Hi,
I understand that when moving from 4.12 to 6, the current sensing method was changed from 2 shunt 'bottom side' to 3 shunt 'in phase' .
Current sensors for in phase applications - specially for high voltage application are pretty expensive, and the only credible option for me seems to be the ACS758 or variants that other members seem to be using.
Questions:
- Is there any benefit of 3 shunt 'in phase' sensing over 3 shunt 'bottom side' current sensing for FOC ?
- Is it still advisable to design a high voltage design (150V) - 10A using the 4.12 design as a base - I was planning on basing it on ST's STEVAL-IPM15B board.
- Is it possible to use 3 shunt 'bottom side' current sensing with VESC ?
Thanks!
For high voltage the 3 battery sided shunts is the easiest way to go.
I assume you are referring to the ones after the low side fets and before Common (-) terminal ?
If so, how to use that with VESC ?
@ripperpc refer to my open source design for an example of low-side shunt current sensing.
https://github.com/shamansystems/Cheap-FOCer-2
Thank you so much for sharing that !
I am planning on using STEVAL IPM15B board that I have lying with me to make a high voltage version.
Do you have a guide as to the firmware changes you made to make this work ?
Thanks again!
I don't have a guide. The VESC firmware is designed to allow easy adaptions for differing hardware. There are certain #define statements that you enable or disable based on certain qualities of your hardware. There also might be some values you have to adjust such as current amp gain or voltage divider resistor values.
The big benefit of using inline sensing versus low side or high side sensing is that you can directly measure the phase currents at any time. The trade off is that the electronics to do this are typically more complicated at higher voltages (>80V) because existing non-isolated bidirectional current sense amps have an upper common mode voltage limit around 80V. Your IPM15B already has low side shunts and differential amplifiers according to the eval kit schematic.
I have been investigating inline sensing and came across the Allegro ACS37612 as a potential option that hasn't been used from what I have seen. The ACS37612 measures the magnetic field from the current flowing through a trace or bus bar and is advertised for higher current applications, but I don't see any reason why it couldn't be used if designed in properly. Accuracy, linearity, and bandwidth may not be quite as good as the ACS770 but it is smaller, cheaper, and lossless.