Resolved Guider Not Bumping Mount Consistently

New problem, alas.

The AO has started not bumping the mount. It's odd; the pointing still is close to perfect, and the tracking is close to perfect when the seeing is good, with very little change in tilt for entire images. So I don't believe that something has happened to the polar alignment. Principally, the problem seems to come up when an image starts, and Maxim lined up the guide star too far from centered, and/or when there is horrible seeing. A variant on that is that, when it takes four images in a row (one hour total imaging time), and the mount doesn't need recentering (not enough aggregate wander) between images, it can go beyond the set threshold after several images. But I see no sign of the mount ever being bumped when the AO gets past the settings, and it obviously should when the tilt gets past the set threshold.

I believe the settings are correct:
ASCOM Direct, using ASCOM Telescope Driver for TheSky (mount is a Bisque Paramount ME); I am not aware of any changes made to setting which might have caused this. It's set to bump when the tilt is more than 75 or less than 25. "Bump time" is set for .15 sec.

The mount moves when I hit the "Bump" button, when it's not guiding (button grayed out when guiding). Maxim guides just fine when I'm not using the AO-X (camera is an SBIG STX-16803 with an STX Guider). Obviously, Maxim is able to control the mount.

Maxim will calibrate the mount, using both the camera control/guide tab (if SBIG Universal is selected as the camera), or the AO tab (if SBIG w/AO is selected as the camera). The AO clearly is well-calibrated.

Sometimes "ASCOM Direct" is grayed out while guiding; sometimes not. I don't see a correlation between this and when it's not bumping the mount.

This problem happens whether I'm controlling things with ACP, or using Maxim stand-alone.

I've been imaging with various SBIG AO devices since 2002, and I've never seen this happen before.

 

It is the same, except I updated this post as I learned more.

Any ideas what could be happening?

 

Ok, I'll increase the bump time.

I have attached a screen shot of the settings.

 

Well, there's good news, and there's bad news.

The first good news is that increasing the bump time may have solved the bump problem. The need for a bump seems mostly to result from the dither ACP puts at the start of each image; this is especially an issue if Maxim didn't properly center the guide star in the little box, or if the tilt is significant left over from the previous image (Maxim does not seem to automatically revert to 50% AO tilt to start each image). At 2 second bump time, Maxim just played ping-pong with itself, sending the guide star corner to corner, probably because it was over-correcting. At .5 second bump time, it didn't seem to be enough. I tried 1 second, and that seemed to work ok, but perhaps was a bit long. So I now have it set at .9 second, which seems to be working fine. Time will tell.

The bad news is that Maxim really does not want to keep the guide chip binned at 3. When the system opened, it was unbinned, which kills the system (I'm guessing it has to do with ACP's guided dither to start each image; it seemed to think that the actual photo taken of the unbinned guide chip was too big). Then trying to coax it into binning 3x3 caused a cascading series of failures, ending with me turning the camera and mount off, then back on, and rebooting the computer. Then it behaved. But this is still a problem, obviously. And, yes, if I set the binning at 3 in the AO box, then close Maxim, it generally comes back up that way. But obviously not always, since I had changed nothing since last time I imaged.

But more good news is that I haven't seen any sign of the 0.2Hz AO rate since I started binning the guide chip at 3x3.

 

...And now, out of the blue, after working fine for a long time, it's not bumping again. At all. No settings changed by me, nothing seems to be changed, but when it dithers just before starting an image, it doesn't bump. Even when it's at 100% tilt.

Merely stopping and restarting Maxim had no effect.

Turning off the camera (SBIG STX-16803), restarting the computer, then restarting Maxim and the camera, seems to have resolved it for now. But it is very odd behavior.

 

Correct. Dithering shoves the guide star across the full guide field by the amount you have set to dither. At this point it’s likely the tip-tilt has maxed out and it’s required to bump the mount to get the moved guide star into a spot that’s within the range of the tip tilt. You will notice (in the log!!) that, after a dither, ACP will monitor the RMS Wander (guider errors) until it gets 5 cycles in a row with RME Wander errors below the user-set limits (both X and Y).

This can take time. If the bumping is too aggressive it will oscillate about the “sweet spot”. On the other hand, if the bumping is too soft, it will take a long time to ease over to the “sweet spot”. Dithering is a bit like hitting the mount with a hammer, the more the dither the longer it will take to recover at the dithered position, if it can recover at all.

This is why I asked you to avoid dithering until you can first get it to work without dithering. Your system is the most complex and fussy of all: German Equatorial, rotator, AO guider, small-ish guide camera.

I’m pleased you found the recipe!!

 

I always am backed up; that's always good advice. But whether or not I am backed up doesn't seem to me to be relevant to why the system sometimes doesn't bump.

In the meantime, I'll just reboot the system more often, as rebooting seems to cure lots of ills....

 

It continues to do this, irregularly. But enough that all of my subexposures have elongated stars. When checking in on the system during the night, I several times have seen it be more than one pixel off center; then it corrects (without action on my part). But the result is lousy subs.

Any ideas?

 

How do you know it’s not bumping? Are you watching MaxIm to see the tip-tilt hit the limit and then stay for a while (multiple cycles)?

Limiting out could be due to other things. It is a complex non-linear feedback control system. For example, if the mount has “rough spots” in RA then bumping (which is a timed pulse) could be moving the mount by a varying amount of RA at different hour angles, and some bumps may be insufficient and/or ineffective.

I’m just trying to open the vista for problem solving here. I can’t recall exactly where but there is a place in the MaxIm camera control window where it flashes a message when it is bumping. By watching that, you can tell if it is bumping or somehow randomly skipping a bump (the current theory). It will take annoying boring time to stare at that place for a long time to detect flaky bumping or to convince yourself that it’s trying all the time.

 

I have seen it be beyond the set limit, and not bump. I have images where the image clearly is doubled (one star becomes two stars, as if the system was content to let things be a couple of pixels off for long enough for everything to be doubled).

Every image has every star very elongated.

I don't know what is causing, but, alas, I am no unable to image effectively until I/we figure this out.

Mark

 

Even the pointing exposures have streaked stars, so I'm not at all convinced it's a guiding problem. We're examining everything we can having to do with tracking.

 

Yeah, there's nothing obviously wrong. Tracking certainly is on, and it's tracking, but not well (tracks well in a macro sense, but not micro; it's as if the entire OTA has a very small, high-frequency vibration). They've checked the clutches and the balance. They're going to clean and regrease. And I've never done PEC (never saw the need), so it hasn't been lost. And this is a Paramount ME; it always could, and still should, track essentially perfectly, even at this focal length, for a lot longer than 10 seconds.

Given that a single ten-second image has elongated stars, and didn't used to have anything but round stars, I don't think the problem is guiding, or the AO. Like you, I'm inclined to think something's up with the mount; the guys at SRO will go through it thoroughly.

Mark

PS--I've tried to renew my support for Maxim, but I cannot sign on for some reason. Any help on that would be appreciated, also.

 

I just spent the last 45 minutes fighting with the website, ultimately failing. It rejected my credit card, although there is no problem with the card. And it doesn't let me look at what numbers I put in (I suspect I put in the wrong number), or enter another card.

I'm stumped.