Asking for Gripper Feedback Advice

ajward

I'm in the process of replacing Wally's defective ankle and the new shiny thing arrived:

Just wondering is anyone has devised any kind of feedback for this kind of gripper? So you don't crush delicate things or stall the servo.

Appreciate any guidance!

Amanda

Beau Schwabe

In prosthetics current monitoring of the motor is used combined with strain gauges. But that might be overkill for your application. Current monitoring of the motor might be doable though, and instead of strain gauges, a carefully placed FSR (Force Sensitive Resistor) would work. I stress carefully placed FSR, because they must remain perfectly flat. A crease will kill the sensor, while a bend will attenuate the full range of the sensor.

Martin_H

I haven't done any sort of feedback. What I do is figure out which servo pulses produce desired grab widths for expected object sizes. I then plan my motion scripts for objects of that size.

GordonMcComb

For cheap you might try the old conductive foam trick. You can find tutorials for making foam sensors various places. Keep in mind the result values may vary, especially for foam that gets old or stays under compression. But what you're really looking for are ratios between fully open and some other sets of closed values you've determined empirically.

Since you probably want your gripper to open and close fairly rapidly, that means a low-ish gear ratio, which means fast acting but low torque. So just from a practical angle the gripper likely won't crush most non-living things. Select a servo with the appropriate torque. That will also mitigate most any gear stripping.

xanadu

You could do something mechanical like use light springs and or rubber bands in place of the arms that connect to the servo horn. Other than that I'd definitely consider monitoring current.

erco

Another trick is to drop the voltage to the servo and just let it stall. Lower voltage means lower stall current. Many of my recent projects have used a single Li-Ion cell (3.7-4.1V) to drive the servos and they're still quite happy. I used one those Ebay boosters to get 5V for the Stamp & circuitry.

You could just add a few series diodes to drop the voltage going to the gripper servo. I prefer diodes to a big power resistor, but one of those could be used too.

Duane Degn

GordonMcComb wrote: »

For cheap you might try the old conductive foam trick. You can find tutorials for making foam sensors various places. Keep in mind the result values may vary, especially for foam that gets old or stays under compression. But what you're really looking for are ratios between fully open and some other sets of closed values you've determined empirically.

Here's Phil's tutorial.

erco

@DD: I'm lovin' your Heat Miser avatar. Keep it!

ajward

Many thanks to everyone for their responses. Interesting ideas to ponder. :-)

Amanda