power line to any hall signal line going to the controller. You should be able to measure the pull up resistance directly this way. 5k would be a typical value. You could try placing 4.7k resistors from each hall signal to the 5v hall power line.

They might be just using crappy hall sensors in the motor too.
 
fechter said:
bchampig said:
I haven't tried this yet, but I noticed that my pole count setting in the Powervelocity programming app is set to 4. I believe it should be set to 8 for the small block motor. I hope that fixes my issue.

thx,
brian

I think it's only using the pole count to sense motor speed for the speed limiting functions. I don't think it will change the eRPM limitation. Worth a try for sure.

5400 RPM is pretty fast. If you have the same issue with two different controllers, then that would probably eliminate a defective controller. Another possible cause might be the response time of the hall sensors in the motor. Typical setup is there are pull up resistors pulling the hall signal lines high and the sensor pulls them down. If the pull up resistors are too high of a value, the signal might be responding too slow. It might be possible to just add some external pull up resistors to the hall signal lines to speed those up. With power off, motor halls disconnected, you can measure resistance from the 5v hall power line to any hall signal line going to the controller. You should be able to measure the pull up resistance directly this way. 5k would be a typical value. You could try placing 4.7k resistors from each hall signal to the 5v hall power line.

They might be just using crappy hall sensors in the motor too.

Thanks, I'll check the value of the existing pullups on the controller and pull them stronger if the pole adjustment doesn't work. Mike from LR has run the same setup(powervelocity controller/small block motor) and his is working fine at high RPM. I guess it could be a crappy hall sensor.

I'll report back after I test.

THX,
brian
 
No luck with the controller pole setting set to 8 poles. Next step is to take the CA3 out of the equation, and then a hall effect science project....uhgg.

Brian
 
Thanks for the info, I will be probably next to get one of those, they seem priced right for an step up to the OEM square taper controller.

G.
 
Back
Top