Potentiometer Servo Feedback (for Robot)

I’m working on upgrading/overhauling a lab robot arm, and trying to figure out a good combination of motor drivers and control boards. I’ve used MESA boards quite a bit, so preference is one of those BUT I’m flexible if there’s a cleaner/simpler way.

The robot arm has 6x motors; 3x appear are 15.5v and 3x are 5.5v. They’re driven by H bridges and the main transistors are rated at 4A continuous and 7A overload.

Each joint is fitted with a potentiometer (I believe a 10k pot), which measures the full range of the joint - the motors have gearboxes, gearing then down to an acceptable ratio.

I don’t want to change the potentiometers, as it gives me an absolute position without having to home etc. However, if there are any well priced absolute encoders, then that might solve a big chunk of this problem.

I’ve looked at the MESA 7i54, which is for this application, but the bridges are Max 2A continuous. Now, that may be sufficient for the 3x smaller motors, I need to take more of the arm apart to get an idea of motor loads. That board has encoder inputs so could be a solution, it requires additional motor drivers for the bigger motors.

Are there any boards - single channel, multi channel etc, that would be suitable for the motor load? Cost is a factor, so whilst MESA make very nice bigger motor drivers, they’re way out of budget. Worst case I may consider the cheapo Chinese drivers if it gets the arm running well (can be replaced later on).

The potentiometers are causing me the grief - I can’t seem to find a single board that has axis control and analogue inputs - they’re all encoder inputs.

Any suggestions? Very worst case, I’m even considering directly interfacing the old motor driver board for the time being via the TTL level inputs the controller board uses. I do have very old circuit diagrams if anybody wants to have a look!

Thanks

hi
you can use the 7i76e and for the encoders the extendend 7i43

so you got a full robot 6 servo setup
the garbige 10k pots shoudt be replaced by low cost 2000ppr encoders about 15USD eatch and the size of 38mm shoudt fit your need

if you got a 3d printer you can easy print a scale hole disk and use othere electronics as a counter direct on the 7i76e

That being the case, should I be looking at incremental or absolute encoders? I found a post from 2011 that implied LinuxCNC has some issues with homing absolute axis? I think PCW didn't seem to think it would be a big issue.

If I go down the incremental encoder route, any suggestions as to homing methods?

for a robot it is usaly a SHUTDOWN event to set the robot to a good start condition
and if on HOME use it is not needed for Home switches then as it is there on the start

if you need it you can use Reedrelais magnetic or simply Cherry5 switches to get it to a max
on DOF6 it is always the max that is homed on arms
the Hand J4 and Base J0 is 360 limited and got a simple 1mm Inductiv that sence a hole

The arm in question doesn't have active braking, so when power is lost, the joints coast. Other than making a stand or something the arm lowers itself on to, there's no reference point maintained. Edit: the joints use dynamic braking via the H bridge motor controller.

As far as I know, there's no provision for limit switches, as the limits are potentiometer based.

The smaller motors (pitch/yaw/roll) are heavy duty RC style servo motors, without any internal control - the pots are wired back to the main controller. I'm not sure how easy it will be to do anything with those potentiometers. Unless there are RC style servos available that I can replace that are SPI etc. Maybe those can be solved via a smart RC servo and the other three bigger motors will need a limit switch or an absolute encoder on the joint.