Potentiometer Recommendations for Arm Position?

My team wants to use a potentiometer in order to position and hold an arm anywhere between 0 and 140 degrees. From what I can gather around CD, a 1 turn linear potentiometer mounted to a bracket and connected to the end of a shaft via vinyl tubing and zip ties would do the trick.

Does anyone have any recommendations for this type of potentiometer?


I like using the US Digital MA3. It’s not exactly a pot, but an absolute encoder, so it gives a bit cleaner of a signal. It also doesn’t have the hard stops, it just wraps around 0-5V continuously, so you can’t break it by accident.

If you want to stay with an analog device, check out McMaster-Carr. To find them, use the search term “variable output switches”. Don’t know why they call them that.

We had good success with one of these our rookie year where we used it on an arm. Don’t remember which one we used. They have 1, 5, and 10 turn flavors.

Potentiometers can be a useful device to do what you’re trying to do. If I were looking for one, I’d buy this with the Digi-Key voucher in the kit of parts:


It may look expensive. This is no flimsy Radio Shack -esque pot. It’s an industrial quality device that will take a beating. The link is for a 1k resistance value. In noisy environments (like PWM motor control) I like some current in my analog circuits. This can be wired directly into a roborio analog input. Put 5V on one of the end terminals, ground on the other end terminal, and wire the center terminal to the signal pin.

Potentiometer survival tip: Only the shaft bushing or the shaft should be rigidly mounted to anything, never both.

QFT. Never use a sensor as a fastener.

Does it reset when powered on/off or will it keep its position?

It does not reset when powered off.

We went with the MA3-A10-250-N Absolute Encoder from AndyMark. Although, we’re not quite sure how to program it. We use LabView and I’m not sure if it should be programmed as an encoder or as a potentiometer for our application. CD searches don’t turn up a whole lot beyond saying to use the get voltage vi.

Any suggestions? Examples?

Program it as you would a potentiometer.

from the AndyMark web page](http://www.andymark.com/MA3-A10-250-N-absolute-encoder-p/am-2899.htm) for this product:

The MA3 is a miniature rotary absolute shaft encoder that reports the shaft position over 360 with no stops or gaps.This is the Analog output version

Yes, that is what the page says. However, if you click on the LabView Tutorial Snippet link, it shows it programmed as an encoder with digital inputs. Furthermore, the potentiometer vi and analog input vi both take an analog input. Hence the confusion.

Thanks for the quick reply, Knufire! In another thread, Mark McLeod cautions against using averaging or normalizing (last post here), which is what the get potentiometer vi does, as I understand it. Perhaps he was talking about the digital version of this encoder.

Hmm. You might want to mention that to AndyMark.

If you download the datasheet it’s pretty unambiguous that the output is an analog sawtooth voltage waveform.

I think you’re a bit confused because of the use of the word “encoder” in different places, where a more specific term may be more appropriate. The Labview tutorial you linked is for a quadrature encoder (Such as the US Digital S4T, E4T or the Grayhill 63R). These measure angular displacement in both directions, and use two digital inputs. If you want to know why, this goes a little more in depth as to how quadrature encoders work: http://www.creative-robotics.com/quadrature-intro

I haven’t looked into the LabView API much, but based on what you said about the Potentiometer VI, I would avoid using for the MA3 and just use the Analog Input VI. Since potentiometers are fully analog sensors, electrical noise will cause small fluctuations in the signal even when the sensor is still (which is what the averaging tries to filter out). This doesn’t happen as much with the MA3 because (presumably) the internal workings of the sensor are digital and there is onboard circuitry that converts the digital signal to an analog sensor.

Second what Ether said, the Andymark example is incorrect. Shoot a quick email to support at andymark dot com and they’ll get right on it.