Very frustrating trying to problem solve. I’m still not sold on HV setups. Seem to be less reliable than 12s from what I’ve read and experienced (although i only worked up to 16s).
I really hope you get it sorted!!! 
5 Likes
yeah thinking about this idea it’s no diffferent that max motor amps. vesc can’t respond fast enough to the current spike, that’s why we’re getting the abs fault.
good on you for doing the experiment tho.
3 Likes
guessing wildly i wonder would switching frequency increase fidelity on current sensing and therefore help here?
1 Like
Yeah, I am curious why the later spike though. There’s definitely no wheel spin happening.
1 Like
That’s what I hoped the time constant thing would do, but what do you mean by switching frequency? Like Bldc?
I didn’t understand the time constant thing, if it’s at detection time i wouldn’t think it would affect runtime other than perhaps getting some more accurate detection in the first place of R and L values?
Switching frequency i’m also a little fuzzy on. I think when those words are used they mean the setting “zero vector frequency”
it’s what shaman had me enable.
and I think it prevents losing track of the motor at higher rpms? … but it’s on my todo to understand bettter.
if it’s what I think it is. it’s increases the “sampling rate” the wave form being generated on the phase wires. makes it higher resolution. at the cost of switching the fets at a faster rate.
so my wild guess stab in the dark idea is that the loop that checks current and adjust dutycycle would also increase frequency to match the new switchign frequency? but… idk… i’m pretty low confidence i understand this bit right.
1 Like
I was mostly hoping to cunninghams law my way to an answer for you. 
Ok so you already have it enabled though? If you learn more I’d be interested otherwise I guess I’ll try to dig as well. But seems straightforward enough, higher resolution in order to prevent the spike.
yah assuming it’s higher resolution of the measurement and control loop. but idk that it is.
and yeah I have it enabled but haven’t gotten back to really pulling on v5#1 since the change. also it’s 2wd so i don’t think i have the wheel slip related spikes.
I think the wheel skip spike is fine with TC, but I’m going to keep this in mind for the other spike if max wattage doesn’t help. Thanks.
For all the days it’s not working right I’m just like what am I doing 
If it makes u feel any better my 4wd was flipping out on me last run i had. My motors starting squeeling at me and kicking in late. One side of dv6 had gone back to default settings 
. Then took a few attenpts at motor configuration again to get everything working properly. Very weird. I’m wondering if my two seperate loop keys plugging in seperately is catching up on me and fucking with Canbus.
In other news it’s hard to tell but the Hubba-hubs flare out ~0.7-1mm more than the Newbee pentagrams hence the vibrations my guess.
I’ll probably print a wedge and see how that goes.
4 Likes
This is the one found under FOC?
According to the note it’s apparently only available on hardware with phase shunts and from what I remember only the Trampa ones have that.
Thats called a draft angle for injection moulding.
2 Likes
Oooh TIL. So was it more a necessary thing with the mould process (i.e. not necessarily planned)? Either way it helps a lot in this case with a tighter fit.
3 Likes
Yes, with injection moulded parts a 0.5 to 3 degree angle is needed so the part can fall nicely out of the mould. So this angle is indeed necessary.
2 Likes
I’ve used a layer of gorilla duct tape to fill out dimensions b4 with good luck.
3dp maybe better, but when its just a bit needed i doubt a 1mm 3dp would hold up
That’s a great idea thanks! Yeah I was also kind of afraid something that thin wouldn’t hold up as well.
1 Like
Ive come to the conclusion that some tires that are deemed shit is actually bad ID/OD fitment and can be rectified with some fine tuning.
2 Likes