While the current prototype shows promising results it is a mess of wires and many pins on the stm are already blocked by the discovery board. Other boards where the powerstep shield can be stacked onto often have no usb port or other issues.
I spent quite a few days researching parts, designing a pcb and preparing the cubemx project for the new layout and chip.
For the next prototype i chose the stm32f411RCT6. This processor has enough pins for 8 analog and 16 digital pins directly connected to the chip while still having usb, uart, internal spi for the powerstep and screw terminals for the encoder.
Some space is available to allow for a larger heatsink if needed. In theory it should be able to handle 80V but in current mode the driver runs best at 12-20V and this is probably what most users will have available anyways. At higher voltages the driver get really hot. In voltage mode not that much but i find the torque reaction to be much smoother and quicker in current mode.
Now i will wait for all the parts to arrive and hope that nothing blows up in the first test.
One other thing still missing is a nice nema34 dual shaft motor. The 34HS59-5004D from stepperonline looks like a perfect match with pre cut keyway and smaller back shaft and mounting holes for the encoder. Sadly it is not available at the moment so i will need to wait more or find a different one.
After a week my parts finally showed up except for the terminal blocks.
Not the most pretty soldering but it was all hand soldered and that lqfp powerstep is a beast to hand solder. Cleanup will be done once the terminal blocks are here.