Surging under PAS but not throttle?

[kiltedcelt]

So I noticed something kind of funky today with my MAC geared front hub motor. When pedaling I could hear sort of an oscillation sound coming from the motor, and even looking down I could see the power must've been coming uneven enough that the fork legs were showing a slight fore/aft flexing that seemed to coincide with the oscillating sound. Now if I stopped pedaling, and simply applied throttle the motor was just spinning and humming along with no noticeable oscillation, either audible or visible, even with trying to limit throttle output to the same watts as what I was getting from the current setting on my multi-position switch/CA V3. A few stats - 52v EM3EV battery, 25amp Infineon controller, MAC 14T geared front hub motor, Cycle Analyst Version 3 with most recent 3.1b upgrade, thumb throttle, 12 pole cadence PAS, and last but not least, the new-ish Grin 2-button multi-position switch which I have configured to offer assist from 50 watts all the way to 500 in 50w steps. Final note - I think I've sort of noticed this oscillation before but it was only momentary and I wondered if it was something to do with the clutch engaging/disengaging, and kind of wondered if maybe I should crack open the motor and see what shape the gears are in. The motor has probably 2000-3000 miles on it at this point. The other possible factor is that the battery was pretty close to being exhausted - it's been EXTREMELY windy here the last couple weeks and so I tend to run the assist at around 350 - 500 watts assist a lot when I'm either fighting those winds going in to work or coming home. So, coming home this evening I was on the last "fumes" of the battery charge, at least by going off of the less-than-accurate battery life icon on the CA display. This battery has 28.3Ah, but I've been getting quite a bit less because probably cold weather and lots of extra watts than I typically use for assist when the weather is nicer/less windy. The most I've gotten recently mileage-wise was about 80 miles, but I know this system as I have it configured can deliver more than 100 miles of range if I'm fairly conservative with everything. So my thought is: Was something funky going on due to the battery power getting low (that lame battery gauge showed about two pixels left while the Ah used was only about 20 - should have another 8 Ah available?), or, OR is something squirrely going on because I need to make some sort of adjustment to the "Gain Adjustments: Minimizing Surging or Speed Oscillations" as explained in the CA guide? Also, adjusting the Gain Adjustments seem to require having the bike raised up in some sort of stand so you can try to recreate the issue and make adjustments on the CA. Is this the case? I'm sure I can figure out a way to elevate the front hub motor/wheel while raising the rear as well so I can pedal the bike and get the PAS to work so I can recreate the issue. One last reiteration - Pedaling under pedal assist = surging, while throttle only/no pedaling = no surging.

 

[amberwolf]

Final note - I think I've sort of noticed this oscillation before but it was only momentary and I wondered if it was something to do with the clutch engaging/disengaging, and kind of wondered if maybe I should crack open the motor and see what shape the gears are in.

If there's no problem while on throttle, and the throttle has the controller providing the same kind of power that PAS does, then there's nothing wrong with the motor.

The other possible factor is that the battery was pretty close to being exhausted - it's been EXTREMELY windy here the last couple weeks and so I tend to run the assist at around 350 - 500 watts assist a lot when I'm either fighting those winds going in to work or coming home. So, coming home this evening I was on the last "fumes" of the battery charge, at least by going off of the less-than-accurate battery life icon on the CA display.

The most likely thing I would imagine based on this is that the PAS is setup in a way that causes surges of current, which under normal conditions aren't noticeable, but at low SoC causes enough voltage sag to get the controller or the CA to stop applying power then reapply it and repeat taht, causing the surging feeling.

If this never happened before, even under these conditions, then it's possible the PAS control is not working the same way it used to, caused most likely by a connection issue. Corrosion between contacts, or bad crimps, etc., could be causing the PAS signal to be different than it should be, causing the CA to output differently than it used to.


If there was a problem with the battery itself (cell going low causing BMS shutdown) the CA would power off along with the rest of the bike. It wouldn't just surge. And it should do the same thing via throttle on hard acceleration--shutdown.


FWIW, I can't view whatever you linked to--if it's a picture you should just attach it to the post, so anyone that can see teh post can see the image.

 

[kiltedcelt]

Thanks for the input. Everything is set up in such a way that the Cycle Analyst controls everything, so you are correct - the throttle and PAS both are getting their input/output the same way. With a full charge for this morning's ride I'll have to see if it does it again. If not, maybe it was a SoC issue with the cadence sensor as you'd suggested. As for the image, that was a screen grab of the pertinent "Surging" section from the CA manual. I see it on my end, wonder why not on yours?

 

[justin_le]

Try reducing WGain. On high torque motors the default value of 15 is often too high and values more around 6-10 are better behaved for constant power feedback.

Or, if you are using the 3.1 release firmware, you can also change your PAS mode from being a constant power output mode to a constant throttle output mode. Then the PAS control is identical to running with the normal throttle and no feedback loop tuning is needed. In this case the different PAS levels will affect the motor speed rather than the motor power.

 

[kiltedcelt]

Justin - thanks for the info. After riding with a full charge I'm pretty sure that the symptom was something to do with either the W-gain like you mentioned or maybe tweaking the PAS multi-position switch. The two-button switch (and the 3.1 CA software), I've got configured to increase power in discreet 50 watt steps. This MAC motor is SUPER-high-torque as well. I'll play with those settings and see what happens. I'll try the W-gain first, as I like having the two-button set up to give those incremental jumps in wattage.

 

[royco]

This happened to me while riding this morning. Turned out, the distance of the PAS to the magnets on the disc was varying due to a not so secure mounting. Pushed the sensor closer to the disc to cancel out the variation and problem went away.