I've spent several months looking for nice industrial grade quality rotory optical encoders on E-bay but just not found anything at nice hobby prices avalible. Finally I stumbled upon B&G micro and they seemed to have a nice price on one without a lot of details provided. I went ahead and ordered two and they arrived today.
They are marked made by OAK/Grigsby, part # 91Q128-43-00410. I assume this means 128 steps per rev.
It has kind of a useless 4 pin connector on the end of the 5 inch 4 conductor solid wire flat cable. I just lopped off the connector and soldered the bare wire ends to a 4 pin header. I placed ground on pin 1 (marked with a stripe) and +5vdc on pin 4. Channel A & B are pins 3 & 2 respectively. I put my two channel scope on pins 2 & 3 and could see very nice quadurture square waves being generated as I turned the knob in either direction. Being optical there should be no contact bouncing like with mechanical encoders and the pules seemed nice and square to me.
The encoder has a nice feel with a solid drag and no detents, feels just like a quality hi-fi volume control feels. The body is 1" square and about 1/2" thick. It uses standard 1/4" knobs, so that is nice.
All in all this is quite the steal for $5 each (or $4 if 4 or more) in my opinion and as I said I've been looking for quite a while. Now I need to decide if owning just two is enough.
Thanks. I use rotary encoders all the time, and I'm always on the lookout for a good one.
I usually use them just for knobs, but I have a couple of measuring gizmo's that need high-res. Ever see one with .1° or better resolution for less than $1,000?
They don't list the "91" series, but the other part numbers certainly match what you have there. The link should give you all the specs you need for the part, if they weren't shipped with the encoders.
Thanks, bobledoux! Unfortunately it's a PWM output with absolute position (usually better for most applications) but I need that quadrature output because I have to grab counter values on each edge. With that PWM frequency at 1khz, I would have to wait at least 1ms for each read, and my surface rotates too much in 1ms. (it's for a kinetics gauge that reads torsional motions on low durometer surfaces during impact)
On my Arduino, by wiring channels a & b to interrupts 0 & 1, I was able to get 128, 256 and 512 steps per revolution depending on if I activated both interrupts or just one, and/or if I set to interrupt on change or falling state. These are my best finds so far this year, I usually only have a couple of good finds a year so jump before they sell out