Continuous rotation of the gear motor would be doable using the OpenServo controller in I2C mode and the US Digital MA3 encoder instead of the pot for feedback. Depending on how you are using that setup, you would have to define the initial position (within a turn - such as a slow move to find a limit switch), but there after, the controller could keep track of the number of rotations, position within a rotation and direction. This ability is not in the OpenServo firmware yet, but it is one of the features being talked about and is certainly doable. The US Digital encoder would need an adaptor or different bracket, which would have a bearing and short shaft to couple the encoder to the motor shaft.
Also, the new servo board design will measure back EMF, which will allow the controller to keep track of motor speed and with some number crunching, relative position, without an encoder. This would not be as accurate as using an encoder, but would be good enough for many applications. But this feature is still in development, so no guarantees.
Please take a look and give me some feedback - either here or there. I pretty much have the electronic design in hand, but trying to sort the mechanical design has delayed the project - so let me know what you think (what you like, what you hate, any ideas) and we’ll get this project done.
Sorry, those boards went to blue smoke heaven at about 5A. But stay tuned, the new boards will do an average 10A and a peak 25A without blue smoke.8) It will be a little while, so there is still time to stock up on 14 gage battery supply wire.
This is using what Panasonic calls a position sensor, which is just a pot without a shaft and bearing. The cost of the sensor is roughly $1.31 @ 100 pcs. (DigiKey). The bearing shown is from McMasters and is $6.10 (quantity 1), but I expect that you can do much better from your bearing supplier. The PCB would be single sided and is a rounded 1"x1" board. This same set-up will work with the US Digital MAE3 (Size 7) - that is the mounting holes, shaft, bearing and connector would be the same.
If you find this interesting, let me know and I will send you the details.
The shaft would have to be made. The shaft sensor end is 4mm (the D is 3mm), the shoulder for the bearing ID is 3/16" (bearing OD is 3/8") and the end going to the pivot bracket is 1/4". I wanted to give an example using a real bearing, that would drop in as a replacement for a standard pot - the only change to the motor bracket is the two mounting holes. The shaft design could be different, depending on the bearing ID and possibly other applications for rotational feedback (I would guess this could be used elsewhere in robot designs).
I have no idea what the shaft would cost, but my sense is that it could be fairly cheap. The good news is that if the pot gets worn, only the sensor board needs to be replace (or just the device). (BTW the pot is rated for 1 million turns)
This is a really great idea. Cliff, I’m sending you a long overdue email. Because my first batch of motors arrived with the 0.1" spaced connectors missing (not installed), I’m open to the smaller connectors you are wanting to use. This is a go ahead for your designs. Thanks for your patience in dealing with me.
