I last left off with a freshly soldered but yet to be debugged circuit. I did work through the debug the next day and after fixing a few miss soldered points and forgotten bodge wires, I and a circuit that could pulse 4.72V on each motor, alternating directions and also read the sensor for each joint.  On two of the voltage control pins I had a 0.7V offset that I could not make go away. It was low enough that I could still reverse the direction of the motor but it did not go to zero when I toggled the pins. It was even their when I took the leg off the board. (Having the headers makes it so much easier to debug.)  There was also an offset on one of the analog inputs.  Since I have an new board coming next week, I will monitor it again there. The point is I have robust proof of concept for the double demux. 

However, show me the money right! I have posted to Hackaday for a couple of years now but never uploaded a video. On the other hand, my son has posted numerous videos that have gotten over a million views, so I think it is time to swallow my pride and asked the next generation how it is done. The goal for this weekend is to show a video of the robot moving, I mean that's what they do right!

 To do that, I need to do the following:

1) Solder the motors to the flex

2) Assemble the second leg.

3) Design and 3D print a spacer between the board and the motor since the headers make it so it cannot be flush.

4) Note everything about the flex I do not like to prepare for a flex spin.

5) Redesign foot based on assembly thoughts in 4.

6) Redesign flex with the following improvements

a) Attach charger to the flex and considered in 5.

b) Change switch mounting locations per 5.

c) Change to two layer flex and see the cost increase.

d) Make the adhesive areas selective? The point is, I would like to maintain some through holes. Especially for the motor.

e)  Move the mounting holes for the hip 1mm laterally to the side with the switch connection. (To the left when the flex is pointing down) This will allow the PCB and the header to be centered on their pads. Presently I need to be creative on my soldering.

On a separate note, I want this project to be a collaborative endeavor. My first collaborator is Doug and he sent me the follow update. He is leading the python charge.  I will add his notes in the next log. He has gotten the I2S working and is working on the wake word problem with Edge AI.

If you are interested in serious contributing to the project, you can request to become a project member. I can then get you started with the hardware you need to accomplish what you want to move forward. New boards come next Wednesday.  The Cad files are available and are not hard to print. I can supply the board files as well. I am not 100% happy with them yet so I have not posted them but I may after I QA the upcoming board.

Two legs are better than one. Now to get it moving.

Better light

So flexible

Aiya