View Full Version : US Digital quad encoder and BS2

03-09-2010, 12:37 AM
Hi all,

I'm continuing to work with my high altitude balloon project, will soon test a continuous rotation servo with a wheel encoder (eight reflective stops for simplicity) and the QTI sensor to read the reflected light.

I'm also thinking about using a small DC motor and US Digital with shaft encoder (E8P) to accomplish the "rotate 90 degrees, stop, repeat in 5 - 7 seconds" motion I previously described at http://forums.parallax.com/forums/default.aspx?f=10&m=426909&g=427481#m427481. Thanks for the suggestions! I've read Phil's thread http://forums.parallax.com/forums/default.aspx?f=5&m=169803, which has been helpful. Information about the Nubotics Wheel Watcher kit and quad encoder, which I understand can be used with the BS2 (monitoring only one channel's output), has contributed to my understanding as well.

My question: can I get away with using a BS2 (and no coprocessor), read around 200 pulses/revolution from the motor/encoder set up and give a stop command in that time frame? A main routine will monitor a photoresistor that will signal (after a camera flash) that it's time to turn the servo or DC motor. That's the only thing that will be happening in code. I like this solution because it eliminates some of the environmental factors (extreme cold, passing through troposphere and high wind) that might affect a semi-exposed IR sensor and the code wheel it must read.

I'd really appreciate anyone's experiences using the US Digital encoder (or any shaft encoder) with the BS2!

03-09-2010, 12:56 AM
As long as the pulses are in the 20 per second range as a maximum, you'll probably be OK.

Heres one I did using the BS2 and a potentiometer-style encoder ( 100 pulse per rev, I believe)



Tom Sisk


03-09-2010, 04:22 AM
Do you plan to read encoder pulses while monitoring the photocell? RCtime can take a relatively long time to execute, based on your component values, and you might miss some encoder pulses. A single-input tach sensor requires only 1 input, a quadrature encoder requires 2, and twice the processor overhead.

Do you really need 200 ppr, or do you only need to rotate 90 degrees and stop? You could achieve that with a square cam & limit switch (mechanical or hall efffect or reed switch) easily enough.

Alternatively, you could cycle a servo that would ratchet a secondary wheel 90 degrees per cycle.

"If you build it, they will come."

03-09-2010, 04:44 AM
Erco, just casting about for an alternative that would let me internalize software and hardware in the balloon payload to protect from elements... I guess my ignorance is showing. I should probably just stick with a continuous rotation servo with wheel encoder and sensor - either IR or reed switch - OR use a simple whisker switch as a mechanical option to meet the "90 degrees, stop, repeat in a few seconds" requirement. I've got everything else, and I ordered the QTI sensor today to experiment with on the Board of Education. I should just keep it simple. http://forums.parallax.com/images/smilies/smile.gif