VESC - Land of 1,000,000 choices

AlexLTDLX

AlexLTDLX

Administrator
Staff member
I haven't even powered the thing up yet. To be honest, I'm scared to power it up. The documentation is sparse, information that is pertinent is hard to come by, you have to spend DAYS parsing forums and guessing. Yes, it may be powerful, but since many settings are followed by the warning: "may damage the VESC or motor" and you're left guessing - yes, literally guessing - like, "select the motor that's closest to what you've got..." and there's a bunch of pictures of motors. But you guess the wrong one, and BOOM.

I love the warning on the top of it: "Read manual before connecting device." I've read the manual. It's short, generic, worthless and passes the buck to the web.

I'm having nightmares of reliving my Megasquirt MS3x experience - powerful hardware, horrible documentation, and the Megasquirt didn't even work right; specifically the VR pickup input. I ended up designing my own circuitry to get it to work. A great product ruined by horrible support and shitty people on the forums (they loved to tell you it was always YOUR fault, when they're running an 80 hp Volvo. And they were always wrong. Every. Damn. Time.)

So far, at least Frank has been super helpful, but I feel bad constantly pinging him via email for what are in my mind simple questions that no documentation anywhere covers. Not to mention it takes forever (he's in Germany).

Why, oh why, is it so hard to make a plug and play high power ESC?

Thanks for letting me rant. A video on this thing will be coming soon. Probably featuring a similar rant.
 
I have spent more time making cables to match particular ESCs than the ESCs have ran - in total. Hopefully, everything works out this time around.

I'm off to make some cables. Again.
 
Great news! I just wrapped up the first VESC test with the TP Power 5780 motor. I was only running one 4s pack, but this was a "get to know you" run, and I spent a fair amount of time just finding stuff in the software.

Video coming in the next day or two.

Next up, while I wait for the MGM/LMT stuff (I'm starting to get a bit concerned, but hopefully they get back to me soon); I'm going to start sneaking up on 16s/full power while the motor's bolted to the Vortech...
 
great news ... so you dug through all/most of the settings jungle on the VESC software.

Particularly interesting this VESC topic. (and once you have the mgm/lmt stuff you would be the first to have both "kits" on my short list 😛)

Btw: MGM do not list the biggest motor LMT 30100 and the biggest 800 amp esc on their website anymore! (you see them only if you know the direct link!)
I hope you are not affected ...
 
Hello guys, we have hid the ESC from our website for now because there's a delay in the delivery of some parts from our supplier. But it will be up for sale in a few weeks from now. We had reserved Alex one since we started discussing his project a while back so he got one of the last pieces we had in stock. We will keep you updated.
 
Thanks Marek! I asked the same question via email, but you answered it here. I'm liking the VESC, but I'm also looking forward to the brute power of your controller - and the ease of use. VESC is a pretty complex beast to implement. Your stuff should be much closer plug and play. It's like the difference between a computerized scalpel and a sledgehammer.
 
AlexLTDLX said:
Click to expand...
yeayyyy! GREAT success!
I could see the sweat pearls on your forehead at some points in time though (especially when the motor autodetect started). And I could see as well the non linearity of the "throttle" so many people had to overcome 😁
It was very cool to see 30000 erpm (10000 motor rpm) with just this relatively little ampere draw on lipo 4s!!! (yeah I know ... no real load on the motor, but still)
Now when you will "feel" your way forward exploring/extending the limits it will become even MORE interesting.
 
isn't 300amps to low ? Also is is 300amp that it can actually run at or is that really a 150amp ESC or is that the point of the testing
 
From what I know of it and from what Alex has said it is rated at 300 amps but has 2 or 3 times that in mosfet capacity built in.

But only time and testing will tell the truth of the matter. Can't wait to see all of Alex testing brings.
 
That's what Frank at Trampa told me - 900 amps worth of mosfets in the thing. Also, they have a go kart that's hitting 30 kW peaks with it:



If only those boys knew what 800+hp feels like on a dead hook. Heck, even 650hp compared to a kart... but still does speak volumes about the VESC when tuned properly.

It also shows how much power an electric turbo really needs...
 
At this point, I don't think I can really recommend this thing. If you're a software engineer just wanting to spin motors and screw with code, it might be for you. But at this point, I've got over 20 hours invested and I still can't even find the firmware they've recommended (they pointed me to a bunch of dead links and apparently, 5.2 = 5.02 and 5.3 = 5.03 to them); much less install it.

Basically, this is a product suffering from brilliant engineers who have a hard time being effective communicators and with no effort towards user experience.

If anyone here has any ideas, I open to suggestions.
 
hmm ... I have no hands on equipment and tools here ... but:

Here is the latest RELEASED firmware location for the 75/300 (you don't want to go BETA now!)
https://github.com/vedderb/vesc_tool/tree/master/res/firmwares/75_300_R3 seems to have "updated status" 2 days old...
(assuming you have the R3 hardware ... it is a different neighboring directory under https://github.com/vedderb/vesc_tool/tree/master/res/firmwares otherwise)
to check your hardware version see here: https://vesc-project.com/comment/8038#comment-8038

and the changelog here https://github.com/vedderb/vesc_tool/blob/master/res/firmwares/CHANGELOG says latest is 5.03!

this firmware 5.03 should work/talk fine to the latest official VESC tool to be downloaded here: https://vesc-project.com/node/17

The confusion comes from the VESC Tool firmware info on the GUI, 5.3 gets confused with 5.30 (which we normal people could associate) ... , whadda lingo crap! It should be read and displayed : Version 5 Release 3 (==03) and not Version "5point3"

all in all this video seems quite recent (well, V 5 Rel 2) and good:

hmmm ... I really hope we can make it run (you want to set up a teamviewer or anydesk session and share your screen in a Teams telco? we could do that together!) ...

otherwise : if you really want to ditch it: make me an offer and I would take it off your hands (and try my luck with it :) )

Having said all that:
Any device allowing you to flash firmware (in so called DFU mode) is adventurous to say the least. How many nerds have bricked their smartphones while jailbreaking them e.g.. LOL
(Here the "tinker boy" can even checkout the source code and do custom binary builds ... OUCH!)
More "end user" oriented devices hide like 70% of the possible settings entirely or make you go through "wizards" which set the parameters in the background for you. And a firmware update (while still risky if interrupted) is as technical as checking the "auto update" box and providing internet access to the device or laptop.
(But have you played with and configured e.g. a OpenWRT wifi router alternative firmware e.g.? ... same tinker crap!)
Here you are at the very sharp literally cutting edge 👹 (with all degrees of freedom ... heaven or hell ... all one step away)
And you must admit: the motor autodetect wizard does kick ass on the VESC tool ... its just all the cognitive complexity while trying to understand all the parameters fully accessible which presumably is the main obstacle ...
 
Last edited:
Interestingly yesterday a new release of VESC tool came out and when I downloaded and installed it I was notified of firmware version 5.3 which I was able to flash to the VESC. I've been emailing Frank at Trampa and he's been helpful. I actually watched that video you posted a link to before - among the many I've watched on this thing. It certainly works and gives you a lot of control. However - the motor startup isn't even as good as the cheap chinese ESCs at the moment - the motor cogs until about 80% throttle and shoots to full speed. Frank told me that it might be difficult to get it better. There's HFI and VSS, both of which I don't know enough about, but VSS was similar start up and HFI just crapped out and the fault LED would light up on the VESC.

Also, there appears to be the possibility that VESC is unable to hit the same RPMs BLDC controllers can hit. This is going to take a lot of experimenting, when all we want is a controller that spins the motor without all this complexity. And if the complexity results in worse startup or blowing up the VESC or lower performance trying to figure it out, well... that's not good.

SQCTS on here got a custom flier ESC that appears to hold up. The downside is no logging and no data output, so we don't know rpm, current draw, etc. The plus side is it's cheap and it works. I don't know how important data logging is once these things become known quantities.

I'm looking forward the getting the big MGM/LMT combo. That motor is sensored, so it should have superior startuo even to the Chinese/APD ESCs. Honestly, once APD fixes their issues (which are all packaging related, seemingly), that was not a bad way to go either.

A friend said a while back, "You're gonna end up making your own ESC." I hope it doesn't come to that, but if I do, you can bet it'll be overbuilt and bulletproof. I'm not saying I'm going to start from scratch, but I might take a Chinese unit and figure out the power side of things and go nuts with mosfets. I'm not an electronics engineer, but I know enough to maybe pull that off - of course, the way I work, there will be lots of semiconductor casualties should that happen.

As for your offer to buy it - I don't really want to ditch it, I'm just tired of literally vaporizing money and it's made me gun shy.

On the plus side, it's spun up the P2 blower on 4s without any real issue... Will it hold at 15s? I don't know yet.
 
Ok - making some headway. The startup is reliant on motor amps. As in, you need a lot of them. I thought that setting to motor amp limit to something low (40 amps in my case) would give me some good safe results. But frankly, startup was jerky and didn't kick in until 80% throttle. With 170+ motor amps (don't remember exactly, but it was between 170 and 180), startup is much better and starts at about 30% throttle. I did raise my battery amps setting to 60 amps (was 40 amps before).

I also wanted the belt to stay fairly centered on the pulleys. It was walking a bit towards the motor. So with some careful playing around with shims (i.e. paper), I found that .003" (.08 mm) under one side of one standoff and .006" (.16mm) under the other standoff kept the belt off the pulley flange. This was only at 4s. I imagine a .005" shim under both would probably work best. And I can tell you, even with 1 of the mounting bolts (nuts, really) loosened, there isn't enough slop and I can't get enough deflection by muscling it to fit two pieces of paper under the spacers (.006"). I think the mount system is pretty good. It might deflect a tiny bit under full power, but the .005" shims should counter that. It may need less if I'm not using paper (which is a bit compressible).

Finally, because I know you may be wondering - 4s at 60 battery amps is enough to get just over 20,000 erpm - or just under 7,000 impeller rpm. The battery amps becomes the limiting factor.
 
i'd like one of these because it is the only one that im aware of that you can control the motor based on Erpm not just percent.. This is useful because you can maintain the same erpm / Boost irrelevant of battery voltage, assuming you have more KV than you need.. Otherwise boost will drop off with voltage drop, which will be very annoying especially if your using a small pack.
 
I don't know if the ERPM limit is something you'd want to use to limit your max speed. At high ERPMs, VESC can be erratic, particularly with inrunners with relatively high pole counts. Actually, in general, VESC should be looked at as a project under development, particularly for our purposes. Not that it doesn't work, but I have run across bugs and other issues, as well as having a whole slew of settings whose explanations are at best inadequate, at worst dangerous. Hopefully, over time it'll become more refined and the explanations for various functions will be more clear. But it is nice to have so much data to work with. That part, along with current controls, is lightyears superior to the chinese ESCs I started with.

Honestly, I think pulsewidth control based on boost is a better way to go (there's no way to limit ERPM dynamically in VESC - at least not that I know of). Ultimately, I think the work you've put in to the arduino code is a much better option for controlling these e-boost systems, and that's where boost feedback control would be best implemented.

FWIW, I apologize for any perceived slacking on my part - over the last few months I've been dealing with a severe health issue in my immediate family, which ultimately resulted in a lengthy hospital stay and may take a while to stabilize (they were discharged from the hospital today).

On the bright side, last night I shot a video with the VESC/P2 combo and got it safely running up to 57,000 erpm on 8s - 23,000 rpm impeller speed. Suprisingly (to me, anyway), the belt drive seems to holding up fine so far.
 
Since i have a dumb ESC im intending on monitoring the voltage and simply increasing the "power" as the voltage drops to achieve the same thing ( constant ish rpm irrelevant of battery voltage )
 
You must log in or register to reply here.
Top