GooserCS

UPDATE: This project is a failure. The bootstrap capacitors refuse to charge except for a tiny blip during deadtime, so the high side mosfets are non-functional. I tried everything I could think of to no avail, so I'm starting over with a design closer to Lepton Gooser3 - OSHWLab 

This is a rework of https://oshwlab.com/dekutree64/Gooser

https://community.simplefoc.com/t/gooser-a-4-in-1-lepton-derivative/5325/1

As before, all 4 motors are controlled by a single STM32G431 in 3PWM mode, and the encoder connectors each have two ADC pins for reading linear hall sensors on the motor.

Changes:
1. Added ACS711 current sensors (two phases per motor).

2. Heavy copper bars are now optional.

3. Board size is slightly larger, 45x52mm.

4. Buck converter is gone (no room, and it was a bit expensive with two extended parts fees). Mid-voltage (3.5-9V) must be supplied separately via the aux connector.

5. Aux connector no longer has an ADC pin, but optionally can have a second GPIO. There isn't really enough space to fit a 4-pin connector, but it can be crammed in if you solder it on the bottom side, or solder wires to the pads. I positioned the holes so you can solder a 3-pin connector on top if you only need one GPIO. Unfortunately the two GPIOs are still not terribly useful since no I2C pins were available. These two are CAN/USB, but would need a CAN transceiver, and USB is high frequency and would probably have noise problems due to the long traces to the pins.

6. Programming connector now has UART, and matches the pinout of https://oshwlab.com/dekutree64/SimpleFOC-Stepstick except for the addition of reset. But you can solder a 6-pin connector to it, and access reset as a touch pad poking out beside the connector.

7. Added a touch pad for BOOT0 so the STM32 can be recovered from all configuration mistakes. It is one of the timer pins connected to a gate driver, but activating it shouldn't be a problem if only the low-voltage section is powered.

This version is not quite as nice when it comes to joining multiple boards together. Easiest to stack them directly like I did with https://oshwlab.com/dekutree64/lepton-v2-minimal-bldc-simplefoc-driver although it is somewhat difficult reaching the motor wire solder pads between boards that way. But this one has pads on the back side too, which are more easily accessible in some cases, and I've made it easy to edit the board outline to give another millimeter or two of solder pad protrusion, which can then be sanded off the top board in the stack to make the bottom one's pads easier to reach.

You can still join two boards side-by-side on a single set of copper bars, but joining more than two boards would be difficult since the ground bars would have to cross the exposed positive rail of one or more boards.

Joining two boards back-to-back on a single set of copper bars is still possible, but be careful when connecting battery wires since the overlapping input terminals are opposite polarity in this arrangement due to one of the boards being flipped.

The connectors can of course be soldered on either side of the board depending on how you want to use it, and the ones along the edge can be straight or right-angle. Or you can solder wires to the connector pads for any that are inaccessible (e.g. the programming connector on stacked boards).

Additional notes:
1. Motor voltage is limited by the 20V abs. max rating of SE3082G gate. I recommend <17V to leave some buffer. Minimum is limited by the mosfets, 10V for full on, but 5V is usable.

2. VCC max voltage is calculated from 500mW LDO max dissipation. 49E linear halls use ~6mA each, MCU uses ~35mA, so max drop ~6V.

3. Drivers 2 and 3 have phase names in reverse order from DRV pin names. This improves routing, allows all motors to have current sense on phases A and B, and avoids confusion on solder pad order. All motors have phase A toward the middle of the board and phase C at the edge.

4. The 0402 capacitors for the current sensors are on the back side, so that requires some manual soldering. But if soldering bulk capacitors on the back, this can be done at the same time. Low temperature tin-bismuth solder paste is easiest.

5. Going by Valentine's recommendation of >1000uF for Lepton v2 BLDC SimpleFOC Driver - EasyEDA open source hardware lab this should have >4000uF of bulk capacitance. There are four pairs of holes for 5mm pitch through-hole electrolytics marked C1-C4, which also double as position pins if soldering copper bars (be careful not to overheat the capacitors). Or if you don't use copper bars, you can solder SMD electrolytics on the back. With or without copper bars, you can solder a bunch of 1206 capacitors along the rails on the back (max you can get this way is about 2000uF with 44x47uF).

Here is a quick reference list I made for myself of which pins/channel numbers are used for each timer, current sensor, and encoder hall sensor.

Timer pins, phase A,B,C
M1 pa8 pa9 pa10  tim1 ch 1 2 3
M2 pa5 pa1 pb10  tim2 ch 1 2 3
M3 pb4 pb5 pb7   tim3 ch 1 2 4
M4 pb6 pb8 pb9   tim4 ch 1 3 4

Current sense pins, phase A,B
M1 PB14 PB12 adc1 ch 5 11
M2 PF1 PA4   adc2 ch 10 17
M3 PB11 PB15 adc2 ch 14 15
M4 PA0 PF0   adc1 ch 1 10

Encoder pins
M1 PC4 PB2 adc2 ch 5 12
M2 PA2 PA3 adc1 ch 3 4
M3 PB1 PB0 adc1 ch 12 15
M4 PA6 PA7 adc2 ch 3 4