Dead simple hall effect pot replacement...

If you wanted an crazy cheap way to replace the noisy pots in flight controls, this is the way to do it. (I originally posted this on SimHQ yesterday)

Here's the parts you need:

http://www.geneb.org/images/parts_required.jpg

The parts shown are a 22MM bearing (just a little shy of 7/8" diameter), two 1/4 x 1/4 neodymium magnets, an Allegro A1302 hall effect sensor and a Bic pen body.

Using a vice, carefully press in a short segment of pen body into the bearing - do NOT hammer it! The pen body plastic is very brittle. Steady force using a vice will "drift" in the pen body quite easily.

Next, glue the two neodymium magnets to opposite sides of the pen body - the magnets should be oriented N-S (they'd come together if the pen wasn't in the way). The way I did mine was to rest the magnet on the inner race of the bearing with a business card keeping it from physically touching the race. I then used "thick" CA (cyanoacrylate) to glue the magnets in place - have some kicker handy to speed up the cure process.

You should end up with something that looks like this:

http://www.geneb.org/images/bearing_..._and_shaft.jpg

Now you need to build the board that will hold the sensor in place. I used a small slice of copper clad perf-board. This holds the 3 pin connector and the A1302 very well.

http://www.geneb.org/images/circuit_board.jpg

Next, you want to get some 1/2" or 3/8" plywood and bore a 7/8" diameter hole in it using a forstner bit. Dont' bore the hole completely through - leave about 1/8" of material. Then drill a 1/4" hole using the same center as the 7/8" hole. I use a set-screw to hold the bearing in, but you could easily hot-glue it in if you're careful to keep the glue out of the bearing races. Tape the bearing in for now so you can align the hall sensor properly. A properly aligned sensor will fit in the center of the pen body with the magnets on either side of it.

You should end up having something that looks like this:

http://www.geneb.org/images/sensor_aligned.jpg

Assembled, it should look like this:

http://www.geneb.org/images/asm_front.jpg

http://www.geneb.org/images/asm_back.jpg

You can connect any kind of pushrod arm, gear, or whatever to the pen body - just don't apply too much side force to it. Too much side stress will crack the plastic. This setup gives 180 degrees of usable travel and works very well! The bearing size that you get should have an inner diameter just a tiny bit less than the widest point on the hexagonal Bic pen body. This allows for a very tight friction fit that doesn't require any adhesive to hold it in place.

You can easily custom design your own bracket for this - this was a simple prototype that proved out the technique and a refined version will end up in my '109 as well as a replacement for the pitch axis sensor I screwed up in my BRFS gimbal build. :)

Here's the device in operation:

http://www.youtube.com/watch?v=CVRjqvjV6_I


g.

wow now thatan answer to a problem i had with my race sim seat . many thansk for that really simple but robust method i know i for one will make use of that -- By The way waht was the range capability i,e lowest reading to highest i understand that you'll not get a 0V reading or +5v so what were you able to get overall for min and Max Values my Pc is playing up ( taking forever to load You tube stuff :-( at work during lunch break ) so could not see full vid -- either way big thanks for a great posting :-)

wannabeaflyer, I get a range of about .13 to 4.92 volts with the current setup.

If you build one, post pics!

g.

Hi,could you tell us how to hook this up? will it blow if reversed? thanks Jim

Geneb

Thanks, a good article. As a matter of interest, what is the resolution, ie how many steps per 180 deg rotation?

David

B1900Mech: The hall effect sensor should be connected according to the data sheet. Being a TTL part, if you hook it up backwards, you WILL let out the magic smoke. This is A Bad Thing(tm). :)

David: The hall effect sensor has no "resolution" per se. The number of "steps" depends on the size of the ADC you're connecting it to. For example, if you're using a 10 bit sample, you'll get 1024 steps. If you're using a 12 bit ADC, you'll get 4096.

g.

Thanks for the info! Do you have to index the center point of your yoke at the max V output?
also,where did you get the 3 pin connectors?

That is awesome :cool:

The center point of the input is pretty obvious - if you build one, you'll see what I mean.

I get most of my connectors from Jameco.

g.

Hi,

just a small question : how many degrees do you need to turn the pen for the fuill range of 0 to 5v ?

does this change with the distance to the magnets or the strength of the magnets?

Greetz Peter

Quote:

---

Originally Posted by pdpo

Hi,

just a small question : how many degrees do you need to turn the pen for the fuill range of 0 to 5v ?

does this change with the distance to the magnets or the strength of the magnets?

Greetz Peter

---

Hi Peter - the operating range is roughly 160-180 degrees. You need to keep the neodymium magnets as close to the sensor as possible in order to get the best result.

g.

FWIW, this method offers 10-bit resolution.

The Teensy 3.0 offers either 13 or 16 bit resolution, I'm not completely sure which, on its fourteen analogue inputs, for $27. And it can be configured to act as a joystick.
Teensy 3.0 + header ID: 1044 - $27.95 : Adafruit Industries, Unique & fun DIY electronics and kits

(I have no commercial interest in any of the manufacturers or retailers of Teensy!)