Make Motor Run Quieter?

[spooknik]

Hello!

I am wondering if anyone knows if I can quiet the motor noise on my e-Bike. I have an e-Bike with Shengyi CMT03, it had an inbuilt controller but it was not possible to program the controller and not the best design. I swapped to a Baserunner and CA3 and everything has been working great for around 6 months. You can see more details in my post history.

Since the swap to the Baserunner the motor is much more noisy. To the point where the neighbors comment on it. It was never whisper quiet, but it was much quieter on the stock controller. Is there a parameter I can tweak to get the motor to be more quiet?

Here is a video of the Phaserunner Suite Autotune running and the motor noise. It seems electronic to me and not mechanical.

Here is a video of the motor just being turned by hand. Everything feels fine by the way, there is no resistance other than the gears.

I attached the settings I have used on the baserunner. The motor settings where found via the Autotune (I know the battery limits a set weirdly).

 

[PaulD]

I would double check your Kv and pole count numbers. Are you calculating Kv based on the crank rpm? what is the gear reduction from motor to crank?

 

[spooknik]

I would double check your Kv and pole count numbers. Are you calculating Kv based on the crank rpm? what is the gear reduction from motor to crank?

Thanks for the reply!

The KV is what the Phaserunner Suite autotune gives me after the spinning test. I am not sure how to calculate this number manually.

Similar story with the pole count, the gear reduction kinda confuses me. As I understand it, the formulate is pole pairs x gear ratio. I counted 9 pole pairs when I had the motor open (picture attached) and the gear ratio is 1:31.1 (stated by Shengyi on the product page). So it would be 9 x 31.1 = 279.9? Again, I just relied on the number the autotune spits out which is 132.

 

[PaulD]

That picture is the stator. You're counting the number of slots, not poles. Look at the rotor and your number of pole pairs is half of the number of magnets on the rotor.

 

[amberwolf]

The KV is what the Phaserunner Suite autotune gives me after the spinning test. I am not sure how to calculate this number manually.

kV is RPM per volt.

So, if you have say, a 36v battery, fully charged to 42v, that sags to 41v during this test, and run a motor at full throttle off ground and get a motor RPM of 400, you would divide 400rpm / 41v for a kV of 9.75.

If there is gearing between the motor and the RPM sensor, you have to also do the math to convert the RPM you read up to the actual motor RPM, so if the wheelspeed was 400RPM, and the gearing was 31.1, you'd have to multiply the 400RPM by 31.1 first, to get 12440, and *then* divide that by 41v, to get a kV of 303.41.

That's all the autotune is doing, is counting based on the other info provided to it, and doing that math.

Similar story with the pole count, the gear reduction kinda confuses me. As I understand it, the formulate is pole pairs x gear ratio. I counted 9 pole pairs when I had the motor open (picture attached) and the gear ratio is 1:31.1 (stated by Shengyi on the product page). So it would be 9 x 31.1 = 279.9? Again, I just relied on the number the autotune spits out which is 132.

The autotune doesn't really know--if you provide the correct gear ratio, and there is an actual wheelspeed sensor (not just one of the motor halls) that you provide the correct pole count for (usually six, sometimes one) it can count that sensor's pulses and the motor hall pulses, and do it's own math for deriving the pole count.

If it doens't know the gear ratio, or the wheelspeed, or both, it can't definitely determine the pole count, as it has no way to know when the rotor has completed one revolution past a particular hall sensor or tooth, etc.

 

[amberwolf]

Since the swap to the Baserunner the motor is much more noisy. To the point where the neighbors comment on it. It was never whisper quiet, but it was much quieter on the stock controller. Is there a parameter I can tweak to get the motor to be more quiet?

Do you get better accelleration with the BR vs the SY?

If so, is the BR's current a lot higher than the SY?

If so, it may just be pushing so much harder on the gears to do this that it makes more noise.

To test this, change the BR's current down to what the SY used, and see if it makes it quieter.


(on my GMAC and PR6, when the motor is powered either in braking or acceleration, the gearing is much noisier than when it is unpowered, even at the same wheel speed).

 

[spooknik]

kV is RPM per volt.

So, if you have say, a 36v battery, fully charged to 42v, that sags to 41v during this test, and run a motor at full throttle off ground and get a motor RPM of 400, you would divide 400rpm / 41v for a kV of 9.75.

If there is gearing between the motor and the RPM sensor, you have to also do the math to convert the RPM you read up to the actual motor RPM, so if the wheelspeed was 400RPM, and the gearing was 31.1, you'd have to multiply the 400RPM by 31.1 first, to get 12440, and *then* divide that by 41v, to get a kV of 303.41.

That's all the autotune is doing, is counting based on the other info provided to it, and doing that math.



The autotune doesn't really know--if you provide the correct gear ratio, and there is an actual wheelspeed sensor (not just one of the motor halls) that you provide the correct pole count for (usually six, sometimes one) it can count that sensor's pulses and the motor hall pulses, and do it's own math for deriving the pole count.

Thanks for the good explanation!

My eBike is a middrive and I don't have a throttle for the speed, just toque sensors on the crank for PAS (set up in CA3). So I am not really sure how to get the RPM of the motor at full speed in a testing environment. However, when biking CA3 reports that the motor runs about 50 RPM when I go full speed, but I am not sure if this number is reliable or not.

So if we take 50 rpm x 31.1 gear ratio = 1550

1550 / 41v = 37.9 kV

Is there any real risk to the motor if I guess the wrong kV setting? It's currently at 5.31, so should I just set it to 37.9 kV and let the autotune run?

Do you get better accelleration with the BR vs the SY?

If so, is the BR's current a lot higher than the SY?

If so, it may just be pushing so much harder on the gears to do this that it makes more noise.

To test this, change the BR's current down to what the SY used, and see if it makes it quieter.

Yes, and yes. I increased the power limit from 250W to 400W.
I tried to decrease the power limit to 250W but it sounds the same.


I would normally also wondering if the controller is doing something strange with the sine waves and giving the motor 'unclean' power, but since it's a Grin controller, it shouldn't be the case.

Update: Looked the motor's product page again. It lists the RPM as n0 (Rpm): 90 and nT (Rpm): 80. So that might be worth trying as well.

 

[amberwolf]

My eBike is a middrive and I don't have a throttle for the speed, just toque sensors on the crank for PAS (set up in CA3). So I am not really sure how to get the RPM of the motor at full speed in a testing environment.

For a "bench test" (vs road test) you can get the wheel offground, and set the CA's throttle menu to the mode "Disabled WOT" which will provide full throttle output to the controller. To stop this you have to go back into setup and that menu and change the mode back to whatever mode you normally use, which will take time, so be prepared for this and have the bike be stable and secure in case you have a lot of wheel wobble. If you have ebrakes connected, you may want to have your hand on one ready to engage it.

However, when biking CA3 reports that the motor runs about 50 RPM when I go full speed, but I am not sure if this number is reliable or not.

If the RPM is reported by a wheelspeed sensor, rather than a motor hall, then you'd use whatever gearing is between the sensor and the motor, and however many poles the sensor itself uses (usually six, or one), as multipliers for that RPM.

Is there any real risk to the motor if I guess the wrong kV setting? It's currently at 5.31, so should I just set it to 37.9 kV and let the autotune run?

As I understand it, it is better to start with something close; the test uses this to base it's starting point on. If it were truly completely automatic, it wouldn't ask for a best-guess number, it would just figure it out on it's own.

Yes, and yes. I increased the power limit from 250W to 400W.
I tried to decrease the power limit to 250W but it sounds the same.


I would normally also wondering if the controller is doing something strange with the sine waves and giving the motor 'unclean' power, but since it's a Grin controller, it shouldn't be the case.

if the motor characteristics are wrong in the controller, then it can certainly drive the motor wrong, which can cause all sorts of wierdness.

It uses those characteristics to do math on to figure out what the feedback it gets from the motor means, for any given signal it sends. If the characteristics are wrong, the results are wrong, and the signal adjustment it makes will be wrong.