Our programmers are dead set on using a potentiometer on the elevator, but I have never heard of a potentiometer that has more than 290 degrees of rotation. Is a potentiometer even usable on an elevator?
1 Like
10-turn potentiometers:
There are also continuous-rotation potentiometers, but I wouldn’t recommend them for an elevator.
You can also always gear-down your motion to a potentiometer to limit the rotation on a potentiometer (see James Bruton’s youtube channel, he does that a lot).
That being said, my personal recommendation for an elevator would be an encoder
There’s also various string potentiometer options available. Aliexpress looks to have some (relatively) inexpensive ones if you don’t need it until next year. 
An advantage of encoders is you can set the home position (down) with a limit switch, that is, reset counts when the limit switch is hit. With a pot, if there is some slippage in the linkage (bent or coupling twists to an unknown spot) then the programmer needs to introduce an offset to get you back running on the field again.
Don’t they fail the qualifications for VENDOR because they take too long to ship/lead time?
EDIT: 10.1 point “C”
ePacket items from Aliexpress usually get to me (SoCal) in 1.5-2 weeks from the seller’s ship time, which is usually around 3-5 days.
You can update the home position for an potentiometer using a limit switch, as well.
My short answer: yes, a potentiometer can be used in place of an encoder on an elevator. However, encoders are designed for this purpose, and are probably a better choice.
My long answer:
In 2015, our team used quite a few potentiometers in place of encoders throughout our entire robot. The pots are fine for telling you about what position an arm or elevator is at, but introduce some unneeded complications. Mounting becomes much harder, as well as coupling the pot to the mechanism (although a string pot does make this a lot simpler). We used some 3d printed pulleys with a large rubber o-ring coupling a driven shaft to the pot. I’ll admit that we didn’t design them very well, but they were constantly breaking mid match.
Another complication is on the software side. A potentiometer can tell you position like an encoder can, but an encoder object has a lot more functionality built in that you would have to write yourself. For example, finding the speed of the mechanism. An encoder makes it a lot easier to implement more complex movements beyond simple PID, such as Motion Magic/Profiling.
Basically, you can make it work either way, but an encoder is designed for the purpose of making precise mechanism control easier, and will work better given the same level of effort.
1 Like
We used a 10-turn pot for elevator height sensing last year. I recommend against the any of the Bourns models with multi-part plastic cases, particularly the blue ones. (There are a lot of knock-offs of this style available on Amazon as well.) One poorly placed impact will separate the case into its component parts and destroy the pot. After destroying 2 of the Bourns-style pots at our first competition we used one of the Vishay 534 (IIRC) series without incident for the remainder of last season + off season.
We have used potentiometers on elevators and arms many times , and they worked well. We always measure the number ornament rotations expected and choose the pot accordingly . There are 1, 3, 5 , and 10- turn pots available at Digi-Key . The advantage of a pot is that it always has home position - no zeroing needed unless you get slippage .
Why exactly do you think only encoders work with motion magic? The Talon SRX doesn’t care what sensor you use for closed loop control when it does motion magic. I’ll admit 10 bits of resolution for a 1m of motion isn’t ideal, but still. If you used a continuous rotation potentiometer, the Talon will even do the wrap-around calculation for you.
If you’re not going more than 9 rotations, and don’t need more than about 3 or 4 degrees resolution, then yes, the 10 turn pots are a good way to do things. If you need better than either of those, it’s encoder time. Fortunately, a couple of motor encoders (Talon SRX, and I presume SPARK MAX) and the roboRIO DIO’s have hardware which takes care of counting encoder clicks so you don’t have to write code which manages the microseconds.
1 Like
Where would we mount a pot? One of the reasons I want to use an encoder is because we can just use a redline encoder and be done with it. Our programmers tend not to really think about design (which is perfectly fine) so sometimes their ideas are challenging. Currently the biggest problem I can think of with pots is the mounting.
Ah.
My team has never actually used an analog sensor with a Talon SRX, so I often forget that it’s an option. That at least partially solves many of the issues I mentioned earlier, though I still maintain that an encoder will be easier to work with in most cases. I could see the value in a string pot on some mechanisms, though.
Our elevator pot is mounted on an l-bracket in from of the motor shaft - and connected with surgical tubing and zip ties . Add a little superglue if you are really concerned with slipping.
I have several times built or guided building a bracket to mount the potentiometer, sometimes mounting at the gearbox assembly bolts, sometimes at the encoder holes. The encoder would be mounted to the shaft with a 1/4" coupler or tubing. My favorite way, though, was just using a 1/4" coupler and some steel wire that was the right diameter to fit snugly in the encoder mounting holes to keep the potentiometer body from rotating. It had a few degrees of backlash, but that wasn’t important as it was for manually controlled rotation of a turret.
Pots of rotary encoders have similar issues when being used on moving devices. Neither is designed for side loading of the shaft and neither will survive pushing on the end of the shaft with any force. I have found that a gear train can be designed to couple moving loads to the pot shaft and satisfy the rotational requirements. An easy way to couple to the shaft is through a piece of surgical tubing. If you use tubing that has an ID of less than 1/4", you will get a tight fit to the shaft. A short piece will give very little slipping to the driven shaft but will give additional safety from over rotation of the pot shaft. McMaster also sells a variety of flexible couplers designed for 1/4" shafts. Select one that will prevent damage to the pot or encoder.
If using a pot will be more work for not more benefit, and taking into account that our CAD team is myself, plus .5 more people, would you say that if we can just toss an encoder on the back of a redline and be done with it that’s a better solution?
Alex,
You need to evaluate your need, the accuracy you would require and the physical environment when choosing an encoder. What you suggest may give you accurate data for speed but not for position. I would usually recommend the addition of limit switches to cal the encoder and prevent movement beyond the mechanical limits of your lift design.
We will be using limit switches at the top and bottom