DavidH
25-04-2013, 06:46 PM
Hello Folks,
After many, many, many moons, have finally managed to put together a conversion for an Astrophysics QMD 900 (originally with stepper drive non goto) conversion to servo drive with a Sitech (Dan Gray) controller. This is proving to be the most successful version to date. Have previously tried toothed belts, helical couplers, model aircraft electric motor gearboxes, etc along with the current motors.
This version uses Pittman servo motors with 10:1 gearboxes, a further 25:6 gearbox, then a right angle gearbox to both worms. There are two Oldham couplers on both axes on either side of the right angle gearbox ( the second one is inside the worm housings and can't be seen in the photos). My last version had the RA motor in line with the worm shaft (i.e. without the right angle gearbox) but the dovetail plate could (and did) hit the motor in this configuration, causing some choice observatory language.
This version so far appears to be the most useable to date. The right angle gearbox takes the motor away from the path of the dovetail plate, and the 25:6 gearboxes give the motors plenty of torque to push the mount around. Speed is quite acceptable running off 12v. No idea of PE yet, but on past experience, the stepper PE will be hard to beat.
One issue I have is determining the exact number of motor steps for rotation of both axes. The motors have 500 step quadrature encoders, 10:1 gearboxes (which are actually about 9.98:1), 25:6 gearboxes, then worms which appear to be about 225:1.
So the calculation should be:
500*4*9.986931818*25/6*225 = 18,725,497 steps per rev.
This works OK for the Dec axis, but falls a bit short for the RA axis. By watching the scope encoder counts, I can get a reading from the RA axis for 180 degrees which results in a tick count of 18,754,680 ticks per rev. Back calculating gives a number for the RA gear of somewhere between 225:1 and 226:1. Is it possible that the RA gear has one tooth which is a bit smaller than the rest? I seem to remember at some stage finding a spot on the gear that was a bit tight, and noticing that the tooth at this point on the gear was a bit smaller than the rest.
Anyway, here are a few photos:
.
Cheers,
David.
After many, many, many moons, have finally managed to put together a conversion for an Astrophysics QMD 900 (originally with stepper drive non goto) conversion to servo drive with a Sitech (Dan Gray) controller. This is proving to be the most successful version to date. Have previously tried toothed belts, helical couplers, model aircraft electric motor gearboxes, etc along with the current motors.
This version uses Pittman servo motors with 10:1 gearboxes, a further 25:6 gearbox, then a right angle gearbox to both worms. There are two Oldham couplers on both axes on either side of the right angle gearbox ( the second one is inside the worm housings and can't be seen in the photos). My last version had the RA motor in line with the worm shaft (i.e. without the right angle gearbox) but the dovetail plate could (and did) hit the motor in this configuration, causing some choice observatory language.
This version so far appears to be the most useable to date. The right angle gearbox takes the motor away from the path of the dovetail plate, and the 25:6 gearboxes give the motors plenty of torque to push the mount around. Speed is quite acceptable running off 12v. No idea of PE yet, but on past experience, the stepper PE will be hard to beat.
One issue I have is determining the exact number of motor steps for rotation of both axes. The motors have 500 step quadrature encoders, 10:1 gearboxes (which are actually about 9.98:1), 25:6 gearboxes, then worms which appear to be about 225:1.
So the calculation should be:
500*4*9.986931818*25/6*225 = 18,725,497 steps per rev.
This works OK for the Dec axis, but falls a bit short for the RA axis. By watching the scope encoder counts, I can get a reading from the RA axis for 180 degrees which results in a tick count of 18,754,680 ticks per rev. Back calculating gives a number for the RA gear of somewhere between 225:1 and 226:1. Is it possible that the RA gear has one tooth which is a bit smaller than the rest? I seem to remember at some stage finding a spot on the gear that was a bit tight, and noticing that the tooth at this point on the gear was a bit smaller than the rest.
Anyway, here are a few photos:
.
Cheers,
David.