Cheap FOCer (VESC 4.12 based design)

[NeonHuzo]

Hello I am interested in this cheap version of VESC. I am a little bit lazy to search trought 8 pages of forum so can anybody link GitHub files if there is any. Did anybody make one them self and how does it perform. Is it much better than VESC 4.12 and is it realy cheaper to make one. Sorry for my english and thanks in advance.

 

[shaman]

Did anybody make one them self and how does it perform. Is it much better than VESC 4.12 and is it realy cheaper to make one.

Hey sorry I haven't updated this thread in a while.

Beta testing took forever to get any results out of the testers. In conclusion, it works ok for under 40A applications and with higher inductance motors. So it depends on what you want to do with it.

I am currently working on the Cheap FOCer 2 that has all my lessons learned when it comes to power electronics layout and design. It will also be VESC 6 based, designed with JLCPCB's SMT assembly service in mind, and be much easier to assemble/solder. I'll create another thread for the Cheap FOCer 2 once I've made further progress.

https://github.com/shamansystems/Cheap-FOCer

 

[NeonHuzo]

Thank you. I am trying to make electric longboard and was interested into VESC but when I saw what were you trying to acomplishe I was amazed. Soo what is voltage range for this ESC. I am planning to use 42 v motor or even higher, and will try to find one with optimal power so that current doesnt go over 40A.

 

[shaman]

The voltage rating is the same as any other VESC 4.12. Abs max of 60V with max operating of 50V(12s).

If you're an esk8er you may have a better time with this forum and this thread.
https://forum.esk8.news/t/re-cheap-focer-vesc-compatible-4-12-redesign/7315/6

 

[NeonHuzo]

Thank you very much. I will think a little bit what will be used in this project. And if I manage to build one FOCer I will send the results.

 

[ZFreaky]

Okay so, I've started assembly on my 0.9 boards, gonna start number 3 soon. Not complaining just wondering if there's something obvious I'm missing with the design. Board 1 worked for the most part, however after 9 amps, it'll throw a DRV8302 fault, could be the DRV, got it from ebay to save cash and the pin1 dot is on the wrong side. Board 2 I smoked the STM and had to replace it cause I put the CAN chip on backwards, Which while I'm here, there isn't a pin 1 marker for or the DRV8302, when trying to do detection, the commutation is very, very weak, won't spin on it's own unless given a push. Again, as I write this I'm thinking it might be a bad set of DRVs I got. I'll see how the last 2 go, overall still excited with this cost optimized design, if there's a fix known for the issues I'm having so far, please share, I'll still press on with 3 and 4 anyway.

 

[shaman]

@ZFreaky congrats on taking the dive into building one of these things!

So the majority of issues I've seen like the ones you are describing come from errors in soldering. One big mistake people tend to make is not soldering the big ground pad underneath the DRV well enough. I tried to help this by adding vias underneath the DRV ground pad on the backside of the PCB as a way to wick solder onto the pad. I made this much better by adding a full through hole pad in v1.0.

There very well should be a little concave dot to mark pin 1 commonly found on a lot of ICs. If that's not there, feel free to post a good image of the chip and I'll let you know what I think.

Give you're soldering a good inspection. Feel free to post or PM some clear pictures of you're board for review.

 

[ZFreaky]

Aight, legit DRVs came in the mail, so I started messing with board 2 since it was the least usable, started with redoing the groundpad, which you were right, didn't get enough solder, need to add more paste next time. The vias also worked great to purge extra solder from the groundpad. This however didn't fix the poor commutation issue, the phases seem a bit stronger, but the motor is still unable to spin on it's own, and it takes forever to rev to speed when given a push, unloaded, very weak. Seems like the phases aren't getting the current they need. I also tried replacing the DRV with a brand new, legit one from mouser, but still no difference. I'm thinking it might be a board issue, I highly doubt I put a part in the wrong place, I'll give board 3 a shot later, can send pics of board 2 if you think it'll give a clue.

The pin 1 dot in the case of the DRV can be seen, I meant the silkscreen pin 1 dot on the board can't be seen.

 

[shaman]

@ZFreaky how does the motor behave during motor detection in the VESC tool?

 

[ZFreaky]

Very weakly, almost like only 2 phases are engaging, won't spin on its own. If you give it a push, it spins, but again, very weak, and very slow acceleration. After a bit, it stops, tries to do the constant speed but can't spin itself, get a bad detection result, with or without a push in the first part. The current in the active sampling window appears pretty noisy as a side note.

 

[ZFreaky]

Okay so just finished and tested board 3 sadly it has the same problem as board 2, except worse, can barely spin even with a push, the movement is a bit chunkier. When going through my parts, I did notice that for the 0.1uF caps I accidentally got 50V rated ones instead of 100V, none are popping but it does mean their ESR won't be as good, so I'll try swapping them with the correct ones and see if that does anything.

 

[shaman]

FYI everyone it seems we got to the bottom of @ZFreaky's issue. It was simply just the old VESC firmware and software that's not the best at motor parameter detection. The newest VESC Tool and firmware are much better at properly analyzing a variety of BLDC motors.

 

[shaman]

Cheap FOCer 2 prototyped and in progress for creating a test batch
https://endless-sphere.com/forums/viewtopic.php?f=30&t=103465&p=1513768#p1513768

I encourage anyone looking to make the Cheap FOCer 1 to wait for the Cheap FOCer 2(CFOC2) if possible. The CFOC2 has many advantages. I'll still be supporting those that have the CFOC1 on this thread.