How to choose potentiometer

thanks but actually there is a problem because i want to limit the movement in +/-45 deg and the smcutility has a min clip value of 255 so with this value and the specs of my 360 hall sensor, if i set the center reference value to 2.5V or 512(adc value) the movement will reach an undesire +/-90 deg. and the solution to avoid this without use gears is changing the limits of the clip value to 512 for example , so im figuring out how to change it in the ino file.

i think a found a solution, in order to represent the movement of one sensor to other, for example a hall sensor 360CW can be represented as a 180cw mapping its input values to the desires output values, in a +/- 90 deg movement centered in 2,5V [512 ADC] the 360cw sensor goes from [255] to [766] and the 180cw sensor from [0] to [1023] , so we just have to map these values in the Feedback variable , im going to try it and check how smooth it goes

These pots are listed at 1000megOhms minimum, I keep reading 10k ohm min/max. Looks like neither of you had issues... but..?

Also, is there any advantage to going with the 180* vs the 360* for the same price? If the resolution is the same (12bit), then the 360 should give you a little more leeway in mounting errors, yea?

180*: 6127V1A180L.5FS TT Electronics/BI | Sensors, Transducers | DigiKey
360*: 6127V1A360L.5FS TT Electronics/BI | Sensors, Transducers | DigiKey

Spending more on 3 pots than I did on all my motor controllers is giving me anxiety. Thanks!

Good point on the higher cost vs probably not having to replace them every few weeks (or so it seems from some of these posts!).

I've checked out that thread and re-read that post several times this morning... from what it seems, no matter if you get a 90*, 180*, 360*, every degree is still worth "12 bits", right? So, if you use 90* of a 90* sensor, you're still only using the same 1080 bits of data that you'd be using if you used 90* of a 360* sensor... I think?

Meaning, a 90* sensor seems like it would be more accurate, but in reality, it's exactly the same. But I assume this can't be the case, because why would they even sell smaller than 360.

I design construction power systems, this small electronic stuff is strangely foreign to me and it's uncomfortable to admit that, haha!

I went ahead and ordered the 180* versions. If I go beyond 90* in any direction I'm in trouble anyway. Thanks noorbeast!

Thank you so much for this recommendation. I was having a terrible time 3d printing gears and designing a system with a hall effect sensors. they were not cheap in Canada (28$CAD) each, but well worth it.

6127V1A180L.5FS potentiometers seem to be out of stock everywhere, has anyone found a suitable replacement hall effect potentiometer?

Thanks @noorbeast i spent a couple hours poking around and I found 90s and 360s in the above potentiometer but as you say 180 is really the ideal for my sim. Trouble is I haven’t been able to find many other manufacturers who make a 180 hall pot. Most are 360s and the only others I have found in 180 are one by CALT that is $42 and one by P3 that is $67 which is really more than most would be willing to pay for a POT. Mouser claims to be getting more in stock in a month or so, so I may just place an order with them and wait it out unless I can find something comparable. This one seems to be a bit of a unicorn!

Amen to that! I found this interesting site that shows stock across all vendors. looks like some company(s) that use(s) these made some massive bulk buys in september and december. Good to know other vendors are expecting them next month as well

Hey Boost, they actually are different but in practice it’s not a huge deal. It’s the same kind of thing with “filling” an ADC. You get the best resolution if you use all the bits but that may or may not matter.

First off, 12 bits is 4096 steps and 10 bits is 1024 steps. With a 10 bit resolution over 180 degrees, that makes each step equal to 0.17 degrees. 10 bits over 90 degrees is 0.088 degrees per step - twice as fine of step size. If the actual range of motion matches the sensor, you can “see”/resolve smaller motions. But use too big of a range sensor (e.g. using a 180 degree sensor for 45 degrees of motion) and the smallest motion you can resolve is 4 times as the size as if they match properly.

For example and just using whole (unrealistic) numbers so it’s easier to think about, if you have 45 degrees of travel and your ADC can resolve 1 step per degree, you could detect a 1 degree movement with 45 total steps. But use a 180 degree sensor and you wouldn’t see a change in output until you move 4 degrees since 45 steps over 180 degrees is 4 degrees per step.

So which range of sensor you use matters but in practice, with enough bits of resolution, it may not be a big deal. There’s other considerations but that’s the jist of it.

The bigger problem comes in going the other way - you simply can’t use a 45 degree sensor in a 180 degree application. In a feedback loop you really don’t want your sensors to run out of range or things can break.

(edited for clarity)

Hi all,

I keep reading that with Arduino 10k is basically mandatory, but I do not understand why. I mean: the way potentiometers are connected, they form a voltage divider between the +5v supply and the tap on the rotary contact. This way, every linear pot will provide a voltage on the center tap proportional to the rotation angle - be the potentiometer 100k, 22k, 10k, 1k and even 1M and 100 ohm (though the latter two would be inadvisable due to respectively noise and power consumption). Can someone e please explain why 10k value keeps recurring in posts?

Thanks

The common, popular options right now for magnetic encoders, usually sold on mountable breakout boards:
AS5600 : 12bit
AS5048: 14 bit
MT6701: 14 bit
TL5012: 15 bit

Same ones used for encoding on FOC to turn BLDCs into servos