#PHOTOREFLECTOR REFLECTIVE TAPE HOW TO#

I still don't know what the counters were counting but by switching the encoder power from the same power supply as used by the servo to the USB power shared with the QuickStart, the noise problem completely went away. I wondered if there were some high speed blips the Spin loop wasn't catching that the counters were but the o'scope traces were nice and clean even at 50us per division (the highest res setting). I still had severe disagreements between the counts from the Spin loop (which were accurate) and the counts generated by the Propeller's counters (which could be very inaccurate). I found my black masking tape from Parallax and the oscilloscope showed nice clean high and low pulses coming from the encoders. Here's a picture of the encoder mounted inside the servo with the gear by its side.

It turns out the Gorilla tape was only causing a small part of the problem. In case anyone is interested, I've attached the spin file to this post.Įdit: See post #77 for most recent version of code.

#PHOTOREFLECTOR REFLECTIVE TAPE HOW TO#

So far, my idea of having an encoder inside a servo isn't working out well ( again).Īny thoughts on how to reduce the noise from the servo motor? The Schmitt trigger made it so the 'ina' technique also picked up the noise (not as much as the counters though). I tried adding a Schmitt trigger to see if it would help. With the motor on, the counters must be picking up some noise because they the start counting very quickly when the gear is moving slowly. This next output is with the motor powered and turning relatively slow. As you can see the two techniques used to read the encoder are in agreement. The above output was with me hand turning the gear. With the servo off, I can manually spin the gear and the two different techniques agree.Įnter number to indicate which mode to use.Ĥ) LEDs display binary count of counter b.ģ) LEDs display binary count of counter a.Ģ) LEDs display binary count of both countersġ) LEDs display binary count of ina changesĠ) LED display encoder pin state (on or off) '' Must be called from cog where counters were setup.ĭebug.str(string(13, 'ledDisplayMode = ')) This isn't as supseptible to noise as the technique using counters.Ĭtra := 010 ' set up posative edge detection from page 159 of the PEK.Ĭtrb := 110 ' set up negative edge detection from page 159 of the PEK.

OldState := 2 ' impossible state so currentState won't match on the first loop I'm using two different techniques to monitor the encoders with the Propeller. I'm getting a lot of noise on the encoder line when the servo is turned on. I've been having some trouble with my plan to add an encoder to a servo. I didn't notice that the orientation made any difference but that probabl y depends on how wide the stripes are. I also used a 10k pullup rather than the 6.8k in erco's schematic. I had no problem with an inkjet striped wheel (I did use the highest quality setting) - see the video !?p=1203122&viewfull=1#post1203122 Have any of you tried a Sharpie to darken the stripes? I've read ecro mention inkjet printed stripes don't get picked up by the sensor. Has anyone used these with a Prop yet? I don't suppose and extra 4.7K resistor on the output line will interfere with the sensor's function? Is there a prefered orientation for these? should the two dots on the sensor be parallel with the encoder lines? Is there in info about how thin the encoder lines can be? I figure I'll try to keep the lines at least 1/8 inch wide. Okay, I'm about ready to drill some holes in a servo case and attempt to read a striped pattern on one of the gears (yet to be affixed).

Photoreflector reflective tape