Stuttering Pulsing On Initial Start

[Braddudya]

I have noticed this on one of my fat tire ebikes as well as my just finished electric pit bike. What can be done to prevent stuttering on initial takeoff? My diy battery RadRover does this sometimes when I am putting a particularly heavy load on it at startup (full throttle, no pedel, uphill). I assumed it was just overloading the motor and or the battery not being able to supply the amperage required. With the new electric pit bike the batteries should be able to supply 200A momentary without issue. The controller should only draw something like 70a peak so I don't think it's a battery limitation. Is it the controller setup? It's a generic controller that came with the 3000w bldc my1020 motor.

I'd love to fix this issue as it makes the pit bike very hard to use and slow to accelerate

 

[999zip999]

Does it have Hall sensors ?
Check your hall wires.
Sensorless controller ?

 

[Braddudya]

Does it have Hall sensors ?
Check your hall wires.
Sensorless controller ?

The pit bike does have hall sensors and I did rewire it to get it turning reverse. I swapped blue and green on the hall wires and yellow green on the phase wires. This was the combination that ran the smoothest and turned the right direction.

I tried every combination of phase wires and all the other combos would just lock up or vibrate as you would expect.

 

[afzal]

Hall sensor signals being noisy due to current flowing in phase wires can cause this kind of issue (not saying that is necessarily the case here), if that is the reason, keeping hall wires away as much as possible from phase wires might make some difference.

 

[Braddudya]

Hall sensor signals being noisy due to current flowing in phase wires can cause this kind of issue (not saying that is necessarily the case here), if that is the reason, keeping hall wires away as much as possible from phase wires might make some difference.

This is quite interesting. They are all just run together in one wire sleeve from. The motor and I also have all the excess pength ziptied in loops on the frame. Right now it is just put together for testing. The loops of hall sensors wire and phase wire are probably a bad idea haha!

I'm not sure this is the issue but I will definitely try separating them. Thank you

 

[E-HP]

Try checking your phase wires to make sure the connectors are tight and making good contact.

 

[Braddudya]

Try checking your phase wires to make sure the connectors are tight and making good contact.

They are attached quite tightly with a terminal block. They should be good.

I opened up the loops in the primary and hall wiring. No difference at all. I can split the loom and try that but I am hesitant as it will make for a more messy routing of wires. Is it bad to run the phase wires with the halls in general? If it's bad practice I will absolutely separate them. They go into the motor together through one port so there is no way to full isolate them.

I have now tried a different combination of hall wires and phase wires that I found on another forum. This combination produces the exact same result. It spins backwards as required but it pulses like crazy on take off. The link below is a video of the symptoms.

https://streamable.com/98lx4n

The chain is not skipping. It does not exhibit this trait if there is no load on the wheel.

 

[999zip999]

Almost all almost every single one of e-bikes on the planet have their hallb wires and their phase wire in the same sleeve. Common practice sounds more like mismatch wires or loose wire connections the small little Hall wires could be in a connection without properly crimped in their pins.

 

[Braddudya]

Almost all almost every single one of e-bikes on the planet have their hallb wires and their phase wire in the same sleeve. Common practice sounds more like mismatch wires or loose wire connections the small little Hall wires could be in a connection without properly crimped in their pins.

That was my suspicion as well. As a test I reverted all of the wiring back. When I did this it spins backwards but it works much better. It was able to do a little burn out instead of just sounding like a rock in a clothes dryer. So that leaves two options. 1. My wiring is wrong. Or 2. This motor is not made to run backwards.

To see if it's my phase and hall swaps I wrote out every possible combination on a piece of paper and tried all of them. I ended up with a new phase wire swap and a new hall sensors swap which seems better but still has the same problem. Logically it's probably the same as it was before but it seems better somehow.

I am truly starting to think these motors just don't work well in reverse as ignorant as that sounds. I have done more research and found lots of people who can't get them to run backwards but they run great forwards. I saw a post on another forum suggesting that these motors have "hall advance" which makes them only work efficiently in one direction. If anyone has experience in this please let me know your thoughts.

It runs great once it's moving!

https://streamable.com/xr5ifx

 

[amberwolf]

Some motors have their halls installed in a physical place that advances or retards their timing from neutral. This means they won't operate the same reversed as forwards. It's not always intentional, sometimes it's just poor design or poor manufacturing.

It sounds like your motor is one of these.

There are two ways to correct it without replacing the motor, but still using hall sensors.

The first is to use elecronics like Burtie's Timing device (no longer made AFAIK) that can advance or retard hall timing to the controller (you can build one with an MCU with three digital input pins and three digital output pins, if you know how to program them (I don't, yet).

The second is to open up the motor, and see where they put the halls. Then determine which way they need to be moved in order to work better in the reverse direction, and move them there. (I recall a website or calculator that helped do this exists, but dont' remember where I saw it). (a variation of this is to just add your own hall sensors in the appropriate place, leaving the originals alone).

Another option is to use a good sensorless FOC controller that can correctly and easily startup from a stop under load (not all of them do; they may rely on the halls to do this startup and then run sensorless).



Note that this isn't the only thing that can cause a motor to not startup well under load. Insufficient power to get it moving for the specific load and conditions is a common one. A controller that just does a poor job of startup sequencing will do that as well. I have two different controllers on the SB Cruiser trike. One is sensorless generic trap, and one is an older Grinfineon sensored sine. If using by themselves, instead of together, the GF smoothly easily starts up as long as it has sufficient power to for the conditions. THe generic just doesn't startup smoothly even if I'm slightly pointed downhill on a driveway/etc., even if it's starting up at the same moment as the GF (which is smoothly starting) the generic often stutters or makes undue noise. It's just got a terrible startup routine. Another sensorless generic I have used has a much better startup routine and doesn't do that sort of thing, but it isn't nearly as powerful, so I only have it on the trike as a backup in case one of the other two fails.

 

[lnanek]

One thing to check is for the zero start options in the advanced settings of the LCD. Some bikes have that feature to prevent starting from a stop using throttle for safety. In which case you can turn it off.

Controllers can sometimes have this problem as well. There's a thread on here with someone with a Grin Tech Baserunner that can't start from a stop either. I have the Baserunner Z9 and experience the same. Have to push the bicycle forward a bit then hit the throttle.

This model controller attempts to use a sensored start then falls back to sensorless, but the logic is broken and if you are starting from a stop and don't get moving fast enough it thinks the sensored approach didn't work, basically. There's a checkbox in the Phaserunner software suite to disable fallback to sensorless, but I haven't tried it yet.

 

[Braddudya]

Thanks Amberwolf. That's what I was worried about. I found a listing for this kit that mentions the controller can run sensorless as well. Unfortunately it has no instructions so I do not know how to try running in the configuration.

I may try changing the hall sensor timing. Never tried anything like that before but it doesn't sound too complicated

Inanek - this system does not have any display. Just a 3 speed switch and a throttle. I will check for my rad power bike though. Thanks

 

[999zip999]

Maybe it's in sensorless mode for bad starts
What motor and controller, links ?

 

[Braddudya]

Maybe it's in sensorless mode for bad starts
What motor and controller, links ?

The kit was the Vevor 3000W bldc motor and controller. The controller model looks to be by24wf01-b. I googled it and didn't find any documentation. There is a loop connection off the controller that is unlabeled. I thought maybe it was the "smart wire" or "learning wire" but it doesn't seem to do anything whether it is connected or disconnected. Maybe it toggles the sensored vs sensorless modes? It doesn't change the behavior at all as I have tried it connected and disconnected. It arrive connected out of the box. There are no other wires in the kit which seems like they would toggle that functionality.

 

[amberwolf]

Sensorless mode is usually done by disconnecting the hall sensors completely. Then the controller "knows" there is no sensors because it shows them all at 5v (all off) which is a state you don't get (for the typical 120degree motor).

Easily tested, anyway, if you have a plug for the sensors you can just disconnect.

 

[E-HP]

That was my suspicion as well. As a test I reverted all of the wiring back. When I did this it spins backwards but it works much better. It was able to do a little burn out instead of just sounding like a rock in a clothes dryer. So that leaves two options. 1. My wiring is wrong. Or 2. This motor is not made to run backwards.

If you get it running smoothly in reverse, what happens is you keep the wiring the same and set the controller's reverse jumper?

 

[Braddudya]

Sensorless mode is usually done by disconnecting the hall sensors completely. Then the controller "knows" there is no sensors because it shows them all at 5v (all off) which is a state you don't get (for the typical 120degree motor).

Easily tested, anyway, if you have a plug for the sensors you can just disconnect.

I will try this! Thanks

If you get it running smoothly in reverse, what happens is you keep the wiring the same and set the controller's reverse jumper?

It goes in reverse but it is not full power. It would be interesting to see if it stops stuttering but I didn't test that. I just saw that when I jumped the reverse wires it would spin in the right direction but with very little power.

I am debating trying a new controller. Possibly something like a sabvoton which has smooth power delivery. I love this whole project except for this one huge problem. I can't even let other people ride it because everyone just goes too heavy on the throttle right off the line and it sounds like it's smashing itself apart. It really takes the fun out of it to baby it off the line everytime.

 

[999zip999]

Still need a real link it may give us a clue.

 

[Braddudya]

Still need a real link it may give us a clue.

Here are a few links.

This is the kit on vevors website
https://www.vevor.com/brushless-dc-motor-c_11227/vevor-electric-brushless-dc-motor-kit-72v-3000w-4900rpm-motor-controller-go-kart-p_010307620598

Here is the amazon listing I bought from (the price was lower at the time)
https://www.amazon.com/gp/product/B099WNGRVZ/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

 

[afzal]

I had a similar issue, it was resolved by filtering the hall sensor signal. The controller used was VESC, since it had the feature of hall signal oversampling, without any wiring circus, issue could be resolved.

If your controller has a configuration software & has option for hall signal filter, that also can be tried

 

[amberwolf]

If you get it running smoothly in reverse, what happens is you keep the wiring the same and set the controller's reverse jumper?

It goes in reverse but it is not full power. It would be interesting to see if it stops stuttering but I didn't test that. I just saw that when I jumped the reverse wires it would spin in the right direction but with very little power.

Some controllers have what is sometimes called a "trike mode" reverse, where they limiit power or speed to ensure it remains handleable by the rider.

Others literally just reverse direction.


With a motor that has advanced or retarded hall timing/placement, either one will not run as smoothly as forward, unless the controller knows what this timing difference is (from neutral) and knows how to compensate for it.

I am debating trying a new controller. Possibly something like a sabvoton which has smooth power delivery. I love this whole project except for this one huge problem. I can't even let other people ride it because everyone just goes too heavy on the throttle right off the line and it sounds like it's smashing itself apart. It really takes the fun out of it to baby it off the line everytime.

I don't know if the sabvotons qualify, but an FOC controller with well-designed firmware can operate a motor very smoothly, once it is tuned to the motor--but you *have* to setup the controller for the specific motor parameters (hall timing, phase resistance, inductance, number of poles, kV, etc) for it to work correctly. Some controllers have a good autotune, but even with those you sometimes have to tweak things for best operation.


But...it might be your motor itself causing you grief rather than the controller, if it does have non-neutral hall timing, and that may cause the same problem with any other controller that expects neutral timing and doesn't have a way to compensate for it.


Anyway, it hasn't been established yet if the motor does have neutral timing or not...or if it is some other problem with the controller or even just sensor signals. I've been trying to think of a way you could easily test the timing without an oscilloscope or building a test rig....

 

[E-HP]

Still need a real link it may give us a clue.

Here are a few links.

This is the kit on vevors website
https://www.vevor.com/brushless-dc-motor-c_11227/vevor-electric-brushless-dc-motor-kit-72v-3000w-4900rpm-motor-controller-go-kart-p_010307620598

Here is the amazon listing I bought from (the price was lower at the time)
https://www.amazon.com/gp/product/B099WNGRVZ/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

The Vevor website shows two diagrams for that controller:

https://www.vevor.com/brushless-dc-motor-c_11227/72v-3000w-bldc-motor-kit-with-brushless-controller-electric-bicycle-scooter-p_010187167720

https://www.vevor.com/brushless-dc-motor-c_11227/vevor-electric-brushless-dc-motor-kit-72v-3000w-4900rpm-motor-controller-go-kart-p_010307620598


Which one matches yours?

 

[amberwolf]

https://image.vevor.com/us%2FZLDJWS3KW72VWZDKZV9%2Fgoods_img-v9%2Fbrushless-motor-f4.jpg?timestamp=1660814248000


https://image.vevor.com/us%2FZLDJWSL72V3KW2BSJV0%2Fgoods_img-v4%2Fbrushless-dc-motor-f4.jpg?timestamp=1631496711000

 

[Braddudya]

Still need a real link it may give us a clue.

Here are a few links.

This is the kit on vevors website
https://www.vevor.com/brushless-dc-motor-c_11227/vevor-electric-brushless-dc-motor-kit-72v-3000w-4900rpm-motor-controller-go-kart-p_010307620598

Here is the amazon listing I bought from (the price was lower at the time)
https://www.amazon.com/gp/product/B099WNGRVZ/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

The Vevor website shows two diagrams for that controller:

https://www.vevor.com/brushless-dc-motor-c_11227/72v-3000w-bldc-motor-kit-with-brushless-controller-electric-bicycle-scooter-p_010187167720

https://www.vevor.com/brushless-dc-motor-c_11227/vevor-electric-brushless-dc-motor-kit-72v-3000w-4900rpm-motor-controller-go-kart-p_010307620598


Which one matches yours?

Its the lower one. The one with no sticker on the top in the photo. The "power output" and "battery output" wires are the ones I was referring to that don't seem to do anything and arrived connected out of the box.

I did just try running it without the hall wires connected. It did nothing without them so my guess is that listing was just inaccurate about the controller being able to run sensorless.

 

[Braddudya]

I would love to find a way that I can test the motor timing to rule it out or in as the culprit.

I do have a Vector 300 controller from Golden Motor plus a programming cable. I have never messed with it yet, I got it second hand buying someones unfinished motorcycle project. I could try configuring that controller to work with the my1020, though I am not sure how hard that is since finding accurate specs for the motor may be impossible.