VESC - Land of 1,000,000 choices

The biggest thing I'm learning - and what you're alluding to - is to STAY OUT OF SATURATION. That's where bad things happen, as we've all discovered by watching things explode/catch fire/etc. That's not the only cause of the failures (of course), but that's the point where efficiency seriously drops off, power consumption shoots up and things get unstable. So what you're thinking (go big on the motor) is the way to go.
 
AlexLTDLX said:
The biggest thing I'm learning - and what you're alluding to - is to STAY OUT OF SATURATION. That's where bad things happen, as we've all discovered by watching things explode/catch fire/etc. That's not the only cause of the failures (of course), but that's the point where efficiency seriously drops off, power consumption shoots up and things get unstable. So what you're thinking (go big on the motor) is the way to go.
Click to expand...
is this posted against the wrong topic ?
 
I dunno - maybe. Somewhere I believe you mentioned getting a motor a bit oversized; I'm about to post the first P2 runs vid in the P2 thread...
 
Somehow I have the feeling a sensored 2pole LMT3080/6 875KV (~1000KV) would play nicely with the VESC. Much better than the TP motors ...
(2pole and sensored both would unload the compute requirements quite a bit ... and the LMT3080s with 30kW peak will do easily 15kW work far from saturation )

The MGM controller is more expensive yes, but then: it is a proven combo with LMT motors (and their weird 6 connector phases).

Sooo ... as much as I like the VESC so far ... it will probably shine with a different motor.
And that different motor will shine "safely" with the other controller ... shux ...

edit: The more I read about hot 6pole D wind TP motors the more it becomes clear: TP motors on high voltage (>12s) are GENERALLY hard to control ... peak currents shoot up due to D wind side effects, 6poles are hard on compute/control without sensors. They challenge even the MGM controllers, which have sophisticated protection mechanisms in place (not frying them necessarily, but they get "limited"). And when they work with MGM controllers they need some exotic settings like borderline advanced timings and accelaration torque limits ...

So ... step one for me : sell my (unused) TP power motor 1000KV motor!

@AlexLTDLX I predict you might face some issues with the VESC (even) down the road if you keep winding up that TP motor higher...
With all your frying experiments there was always one invariant: TP Power
 
Last edited:
Like I said in the YT comments, I want to make some test this winter.. I intended to make the tests this summer but I don't want to mess with the set up to go in and out of the car.. the test you've made is super simple and can solve the problems I have! It's nice to see the same project but different set up and combo like @cmoalem talk!
 
It's still -25°c the morning (-13°F) but the summer is really soon! Tomorrow is my birthday so in 2months I will slowly put the turbo back in the car! Still need to work with the all cherry sensor.
 
cmoalem - We'll find out soon enough. This morning I ordered some XT90 connectors to test the VESC/P2 setup with the LTO packs and in-car circuitry. FWIW, the MGM/LMT combo I have coming is sensored. In fact, they emailed me last night telling me the motor was on the way to them and they install the sensor.

WB - Happy birthday! I agree and am glad you can use some of the info. And that's one of the reasons why I went ahead and built the P2 unit - so I can see (and maybe work out some issues) so that other folks don't have to do the guesswork. To be honest, I really enjoy that. Even though with all the fires and explosions I'm a bit nervous - but that makes it even more exciting.

BTW - The belt doesn't seem to stretch much. It did stretch a bit at first, but not so much now. Of course, that could change at higher rpm and higher load.
 
FWIW, VESC is on one hand, a great learning tool. On the other hand, there are so many poorly or not documented things that I'm literally spending hours trying to solve mysterious issues that could be solved with 1 centrally located, evolving document with proper links. By way of example, I'm trying to solve an issue where my battery current draw seems to be limited to 150 amps. Only recently did I discover that there's a "no hardware limits" version of the firmware; but nowhere can I find what those limits are. Even more aggravating, I STILL don't have a clear idea on how to install any firmware except for what it auto suggests. Of course, it certainly doesn't automatically suggest firmware with no hardware limits...

Be aware, if you go the VESC route, you're going to spend literally DAYS or WEEKS just trying to get your motor to turn properly. It's a good learning experience, but not good if you just want to get up and go.
 
oh ... and one more thing @AlexLTDLX : what is your current setting for switching frequency? ( in FOC mode a higher value than 20kHz (40,60) seems beneficial for the top speed as well! in DC (BLDC) mode going higher than the default 20kHz needs possibly good cooling of the powerstage!)
 
I'm going to look into that. Somebody on the VESC forum sent me a video Benjamin Vedder did about manually dialing in a motor. And ultimately, he ended up doing exactly what I did - doubling the observer gain. But I've still got this mystery 150 amp battery limit. I should finish building the blast shield today, and then I'll be able to test some things. My pullies/belts should be here this week. I'll start with the 16 tooth motor pulley. Right now, I think there's something in the firmware possibly (there is a no limits firmware you can install), or maybe Frank's right, and it's just torque limited. But it doesn't look like it because I was able to go over 150 battery amps before doubling observer gain (I got to 160 amps before I started getting current spikes).

At this point, consider VESC a work in progress - you can certainly make it work, but it's far from plug and play. Or even plug, run some wizards, and play. It'll get there most likely.
 
There is much more than observer gain. You will need to play with it up and down significantly 60-2000
There is also motor flux linkage and inductance that may need to be moved up to 20% each way. Ensure absolute current limit is reasonable and go up in stages of 20 phase amps.
I have a VESC in a much different application and thus far no real problems and it puts out 10kw..
 
I have only used 130 battery amps so far because the gear ratio is extremely low.
I'll turn it up significantly next week after the next sprockets arrive. I am using a VESC 100/250 so less amps than you will be able yo run.
 
So I'm still having issues. Jacking up observer gain let me get to 95,000 eRPM (24,600 impeller RPM) with the 1" restrictor in place, but I'm getting current spikes again. I've tried re-running motor id, played with all kinds of observer gain settings, but no luck. I've been at this for a few weeks now, and am making progress, but it's frustrating. cmoalem's suggestion, along with disabling a couple of things, are next on my list.

FWIW, I've started a thread on the VESC forums about this: https://vesc-project.com/node/3436#comment-9407

If I could've just run it out of the box with a few tweaks, it'd be on the car by now.
 
Two nights ago I was able to get things up to about 125,000 eRPM. I raised the frequency to 40kHz, and played more with observer gain. We hit 94.4% duty cycle (max is 95%). This is with the 1" restrictor in the discharge. The motor pulley and the belt get the hottest. This was with a 14 tooth motor pulley and 18 tooth impeller pulley. Impeller RPM is 32,400. Now I just need to make sure it's repeatable and can accelerate quickly - then time to install it.
vlcsnap-2022-03-02-22h03m56s360.jpg
vlcsnap-2022-03-02-22h06m08s372.jpg
 
I think i'll have to get one of these at some stage.. as these seem to be the cheapest with fantastic features without exploding
 
You must log in or register to reply here.
Top