What causes wildly changing values on motor detection?

I’m a Vesc noob, and have wasted hours trying and failing to get a set of hub motors to not crackle and crunch at 50% or more throttle.

Every auto detection I ran would have different resistance, inductance, and manual detection would be the same. each time the values were way different, and i never got the motors to not crunch under 50% throttle.

Here’s two back to back auto detects:

I have another set of even junkier hub motors that I tried, and they actually seem to work OK on Vesc, though on manual detect would still offer different resistance each time.

I’d love to dial in the larger more powerful hubs but had to give up, or risk losing my mind.

1 Like

sometimes vesctool gets confused and you have to try again, your second go looks right.

also make sure you have the correct pole pair count, or number of poles I forget which it asks for

2 Likes

I would be even more concerned with those inductance levels measured. I wouldn’t use these unless you can get detection figured out, and not only look at resistance but make sure the other values measure reasonably close as well.

Not sure if this helps in your case, but I recommend trying to redo detection with every port emptied out other than the hall sensor ports. That can help sometimes.

The default value in vesc tool is probably wrong for hub motors. However this shouldn’t affect anything other than the speed that vesc tool reports.

1 Like

Do detection from the foc or bldc page and wizard auto detect isn’t nearly as accurate. With bigger motors you literally can’t use the wizard auto detect.

Only the sensor ports were filled.

On real time data, unloaded on bench, the MPH is way way off, at least double actual.

I believe i have 20 poles, was setting 10, not the default 14., and 10 inbthe ‘additional info’ menu too.

What motor. Maybe u have wrong amount of pole pairs.

I just went for a ride with my smaller vesc’d hubs which last night ‘worked OK’ but that was relative.

They are prone to making a different type of noise than larger hubs, more a muted crackle at higher throttle, but don’t seem to lose all torque when crackling, as do the larger.

On braking hard its like one motor is giving out and i can see and hear hanger twist underfoot.

I manually detected these, applied and wrote to both motors.

the smaller motors have 18 poles for sure, i entered 9. I can feel 18 detents spinning 360 by hand and a picture of the guts show 18.

The larger motors(puaida) have 20 detents, I enter 10. the translated instructions sat “the number of the rotor…20”

Both these sets of hub motors worked fine on a puaida/lingyi esc.
They Were Much quieter too. just weak brakes.

That esc started randomly disconnecting from remote requiring i stop and cycle the power button to continue.

I don’t see any larger motors on their site.
It’s not possible to tell how many poles a motor has feeling for cogging or indents. Maybe u can see inside to count magnets?

2 Likes

The puaida motors are just their 90mm ones. they do have supposedly higher wattage hubs with 110mm airless rubber wheels.

My smaller 83mm hub motors are exhibiting similar weirdness on vesc, and i know i have the pole pair number correct on these.

I’d love to get either set of motors working even 80% as well as they worked on a lingyi/Puaida esc.

I’m too frustrated, and confused, to continue trying to get this Vesc working with either set of available hub motors.

Press RL button
Resistance will vary from 102 to 220 ohms.
L from 220 to 493.
ld-lq from 226 to 1461 on back to back manual detections.

It says motors wont spin on Rl detection, but they do a whiny 1/4 turn.

On L, they spin up, yet sound like a can of rocks

I have no idea if it is a firmware issue, hardware issue, skill issue, or some combination of all 3.

Well, I know it is a skill issue, as I can read the help menu, and read the words, but they don’t register.

I was looking forward to tweaking and tuning motor performance, but I cant even seem to find a functional starting point.

Appreciate the help, but I’m waving the white flag, and exploring my other options.

I’m counting 18 poles, 20 magnets on these 83mm hub motors

Was setting motor poles as 9 on the wizard. eskate, medium outrunner.

How good are your solder joints? If one is loose or not wet out properly it could mess things up.

Get some tape and do one motor only with everything just taped to the deck. Test on garage floor. With one motor it’s easier to test. Get a good detection and settings figured out. If it doesn’t work and you get that cracking chudder badness - Change the vesc side but keep the same motor. Same chudder - change to motor. Something is busted it shouldn’t be this hard my man. Time to start eliminating things

Xt90 solder too. Sometimes iffy joints will make weird things happen.

Have you opened the hubs? I know you said one of the bolts is stripped? Is that the motor that is making noise? Open up the other one and make sure nothing is funky

3 Likes

The 6.6 came with MT60s on the 12 awg phase leads.

I’ve not popped off the cover and seen their solder joints.

I made MT60, to MR30 adapters with short lengths of 16awg silicone wire, I recall being pretty happy with the quality of those joints when i did them.

Ive not inspected the factory solder joints on the hub motor mr30s.
They did seem to separate far easier from, than Id like.

I guess i could cut off mr30s, and solder MT60s onto the 18 awg hub motor phase wires and eliminate the adapters.

Its so damn hot and humid in the garage, I need to set up a workbench inside The AC, but dont really want solder fumes in the house.

The Puaida hub motors, I’ve never tried to take apart. There was no reason to. They worked smoothly on the puaida esc. If it did not start disconnecting, I likely would still be using it.

The JK hub motors ive had apart a lot. One of their torx head m3 screws inhibit huge tightening torque, but removal is no issue. Power delivery on them was also smooth on the puaida esc. brakes being weak was only significant performance complaint with that esc on those motors.
The Puaida hubs I could feel the throttle ‘steps’ if accelerating slowly, and brakes were even weaker with larger diameter.

1 Like

Had similar issues with detection values, as above was caused by a poor solder join.

3 Likes

I was pretty happy with my soldering of 10awg loopkey circuit and XT90 onto the 6.6 mini 's 8 AWG, as well as the MR30 to MT60 adapter.

I was not happy at having to use an adapter, nor the loop they had to do to fit in the enclosure, but I wanted future adaptability to easily and quickly migrate components, and compromised.

We are Finally due to get some much needed rain soon, which should cool down my workshop, and give me time to tinker, with less guilt.

I’m thinking I should eliminate the mr30 to mt60 adapter.

I do have new MR 60s I can put on motors, and esc, which would allow a cleaner enclosure too.

Thanks for the input.
Much appreciated.

Then set it to 10 pole pairs and you should be counting magnets.

3 Likes

Ok. will be lowering white flag and trying again, relatively soon.

To clarify a bit, that picture of my hubs was not the 90mm sleeved puaida hubs, but the Jking 83mm hubs which were advertised as 350 watts

I’ve not yet taken the Puaida sleeves off, or the 400 watt Puaida motors apart.

They came with this documentation:

Their website is whack, but they sell these ‘450 watt’ hub motors too, with the airless 110mm rubber.

No idea about the actual size of those motors.

1 Like

that documentation is terrible engrish. “slots” is not the number you want. they have the pole count as “number of motor” which is actually # of magnets. that number divided by 2 is your pole pair count.

this explains literally all of your problems. use 10 in the pole pairs.

On the puaida motors, most detections were set at 10 poles.

Still the same huge variance on back to back auto or manual detection.

When I tried the JKing motors i set 9. as i counted 18 actual poles, and didnt know better.

I will proceed as if my issues are questionable solder joints and MR30 weaknesses.

Those high pitched noises on RL detection, I hypothesize that the weak sauce microbullets inside the mr30’s might be vibrating at high frequency too.

The puaida motors also detecting at 30 ish amps also tops out the mr30 rating.

Could my issues be summed up by a poor connection?

I’ll be switching to MR60 on all phase connectors

very normal for hub motors

very likely

1 Like