VESC-Tool 2.04: FW 5 >> A BIG STEP FORWARD

It’s called Sensorless ERPM HFI in the Motor Settings > FOC > HFI screen

Lower that to 750

You can thank me later

1 Like

Synopsis time:

what’s it do? does it make the magic music last longer?

It just turns off HFI really quick once the rotor gets turning. In fact you can lower it even more than that, try 500. It only needs HFI when it’s not moving at all.

Once it turns off, you can dump as much power as you want into the windings. So basically it increases the window during which there is no cogging risk for extra high motor current levels.

5 Likes

THIS

interests me… thanks I’ll giver a go this weekend an report back if I’m a fan of your unfound gem!!!

thanks meng!!!

1 Like

A post was merged into an existing topic: Derail Jail 2020_May

Just wanted to let you guys know, the stuff regarding max duty cycle is on Vedder and My radar.

Vedder has pretty good theory on what is going on, he made the maximum duty cycle limit work better which is somewhat ironic. Basically before as you approached max duty it wouldn’t correctly prioritize Q-axis current (the current that makes torque) over D-axis current (no torque). Now it does this more efficiently which is good, the problem is it causes a more abrubdt feeling torque loss when the limit is come up against and can cause oscillations with some setups since it is a rapid change. The solution is to soften the change as per the setting @Trampa is pointing to.

I’ve talked with Ben about the best ways to add this into the wizard for next release. It does cause a slight decrease in top speed power in the last 5% so people on e-bike or expert users and other applications might not want this limit imposed without being informed.

13 Likes

I have that limit set to 85% on all 4 escs, but still getting oscillations on profile limit 40km/h. When I set 30km/h it’s ok. My setups max should be about 50km/h.

What if it’s also caused by every motor having slightly different erpm? Like when wheels have different pressure, so on the same speed you get little different rpms. But traction control erpm difference is at 3000 erpm…

I just lowered mine from 2000 down to 1000 and it was all sorts of fucked up. What works for you might not necessarily be best for everybody else.

Has anybody else experienced weak torque on startup with HFI? My board only delivers torque once it is well and truly rolling, no off the line punch. Funnily enough tho, in ‘current’ mode, reverse has heaps of punch :thinking:

I tried inverting motor directions with the same result. Forward sucked, reverse was great.

Wtf?

5 Likes

Well the solution is right in your face. Change stance and use Reverse as forwards, forward as reverse. duh.

2 Likes

That’s a work around, not a solution.

3 Likes

I’ll see if I can get similar behavior this weekend.

1 Like

You may want to change the % limit for the start also then. These limits are essentially P controllers which can become unstable in certain setups.

2 Likes

Personally I wasnt satisfied with the workaround duty cycle current limit start “fix”. It made the full throttle motor wobbling issues go away but my top speed was lacking. It felt like my board was sagging while reaching top speed.

I understand that’s what the duty cycle current limit start is doing but I couldn’t continue to ride with it unfortunately.

Major bummer because I was really liking the profiles feature.

Speed limits, set in the profiles, automatically apply a duty cycle current limit.

1 Like

https://digitalcommons.unl.edu/cgi/viewcontent.cgi?article=1343&context=electricalengineeringfacpub

Interesting paper I found. It might provide an alternative HFI strategy that allows for greater power flexibility.

I just tried on a new skate, and the HFI sucked ass at any kind of low setting.

I think it might be duals, maybe duals have to be high.

The single drives, you can put it at like 750erpm and it kicks ass

2 Likes

Hey @deodand what phases are the vesc current sensors on?

All :thinking:

Three shunts, three meassurements, one per phase

1 Like

How’d you go over the weekend mate?

@kalebludlow
4.x hardware only has shunts on phases A and C I believe

And during HFI and low modulation it only uses the measurements from two phases, can’t remember which two, it’s somewhere in the code.

3 Likes