• Hello ES! We could use some help to get us past the finish line on building the new knowledgebase for the forum.
    Can you donate? Please see our fundraising page. Thank you!

Having a custom VESC controller built.

Joaquin Suavve

Established
Joined
Feb 27, 2025
Messages
237
Location
Guadalupe, California
It needs to be an odd shape (not square) and 60 volt/ 80 amp.

What are military drones using? I'd sure like to distinguish/ understand the difference between RC and what the DOD is using.

BTW: I have no problem paying top dollar for someone with the right skill's to help me!
 
It needs to be an odd shape (not square) and 60 volt/ 80 amp.

What are military drones using? I'd sure like to distinguish/ understand the difference between RC and what the DOD is using.

BTW: I have no problem paying top dollar for someone with the right skill's to help me!
I would ask the folk on VESC discord. It looks like they do exactly this work and do this for money, at least they advertise such services. There was a guy who bragged he can design VESC controller and firmware for it from scratch withing a weekend so it looks like this work is not that expensive, just a few days of labor. I tried to contact some of them regarding such work and had no response though, but maybe you will have more luck dealing with them.

Military drones are using is what is the most cost effective in a given situation.
 
Last edited:
Yeah there are a good number of people who can do this although you're better off asking someone to modify a tested design or honestly why does the shape have to be so specific?

The thing about VESC or any traction designed speed controller is they are not the same use in many ways as a rotor speed controller. A rotor like in a drone, boat, plane, etc has a well understood inertia and fairly predictable drag response and is spinning at a fairly high RPM most of the time. A traction motor like in a bike or car has to start from a dead stop and ideally apply maximum phase amps and not skip a beat, similarly you hit a bump and the load spikes up super fast, has to keep control.

If I was building a system from scratch I would use a VESC but I would use it with a magnetic encoder, assuming you don't do a terrible job of protection the encoder wiring from EMI it will be very reliable, predictable, require very little tuning. Just better in almost every way than using hall sensors and sensorless. Only downside is very high RPM but you probably won't be using a motor that will spin that fast in any vehicle application. These reasons are also why every electric production car's motor that I've seen uses a resolver or encoder of some sort. VESC supports many different encoders/resolvers/etc it's just the tiny chip magnetic ones work quite well and are very easy to install if you have access to one end of the motor shaft.

Also like 60v/80amp you can make a tiny VESC so maybe the shape doesn't matter, like a Team Triforce A100 is absolutely tiny and I think at the moment Team Triforce makes some the highest quality VESCs around, they are not cheap (but maybe not so bad if you were ordering a good quantity of them) but with most VESCs you get what you pay for.
 
I run a VESC on a middrive and 2 hub motors before. No problem using them for that purpose. No real hassle tuning them either. The only thing that's difficult to get working is PAS and torque sensing, because they didn't think about bikes much when designing it :)

But yeah i second that the VESC forum would be the best place to find what you're looking for, because there's only a handful of VESC devs on here compared to there.
 
Spintend ubox lite is pretty small, a bit bigger than the one linked above but much cheaper.
 
Not with PAS. PAS is integrated in the VESC code by default. But with torquesensor and with a Bafang display. Merged from the Luna Cycle fork to the Flipsky hardware. Of course it can be merged to any usual VESC variant. I asked in the discord developer group, but there was no interest to make this features available for all hardware. 🤷‍♂️
There is a legal reason beind that. Benjamin Vedder did explore such possibility to include torque sensors support in firmware and asked on the VESC forum if there are other than bicycle applications for torque sensors but ended up not finding any. They can not include this support in VESC firmware because of it will render their controllers have a prohibited feature (overpowered motor controller with torque sensor).

They can legally include PAS though because of PAS is just pulses signal to command the motor to rotate which can be used beyond bicycle applications (for instance it can be used in robots to actuate a mechanism with button which looks the same as PAS on the code level).
 
Last edited:
They can not include this support in VESC firmwar
It is already implemented in the Luna Cycle hardware config and the PAS app officially. So there can't be any legal issue. But with the Luna implementation the torque is read from the CAN bus as the M600 has a CAN sensor, I've implemented it for reading the torque signal from ADC2.

 
Last edited:
Yes, as long as you are not part of the VESC team you are free to code whatever you want like Luna did. But Benjamin Vedder did not move forward with coding torque sensors himself because he did not find applications for torque sensors other than bicycles and overpowered bicycles controllers are prohibited in his country and he did not want possible litigations. There was a discussion on the VESC forum long time ago about it. He ended up implementing only PAS. That also explains why the VESC team did not release the VESC firmware for the Innotrace controller and kept in secret that Benjamin Vedder himself worked on it until I accidentally messed up their plans and even now after they decided to release it they do not share the installation instructions.
 
Last edited:
?! That's completely rubbish, I think. Marcos from Luna Cycle is a honoured member of the inner circle of VESC developers. His implementation is part of the official VESC code, not the bungling of a hobby programmer like me.
 
Well, that's how it looks like according to the old discussions on the VESC forum regarding implementing PAS and torque sensors and the responses given my Benjamin Vedder on that forum publicly regarding it.


PAS can be technically implemented without legal issues, torque sensor not so much.
 
Last edited:
In the EU, so in Sweden, there is no rule regarding the use of a torquesensor in an E-Bike. There are just two things to consider: max. 250W rated continuous output and no motor assistance above 25 KPH. The peak power is not limited. Definition of the rated power is discussed often. Bosch has the simplest and most convincing explanation.
Rated continuous output

Rated continuous output describes the mechanical power that your drive unit can provide continuously without the components overheating or being damaged. Under EU law, the mechanical rated continuous output is limited to 250 W – this applies to all eBikes and eBike motors.
So the rated continuous output has nothing to do with the controller, it is only defined by the properties of the motor. With 251W continuous power it has to break down per definition. I wonder what tricks the OEMs have to sell their systems as legal 250W motors... ;)
 
Last edited:
In order to be legal troubles free VESC positions their motor controllers as just motor controller without specifically for pedal assist bikes and it is up to the the people how yo use it. There is no rule regarding the use of a torque sensor in an E-Bike but implementing torque sensor on over 250W motor controller would render this controller over 250W controller specifically for bicycle because of there are no other applications for torque sensor controls motor and the responsibility is ow not on the person but on the VESC and there are people or companies or politicians who could potentially try to use it against VESC or try to take an advantage of it harming VESC.
 
Last edited:
Just stop that rubbish, as said the torquesensor has no legal relevance for "EPAC"s (Electrically Power Assisted Cycle) and it is implemented in the official VESC code already.
The controller can't change the "rated continuous output" of the motor. You can power a "250W" motor with 10kW, if you want, that's not illegal in Europe. But the motor has to fail with 251W continuous output. :whistle:
 
Last edited:
Yes, this is a defense point in case of litigations. Another defense point is it looks like the torque sensor support code was not implemented purposefully by the people who makes VESC branded controllers and was implemented by people from other country for other company (or VESC team made it looks like this) and this code was just merged with the main VESC fork according to the open source requirements. This trick gives the VESC team some defenses in case of possible legal attacks and deters the ones who would like to bring a case against VESC.
 
Last edited:
Thank you all for the input and information!

But come on you guys! Lets keep this thread on track! I doubt that whatever I'm up to is going to need a torque controller other than my left wrist!

Of course I'd prefer an "off the shelf" controller! I maybe will be finding one that is close to the size I need, then doing some doctoring (like removing the case ) that "MAY" work... But I am exploring the possibilities if it won't!

Other than frame geometry, wheel size (probably 29"), and rear hub options... I'm starting with a "blank sheet" (term from back in the drafting days) and will be trying to put a lot of motor, motor control, gear reduction, AND cooling in a very small place!
 
Last edited:
Of course I'd prefer an "off the shelf" controller! I maybe will be finding one that is close to the size I need, then doing some doctoring (like removing the case ) that "MAY" work... But I am exploring the possibilities if it won't!

Other than frame geometry, wheel size (probably 29"), and rear hub options... I'm starting with a "blank sheet" (term from back in the drafting days) and will be trying to put a lot of motor, motor control, gear reduction, AND cooling in a very small place!
We just finished the new VESC based INNOTRACE Shrimp motor controller can easily handle 60V 80A you need. That controller got the highest power density among all the existing motor controllers in the world. It also has specific shape and design approach allowing to integrate it in tight spaces in the motors. I am sure it will fit wherever you want to place your controller. You basically can not find any better for your application.

 
Yes, this is a defense point in case of litigations. Another defense point is it looks like the torque sensor support code was not implemented purposefully by the people who makes VESC branded controllers and was implemented by people from other country for other company (or VESC team made it looks like this) and this code was just merged with the main VESC fork according to the open source requirements. This trick gives the VESC team some defenses in case of possible legal attacks and deters the ones who would like to bring a case against VESC.

Maybe we should collectively come up with an 'ebike controller' kit that includes torque sensing, pas, and 3 speed switches.
I'm fairly happy with my throttle only VESC, but it's lacking in those creature comforts.

Joaquin; since you cannot buy a 'shrimp' at the moment;

It's probably safe to think of the spintend lite as a 33A ( battery ) continuous controller, maybe more if you can add some thermal mass / a heatsink.
1783057410566.png
Can't immediately think of a more suitable controller you can buy today. Maybe if you find it, it costs a lot more. Spintend quality is pretty good for the dollar though.
 
The solution already exists, at least for this controller

 
I bought a flipsky and it turned me off from the brand:
- really poor quality construction is apparent, as i'm handling the controller, the wires are falling out of the connectors as if someone didn't bother to crimp them fully.
- shipped me a DOA unit.
- some included accessories are flimsy to the point of being dangerous.

It did cost a bit less than the Spinted but.. it was more of a 'cheap' product instead of a 'frugal' one
 
The Luna Ludicrous V2 firmware can probably be adopted to any VESC controller so someone familiar with the VESC firmware probably can do this so the controller will work with the cadence/torque sensors and with the Bafang CAN displays and you can select the assist level with the Bafang display. We know how the cadence sensor, the torque sensors and the display should be connected, we just need the firmware.
 
Last edited:
someone familiar with the VESC firmware probably can do this so the controller will work with the cadence/torque sensors and with the Bafang CAN displays and you can select the assist level with the Bafang display
It would be possible to implement the torquesensor and display support into the default VESC code and make it selectable in the VESC tool, so you could use it with any VESC hardware. But there is no interest from the VESC community, I asked on the discord channel some time ago already. I offered help for the embedded side, on the VESC tool part, I can't help.
 
Back
Top