- 7,699
- 3,787
Another excellent weekend, I had three chances to try out the various suggestions. Unfortunately, I have mixed results. To be fair, I've also noticed that over time, I have been setting more and more stringent requirements on the images I retain for stacking.
The two main things I tried this weekend were: better alignment and better mount balancing. Again, here's my starting point- a stack of consecutive images (@400%) showing oscillatory behavior in right ascension (RA, vertical) and drift in both declination (DEC, horizontal) and RA:
At 400%, these RA oscillations are around 250 pixels, corresponding to 1.6 arcmin actual angular motion. Not sure if that's within spec.
One of the challenges is the lack of a meaningful user guide for the GoTo drive. The drive does output some useful metrics- I get live updates for 'mount balancing' and 'alignment model parameters'. If you check out those sites, you will note the relative lack of useful information- those URLs are the Gemini manual.
Mount balancing:
The GM8 user guide is pretty clear about the mount balancing procedure- mount the scope and adjust the position of either the telescope on the sliding dovetail (for DEC) or the counterweight (RA) until the mount holds a static position.
One clue I have about the possible underlying problem is while I have to position the telescope dovetail accurately to within 1 mm for DEC balancing, I can slide the counterweights within a 2-inch range of travel and the mount still stays (apparently) balanced. That is, the DEC axis rotates very freely but the RA axis seems 'resistive'. It's worth mentioning that the GM8 manual suggests leaving the mount slightly unbalanced... "This is done so that the worm gear is pushing against the slight load", resulting in improved tracking.
Looking at the GoTo drive output, the numbers have been very resistant to change- the 'Y' balance (DEC) stays at a constant 6% and a -2 offset, while the RA ('X' balance) rapidly oscillates between 6%-10% (poor hard-working jalopy motor!) and with the same -2 or -3 offset. Those numbers don't change, regardless of how well (or intentionally poorly) I balance the mount. I do understand what the PWM % numbers mean but have no clue what the 'offset' means- and I believe the important readout quantity is 'offset' which should be close to 0.
So on different nights, I moved the counterweights to different positions to see if there was any observable effect, first is with the counterweight too close to the mount (telescope heavy) and the other with the counterweight too far from the mount (counterweights heavy):
No substantive difference, the oscillations in RA are the same magnitude- 250 pixels (plus or minus). The reduction in DEC drift is discussed below...
One possible cause of the RA oscillations would be friction (somewhere) in the RA drivetrain, I think it's called 'stiction'. Although I store it with a dust cover, some of the gearing mechanisms may need to be cleaned after 10 years of use, but I'm not sure which. I've taken the thing apart before, so maybe I have a project...
Mount alignment:
One thing I tried did (minimally) impact the RA oscillation amplitude: better alignment. After I aligned to the same few stars 3 or 4 times, the model fully populated its parameters and the mount balancing changed- the DEC PWM was reduced to around 2% and the offset was 1, the RA PWM was still bouncing between 6-10% but the offset was also around 1. Here's what resulted:
These RA oscillations are around 200 pixels, which is better. Also, the RA drift is gone. Now, I'm not sure why (electronic) improved alignment impacts (mechanical) balancing, but that's what the drive output indicated.
Anyhoo... here's an updated image with Neptune and Triton:
The two main things I tried this weekend were: better alignment and better mount balancing. Again, here's my starting point- a stack of consecutive images (@400%) showing oscillatory behavior in right ascension (RA, vertical) and drift in both declination (DEC, horizontal) and RA:
At 400%, these RA oscillations are around 250 pixels, corresponding to 1.6 arcmin actual angular motion. Not sure if that's within spec.
One of the challenges is the lack of a meaningful user guide for the GoTo drive. The drive does output some useful metrics- I get live updates for 'mount balancing' and 'alignment model parameters'. If you check out those sites, you will note the relative lack of useful information- those URLs are the Gemini manual.
Mount balancing:
The GM8 user guide is pretty clear about the mount balancing procedure- mount the scope and adjust the position of either the telescope on the sliding dovetail (for DEC) or the counterweight (RA) until the mount holds a static position.
One clue I have about the possible underlying problem is while I have to position the telescope dovetail accurately to within 1 mm for DEC balancing, I can slide the counterweights within a 2-inch range of travel and the mount still stays (apparently) balanced. That is, the DEC axis rotates very freely but the RA axis seems 'resistive'. It's worth mentioning that the GM8 manual suggests leaving the mount slightly unbalanced... "This is done so that the worm gear is pushing against the slight load", resulting in improved tracking.
Looking at the GoTo drive output, the numbers have been very resistant to change- the 'Y' balance (DEC) stays at a constant 6% and a -2 offset, while the RA ('X' balance) rapidly oscillates between 6%-10% (poor hard-working jalopy motor!) and with the same -2 or -3 offset. Those numbers don't change, regardless of how well (or intentionally poorly) I balance the mount. I do understand what the PWM % numbers mean but have no clue what the 'offset' means- and I believe the important readout quantity is 'offset' which should be close to 0.
So on different nights, I moved the counterweights to different positions to see if there was any observable effect, first is with the counterweight too close to the mount (telescope heavy) and the other with the counterweight too far from the mount (counterweights heavy):
No substantive difference, the oscillations in RA are the same magnitude- 250 pixels (plus or minus). The reduction in DEC drift is discussed below...
One possible cause of the RA oscillations would be friction (somewhere) in the RA drivetrain, I think it's called 'stiction'. Although I store it with a dust cover, some of the gearing mechanisms may need to be cleaned after 10 years of use, but I'm not sure which. I've taken the thing apart before, so maybe I have a project...
Mount alignment:
One thing I tried did (minimally) impact the RA oscillation amplitude: better alignment. After I aligned to the same few stars 3 or 4 times, the model fully populated its parameters and the mount balancing changed- the DEC PWM was reduced to around 2% and the offset was 1, the RA PWM was still bouncing between 6-10% but the offset was also around 1. Here's what resulted:
These RA oscillations are around 200 pixels, which is better. Also, the RA drift is gone. Now, I'm not sure why (electronic) improved alignment impacts (mechanical) balancing, but that's what the drive output indicated.
Anyhoo... here's an updated image with Neptune and Triton:
