Phase shunts above 16S is difficult. 18S is about the max. Reason: The shunt amplifiers are not rated higher. INA240 can handle 18s but that is borderline and personally I would not operate it beyond 16S. That part is also unobtanium since ages.
The only other alternative is the AD8418 and that part is also not easy to purchase in volumes. Each ESC needs 3 of these.
So low side shunt designs are easier and cheaper to produce. There is more part availability and the parts are cheaper.
For sHFI the voltage tuning is important. It also depends on the motor Lq-Ld difference value.
I will try to make instructions on tuning soon.
FW issues for third party HW should be addressed with the company offering the HW. Vedder is not coding that FW. He recieves pull requests and integrates the FWs.
Let’s stop using the phrase battery shunts. No VESCs use battery shunts. There are either phase shunts or low-side shunts. You can use either two or three phase or low side shunts.
The low side shunts are between the low side mosfets and the ground of the systems phase shunts are on the output of the phase between the high and low side mosfets. So far there are no dual hardwares that use phase shunts that I am aware of. Only very recently did TI release a bidirectional phase shunt amplifier (ina241) capable of 100V common mode operation that has PWM rejection. Before that any hardware that had phase shunts couldn’t run beyond 80V so between that and cost is why all Stormcore currently feature low-side shunts.
It is something I’d like to look at as a potential improvement in the future. Mostly there are not so many major general advantages to phase shunts but it can be more robust in some cases because it allows you to filter the phase current measurements. The addition of the ability to measure current in both V0 and V7 also now enables silent HFI to be possible at half the needed switching frequency which is a new advantage that is very interesting.
The latest Stormcore do feature phase voltage filters which help with low speed sensorless operation mostly and improve accuracy of motor detection routines.
I don’t know whether it’s correct or not but the interesting thing is just from looking in the hardware config files in the VESC source code, the MakerX DV6 is a dual that apparently (according to the files) has phase shunts and phase filters. Although the file is called “DV6 Pro” and not just DV6 so maybe that’s a single I’m not aware of?
Thanks for the clarifications, I’ve updated a few of my posts to just call them phase or low side shunts.
He’s one of the guys who designed the Unity and Stormcore controllers. He’s specifically talking about single mcu (STM32), dual motor controllers like those two. That was a significant design change from Vedder’s schematics, and it had to use an STM32 with more pins along with a lot of firmware changes to be able to control both.
All other dual controllers are really just two single controllers, each with their own STM32, put onto the same PCB and they communicate over CANBUS.
It does have phase shunts, but the difference between low side and phase shunts are about the placement of shunts rather than the look. Usually phase shunts are going to be visually very close to the phase wires.
They are simply resistors and on VESCs usually look like this
The comment just made it sound like he was saying they just don’t exist on any VESC My bad if I was supposed to know he only meant on Unity and Stormcore
I know that and they are 100% between the phase wires and everything else. They even often come with some of the solder from the phase wires touching one leg of the resistors. I had to look way too long at different photos of their FESC to ensure that it won’t be an issue or cause a short. Some of their own product photos will show them touching and others don’t have the solder touching them at all
The hardware layout resources directly from VESC project also show the resistors as connected directly to the phase wires, which further makes me think it should be fine. I’ll probably try and get them to not touch if I can find a way to remove the conformal coating on the entire area at once. It does melt of with a soldering iron but only where it directly touched, and I think it isn’t helping when I try and remove all the unwanted solder on the resistors.
There is the possible issue of not knowing which leg of the resistor connects where on the PCB but I think it is safe to guess that the side closest to the phase wires is connected internally (via the pcb traces) to the phase wires.
I had another response from Lacroix and they confirmed again, in conjunction with one of their engineers, that the 60D+ has high-side shunts and phase filters so go figure. They are going to look at the hardware config files…
My bad if I was supposed to know he only meant on Unity and Stormcore
I had to look way too long at different photos of their FESC to ensure that it won’t be an issue or cause a short. Some of their own product photos will show them touching and others don’t have the solder touching them at all 
