New member here, thought I should say hi to the community. My name is Aleksander. I’m an electrical engineer, 25 years old, and my hobby is to design electronics in my spare time. I was a part of a formula student team while I studied so I have some experience with motors and batteries (400V battery and 80KW PMSM motor).
I’m planning to design a ESC (using FOC algorithm or similar) for educational purposes. When I’m done with that I’ll probably begin with a BMS design.
My questions are:
What do you guys think would benefit the esk8 community?
Anything that is lacking in the open source section?
Any problems with the VESC?
Are there any other hardware designers / software developers on the forum?
I’m hoping to get some discussion in this thread regarding the open source hardware available for esk8. It doesn’t have to limited to ESC, but I’m going to work with ESC design first (again for educational reasons) .
an alternative to the VESC probably wont hurt.
i’m actualy more up for a small scale and intelligent BMS.
something combined would probably be the crown jewel in this category
Welcome to the forum. Usually when people start designing challenging electronics like a foc-esc or a smart bms (see raphaelchang) things get complicated on the software side, leading to uncompleted projects.
Why don’t you start with something simpler like a high amps switch (150amps) with directfets or a non-IC bms?
Btw maybe a lipo charger like the imax b6 just on a pcb and maybe with changeble settings to adapt it to it’s battery pack.
Because, maybe I am wrong, that such a charger charges the batteries a little bit better than a bms but it is pretty anoying to plug the batteries out everytime u charge. With this pcb you can insert it in the enclosure and have a charging plug for the needed voltage
So than you have the best of both worlds, the best charging, and the user friendliness of a bms system
The largest barrier and creates the most problems is the VESC (for me included). One reason is not plug and play and the other is poor build quality. I understand that some of the folks that are very skilled in electronics like all the adjustability, but the majority of us just want to quickly set up our equipment and go ride. The FOC BOX is the closest product I have seen so far but still not plug and play enough. Should be as simple as setting up an ESC for an RC car with same reliability as say “Castle Creations” or Novak" products. If it had a BMS as part of the package would be that much better. Again, plug and play. Hook it up and pick your battery size (6-12S) and done. Another thing that would be a huge help would be a system where you can combine battery cells to make a larger battery without having to weld the cells together. I have seen some plastic battery cell holders that you just snap the cells into place and combine into a larger battery. If there was a frame that would accept cells to make between 6-12S with no welding required (just soldering connectors) anyone could do it. Maybe it could be 3D printed with some nickel inserts or something?
they only reasons you dont see these things beeing used in sekates, is because those methods waste alot of space and also dont handle vibrations very well. the last thing you want on your board is failing breakes because the battery decided to have a small outage. if you need an easy method, go with lipos
PS: the Focbox actualy is a VESC, it just has different components and a different formfactor. both have the same “experience”!
I mean look, everyone here most of the chances are running FOC mode (when they can) as it’s silent and doesn’t catch police attention. It provides more torque anyway, so not supporting BLDC mode that doesn’t matter so much. If it runs on both BLDC and FOC is the best. Yet also need to support sensored and uncensored. It’s even more marvelous if the BLDC tool is much simpler. People want more voltage, more current, higher max erpm. But again if you can make simpler version of VESC with 80 euro price range, thats a hard contender for VESC. More or less soooo many people have their VESC / FOCBOX broken in away. If the fragile components is plug and play replaceable using something like this https://cdn-shop.adafruit.com/1200x900/1281-00.jpg That would make replacing broken things soo much easier.
just make something that does foc and doesn’t break! the rest is unimportant. vescs and focboxes break nonstop its ridiculous and at this point it seems there’s almost nowhere to get one of these vesc versions other than the crazy expensive 6, the lackluster focbox, or some Chinese crap that might not even do foc. no more ollin and it seems no more axle these days or at least theyre so backlogged they might as well be gone. please make a simple cheap foc machine that doesn’t break! you will be rich and we will all have it
VESC 6.4 pcb designs open-sourced. A pcb that’s skinny (35mm) and long (120mm) versus square. (65mm * 65mm) We have options for square already. Easier to fit in tight spaces. Also preferably a way to have two paired on the same PCB, possible able to break them apart into 2 pcbs (the canbus lines etc would just sever at the board break) if needed for single motor ESCs. And with protected DRV and CAN chips so they aren’t blowing all the time.
It’d be great to have “trapezoidal FOC” mode and “sinusoidal FOC” mode — instead of BLDC and FOC modes
Someone should offer kits to make your own battery packs, assuming you have minimum set of tools.
Maybe offer pre spotwelded 3p or 4p packs with balance wires that can be connected modularly into a 6s or 10s configuration with a simple charge only BMS?
I would love to try and build my own pack but I don’t have the confidence (or a spot welder) to do it. Similarly, while the process may be theoretically straightforward, I tend to be reluctant to try this sort of thing for fear of destroying or shorting cells, or it all catching on fire.
If I could buy all the compatible parts and assemble them following a set of instructions that were easy to follow (read: idiot proof) and procedurally correct (EE certified), I would be very interested. It would hopefully reduce the perceived complexity of assembly, especially where a BMS is desired for ease of charging.
One could get into the argument about the ease/ difficulty of shipping the 3 or 4p modules internationally, but depending on cost, you could potentially have a ready and waiting international market, especially when your local battery builder wants to charge you arm and a leg for a pack.
The current vesc has components rated to just over 13s voltage as I understand it. Many use it at 12s (some 13s). 10s has shown to be more reliable though so what we really need it would seem is a esc with components rated to at least 15s or 16s voltage so it is reliable at 12s. Essentially we are using our gear at its limit with little to no headroom so it brakes down often…
But good electronics design is not so easy and since your software is relying on your hardware to control your motor, if your hardware is not well done then your software’s performance will also be limited by proxy.
Hardware costs money, software is free. This was the saying in good days of Nokia.
I don’t want to sound rude, but that comment just seems so layman to me.
I am electric/software engineer I can design ESC board in a day software will take 1-3weeks. So here is an example for you why electronics are easier.
Plus the motor control is I think the easiest stuff to do because it only depends on 3 x half bridges, current sensors, voltage sensors and that’s all… For e.g. I am working at the moment with a main control unit which would do like data logging, aggregation of data from multiple vesc and etc. So that board is way more complicated than ESC
