Home-made anaQuad discs!

A project log for anaQuad!

Reliably increase your quadrature-encoder's resolution using an ADC (or a bunch of comparators?)

eric-hertzEric Hertz 04/30/2016 at 09:295 Comments

@Logan linked an interesting and easy-to-build encoder-disk method in a comment over at @Norbert Heinz's #Self replicating CNC for 194 (or more) countries project (which has a bunch of ideas for various positioning systems)...

Check out this guy:

I'll let the image speak for itself:

So, the current implementation of anaQuad(4x) would give 16 positions with that disk, per revolution. But, that could easily be bumped to 32 via software, and there's no reason the disk has to have only two "poles".

I think something like this would be easier to build than the ol' "slotted" style disks, since slots would require a tiny (or masked) sensing-area, etc.


danjovic wrote 04/30/2016 at 12:13 point

Clever idea! I think that printed magnets with hall sensors would work as well as an analog quadrature disk.

  Are you sure? yes | no

Eric Hertz wrote 04/30/2016 at 20:54 point

Good call. 

Even, maybe a regular-ol' disk-magnet, assuming its poles are both in the same plane. At some point, of course, those magnetic position-sensors designed for the task would probably give much higher resolution (and accuracy)... Though those things promise resolutions in the thousands per revolution, I wonder how accurate they can really be considering a tiny bit of misalignment would completely change the fields entering the sensor. Can't recall who it was, off hand, but someone 'round here set up a nice jig to test one of those out with good results.

Although, now that you mention it, another decent sensor idea for e.g. linear tracking might be just to record a sine-wave on an audio-cassette tape and stretch it out. Hmmm... I wonder if hall sensors can detect those tiny fields, 'cause I think a regular pick-up would only work if it was moving (and its amplitude would change with speed?).

Mind if I thank you for the skull, here? I think I've spammed your profile page a bit. ;)

  Are you sure? yes | no

danjovic wrote 04/30/2016 at 21:22 point

You welcome! I am always grateful for your skulls!

  Are you sure? yes | no

Eric Hertz wrote 04/30/2016 at 21:01 point

Hmm, got me thinking a bit more... A triangle-wave (printed or magnetic) would remove the need for anaQuad altogether, a simple *single* ADC-reading (and comparison to the prior) would be all that's necessary to determine direction. Only confusing-factor software-wise would be what happens when the sensor is "too large" and averages out the high-to-low transition... Might look identical to a fast-reverse. Guess two sensors are still a good idea.

  Are you sure? yes | no

Eric Hertz wrote 05/02/2016 at 08:51 point

Hrm.. by which, of course, I meant sawtooth-wave, or a right-triangle, maybe...

  Are you sure? yes | no