I roughly know about origin of the problem with the motor drives in the stock Solo, i.e., have read that it is a shift in gnd level at high currents such that the cube’s 3.3V PWM voltage may become insufficient for the ESC input to reliably detect the PWM signal, and it’s “solutions” such as to install a green cube with 5V PWM signals or the infamous level-shifter-mod in the stock cube also resulting in a 5V PWM (or to software filter as in the stock Solo’s firmware).
I’d like to try now a different solution, which essentially consists of applying a clean 3.3V PWM signal directly at the motor pod connector, but am not absolutely sure if it will safely solve the issue since it’s not absolutely clear to me where exactly the technical issue for the gnd level shift “sits”.
Specifically what I’d like to understand is if the large gnd level shift is in fact produced on the motor pcb itself, so that even if one would provide a 3.3V PWM signal directly at the motor pcb’s input connector (using also a good gnd connection) it would not help, or if the gnd level shift is so to say produced along the current’s path from somewhere on the main board to the motor pcb, in which case a close 3.3V PWM signal would resolve the issue, and which I kind of find the more likely scenario.
THX in advance for any helpful insight,
PS: what I consider to do is to install my UAVCAN-ESC bridge (or UC4H ESC-Actuator node in my speech) in-between the ESC PWM input and the cube’s CAN port, very close to the motor pcb, such that the motor pcb would get a full 3.3V swing at its input connector.
I believe the ground lift occurs on the motor/ESC negative battery wire due to high motor currents during transients. This is why the fix involves softening the motor responses.
It’s amazing that 3DR made this mistake, since Ardupilot people knew of the issue for several years before before.
not sure I read that correctly, but I think you imply that the gnd shift is exterior to the motor pcb, right?
let’s hope someone else has additional insight, I would feel more comfortable not having to rely on my or your believe
I know this mod well, and sadly it doesn’t answer my question if the mod I intend to do works
many thx anyways
as-is , you cannot be sure to detect incoming 3v3(relative to the cube) on an motor pod. except if you added a buffer or a level shifter that would handle anything above 1v as high, and output 3v3(relative to the pod)
I played around with that, but did not design anything that was simple and foolproof enough to easily add to all pods ,as I did not wish to add complexity, or make it more vulnerable to mechanical stress, and users fiddling with pods.Same story for my idea to up the voltage right on the Solo motherboard, as a plug-in-between the motor connectors and main PCB, one could easily add some small PCBs one for each motor, one would be a tight fit, and all woudl need external 5v supply, as there is no clean 5v supply in the 6-pin connector.
thx for the further details
unfortunately, they still don’t get to the bottom of my question
I think I understand well that I “cannot be sure to detect incoming 3v3 (relative to the cube) on an motor pod”.
My question would be: “Can I be sure to detect incoming 3v3 (relative to the motor pod’s connector) on a motor pod?”
I heavily researched the web (I did before, but now again) in order to find electric schemes or layout files for the Solo motor pods, but did not find anything (which of course might be due to me not searching sufficiently well).
Thus, question pl, anyone knows where to find such info on the motor pod pcb’s?
ok, I think I have figured it out, and am quite sure that “my” concept will work safely.
THX for your comments, much appreciated.
(wasn’t there a way to mark a thread as “solved”? can’t figure it out anymore)