I set up a cordless drill to crank the spindle while I fettled the position of the bracket to get the correct duty cycles from both sensors and the correct phase angle between them. I want to see 50% +/-5% or so on the duty cycle and 90 degrees +/- 5 degrees or so for the phase angle.
Here's the sophisticated setup for cranking the spindle. You can just about make out the cordless drill driver behind the machine vise:
And after a bit of finessing of the position (trial and error, with a little bit of understanding), I ended up where I wanted it. Moving the sensor closer to the teeth reduces the duty cycle. This makes sense when you bear in mind that without a tooth to sense, the output is logic high. So the closer the sensor is to a tooth, the longer the output stays low. The skill is to move the bracket to correct the position of one sensor without screwing up the second. Luckily I had already (very skillfully!) positioned the sensors relative to each other so that I didn't have to play with the phase.
Forward (clockwise spindle rotation):
Yellow duty cycle - 50%
Blue duty cycle - 51%
Phase angle - 86 degrees
Reverse (anticlockwise spindle rotation):
Yellow duty cycle - 49%
Blue duty cycle - 50%
Phase angle - 88 degrees
Job done. The frequency of the signal with the cordless drill flat out is around 980Hz. With a 48t gear, that's about 20 revs per second, ie 1200rpm. And at 6000rpm, the signal frequency would be 5 times that ie about 5kHz which is well within the claimed bandwidth(s). There is some disagreement on the various datasheets as to the max frequency it seems.
Having checked that the bracket and sensor fixings are well tight, I can now start to reassemble the head. I'll try to spin the motor before every single component and fixing is in place in case I have to fettle it any more.
This is what the assembly looks like finally. The gap between the sensors and the gear is about 1.5mm.
No comments:
Post a Comment