I’m looking for something that can be used as a passive (unpowered) shaft encoder which also provides direction of rotation. I can envision two types of devices. One would be a potentiometer that would sweep through some resistance (0 to 10k for argument sake) in a single revolution (360 deg) then would start all over once you turned past 360 deg. In other words I need an infinite turn pot that repeats the resistance taper with each revolution. An alternate configuration would be an infinite turn pot that had say 120 or more discrete resistances per rev. Any ideas where I get a FIRST legal device as described? How about ideas for making one?
Thanks,
James Jones
Engineer (Mech) / Coach
Team 180 SPAM
It sounds like what you want is a grayscale mechanical encoder, which you can get from digikey and connect to the digital inputs.
This won’t give direction of rotation. But, you can keep track of last iterations value and determine which way you’re going.
You just have to make sure that you don’t go through more than 1/2 revolution during your cycle. If you are going to go through more than one revolution, you can attach your encoder through a gear.
I don’t think there are any infinite turn potentiometers out there. If there are, they probably have infinite resistance too, which would make them incompatible with the RC analog input.
Other legal (last year) ways of measuring continuous shaft rotation include optical sensors, optical encoders, and hall effect sensors. These options are all powered and may require a custom circuit to generate a reading that the RC likes.
The above mentioned encoders are probably best, however if you really want to use a potentiometer, there is a way to modify pots for continuious rotation… well sortof (there will be a region of several degrees where the resistance will be an open circuit.)
But anyway here’s how:
1)carefully bend back the tabs whichhold the metal back on.
2)Directly opposite the point where the wiper contacts the circuit board, there is a little part that protrudes out further thna the rest of the platic disc. Cut it off. Alternatively you can bend out the little indention in the metal cover.
3)put the thing back together.
I do not recomend this.
There are a lot of ways to handle this one. You didn’t say how much resolution you needed so let’s just look at one tick per revolution. There are reed switches available from lots of places. These switches close when in proximity to a magnet. Make up a mount where you can get two switches slightly separated (i.e. slightly different angles close to the shaft) Attach a magnet to the shaft and as the shaft rotates first one and then the other switch will close. All you have to do is sense which switch is turned on first and you will be able to tell direction. If you want more resolution, add more magnets or more switches.
What exactly will you be using this for?
Pots are available off the shelf that have no stop at the end of the resistance element, and the element seems to extend further than the usual 270-290 degrees. I would worry that they might get a little noisy when used on drive axles, even more so on higher speed gears.
We tried an optical encoder system using the Banner sensors and black & white paint on the wheels, feeding a PIC. It proved unnecessary, and wasn’t used in competition.
Hello,
Can you put a cam on the shaft? If you can, then throw 3 limit switches so that they get turned on and off as it rotates. You can tell the direction and speed by what order they get pressed, and speed by the amount of time between presses. To increase you resolution, add more switches.
Good Luck,
Kevin
Potentiometers which recycle at 360degrees do exist. It took a couple hours of searching through local surplus stores, but I found 10, over the course of the build period in 2003.
We used these on our swerve system.
If you have any questions specific to our implementation of the system, I’d be more than happy to answer them.
At the transition point from lowest to highest resistance on the pot, you’ll most definitely have a couple degrees of deadspace, which can be fixed by using two potentiometers for each enclosed system, one facing one direction, and the other in the opposing direction. When you’re oriented in one of the two deadzones of the system, you can decide to take input from whichever potentiometer isn’t in its own deadzone. However, for our robot, we did no such thing, and it worked fine, just a thought for future seasons.
This will give you a resolution of 256 ticks per revolution, which is better than some of the lower-end rotary optical encoders carried by DigiKey, which are not too cheap, if I recall correctly.
As for FIRST’s regulations, you can use any pots (as long as they’re 100k, if I recall correctly).
I remeber a while back on the official First forums that some team wanted to use a steper motor as a encoder. They said it was ok. This might work.
*Originally posted by Adam Y. *
**I remeber a while back on the official First forums that some team wanted to use a steper motor as a encoder. They said it was ok. This might work. **
Just off-hand, one of the problems I could see with this (or using any relative encoder with a decent resolution):
A stepper motor (as far as I can tell, from what I know about how they work) would act as a relative encoder, letting you tell which coil of three or four (or however many the motor has) is “active”. I’m thinking of it as being a little dumber than quadrature output from an actual encoder. With a high enough sampling rate, this would be fine, because you could just continuously sample the outputs from it and track changes (keeping a sum of positive [CW, arbitrarilly] and negative [CCW, arbitrarilly] changes, for absolute angle). Assuming you use a standard stepper motor (200 ticks per rotation), in order to keep proper track of absolute angle (without using a custom circuit as a latching system), you’d have to keep the rotational velocity down to approximately 0.2 rotations per second (assuming the RC samples at 40 cycles per second, which quite possibly won’t be the situation, next year crosses fingers).
It is an interesting thought, though.
Thanks for all the info. I’m going down the 360 deg pot path for now. I’ve identified a few potential vendors on the web but looks like they are a little pricey and 100k is not common. If anybody knows a good source let me know.
As for what I’m wanting to use it for…well the simple answer is a shaft position sensor. I was hoping to relieve some of the work load on our electrical guys by having a sensor that did not require a bunch of custom electronics to interface with the oi or controller.
James Jones
Engineer/Coach
Team 180 SPAM
*Originally posted by JamesJones *
As for what I’m wanting to use it for…well the simple answer is a shaft position sensor. I was hoping to relieve some of the work load on our electrical guys by having a sensor that did not require a bunch of custom electronics to interface with the oi or controller.
All the more reason to have (a) project(s) going over the summer. Having a self-contained system for doing exactly this before the season starts (prototyping only, obviously) could be a very useful and educational endeavor for the kids.