G11 Star Elongation: ONLY NEAR THE MERIDIAN
Posted 21 September 2012 - 02:39 PM
Then things go to Hades and I get E-W elongation.
I have done my darndest to eliminate any possibility of flex and AM USING VERY SHORT FLs (caps intended). I get this elongation even at 180mm FL. Balance does NOT seem to be the problem. I've tried both equi-balance and east-heavy to VERY east heavy with no apparent difference. Imaging load is light . . . less than 15 pounds with a 180mm camera lens.
I guide with the STi guiding kit and PHD.
I've tried all I can think of to fix this frustrating problem.
The crux of the matter is (again, caps intended), IT ONLY HAPPENS WHEN IMAGING NEAR THE MERIDIAN!
Thirty minute subs look great well off the meridian. Within 15 degrees I have to cut back to 3 minutes or so.
Posted 21 September 2012 - 03:21 PM
Posted 21 September 2012 - 03:46 PM
Posted 21 September 2012 - 05:26 PM
Have you noticed any cable bunching, anything restricting the movement of the RA axis? What does your guiding graph look like when approaching the meridian? Is it sending really large pulses E/W to try to correct, or is it behaving normally (which indicates some form of flexure)?
What kind of power supply are you using? If your powering off a battery or 12V supply, when the load shifts and it tries to draw more current the voltage may sag causing erratic behavior. I run mine with a custom 20A 15V power supply when I've got A/C power. Otherwise, I run a car battery through a power inverter and then to the power supply to guarantee proper voltage delivery (I know this is inefficient, but I was lazy and didn't want to build my own DC/DC step up power converter for running off the battery).
In my case when running the G11 off a 12V power supply or battery I had all sorts of speed control issues and strange random slews, etc. When I moved to the 15V 20A supply all those disappeared.
Posted 21 September 2012 - 08:07 PM
Posted 22 September 2012 - 06:23 AM
1/ I am unsure if this was always a problem or a recent occurrence. I always had some poor subs but only recently sleuthed it out to being only when very near the meridian.
2/ Power? Not likely. Using AC.
3/ Symptoms point to flexure but I can't see how. I am away from the gear now and can't post pics. I've tried several configurations. Latest is a Casady plate bolted directly to the G11. Camera (ST8300M) is bolted directly to a Losmandy D style dovetail through the 1/4”X20 tripod mounting screw. Lens is mounted to camera and is firmly supported under its objective end. STi guide scope mounted securely “Joey” style to dovetail. Everything is TIGHT!!!! I've used other configs too that were perhaps more prone to flex. My problem led me to this which I just can't see flexing unless the whole D plate is drooping or twisting. Seems inconceivable.
Flexure only at the merdidan and with such short Fls seems highly unlikely with this configuration. Weight is absolutely trivial too considering what this mount should be able to carry with ease.
4/Guiding seems normal. Nothing I can see unusual going on in the magnitude or size of the corrections . . . again pointing to flex I suppose. But only very near to the meridian??? And at very short FLs???
Posted 22 September 2012 - 09:02 AM
One thing that does change as you approach and drop away from the meridian is the amount of torque required by your motors ... the amount depends on how much out of balance your setup is.
I'd vary your balance a bit and see if you can either make the situation worse of better.
Have you ever manually checked the worm/ring gear spacing or tightness around the whole circumference of the ring gear? You're looking for any tightness or binding areas as you slowly drive the worm (without any load on the ring ... and then with your normal load).
Posted 22 September 2012 - 10:14 AM
If you aren't seeing any changes in guiding, it does seem less likely to be power, gear binding or a balance shift. Power could still be an issue though if your AC/DC converter is a 12V lower amperage model. The only reason I push this point is because your issues seem eerily similar to mine when I was running with a 12V power supply. It could supply 30A, so power delivery was not an issue. This told me the G11 really needs 15V+ (and probably over 10A, at least when slewing) to run correctly.
Does the imaging lens have moving elements? I've never really though about this before, but it does seem like most camera lenses are designed to be held horizontally, not vertically, so I just wonder if something is shifting around in there.
Do you know anyone with an off-axis guider setup? It might be difficult to get the off-axis guider between your camera and lens, but you could stick their whole scope/guider/camera setup on your mount and at least rule that out.
Posted 22 September 2012 - 11:18 AM
1) Lens barrel creep. One of my lenses has this nasty habit.
2) Your mount needs to be tightened up. I had all kinds of imaging nightmares until one day I said I have had enough and I just pulled the entire mount apart and rebuilt it. After tightening everything up, my problems went away.
Posted 22 September 2012 - 05:21 PM
1/ I am on the road and sort of unprepared to answer some specific questions . . . like what PS I use. It will be a week or so before I get home and have the gear available.
I have only camera lenses. I can't rule out movement of some internal elements but focus seems to hold true. Also, if this were so, a sub or two would be ruined and then the offending element should settle in and subs would be good again. Doesn't seem the case. All subs very near the meridian are poor. They become poor and they stay poor until I shift away from the meridian.
2/ Unfortunately, off axis guiding seems just not possible with EOS lenses period.
3/ I am reluctant to pull the mount entirely apart as Hilmi suggests. I have inspected the worm-gear spacing on the RA
axis and all seems well there . . . but my examination was very inexpert.
4/ I have nobody with whom I could swap gear to rule out the mount. So far I have ascertained the problem is definitely occurring with both the EF180mm f/3.5L and EF135mm f/2.0L EOS lenses. I have used other EOS lenses but have not specifically looked for this problem at the meridian yet. I will do so after the current imaging project with the EF135mm f2.0L is complete. They are of longer FL and the problem, if it occurs with them too, will be proportionately worse.
The problem I am encountering seems similar to what Steve Cannistra reported on this web page, i.e., inconsistent tracking near the meridian with his G11. His "fix" was a Temma 2.
I'm hoping for a . . . er . . . less drastic solution.
Posted 22 September 2012 - 05:45 PM
I've owned my A-P Mach1GTO for about a year and it is just about to need gear mesh adjustment because I can feel a very slight free play in RA axis. When I first purchased it brand new, the RA axis free play was tight.
Maybe your G11 just completed a break in period and needs RA mesh adjustment. From then on, it may not need adjustment for a while.
Does it matter which part of the sky when your mount is near the Meridian?
Posted 22 September 2012 - 06:05 PM
Worm adjustment rules itself out (because of the 4mn period).
I've seen balancing having some minor effect on star elongation. But if I understand correctly you can guide perfectly with 30 mn exposure until you are too close to meridian at which point 3mn are the max you can do. That is not a minor effect to me... It has to be something else. I mean, if I can 30mn perfect exposures it means that my balance really is good enough.
I always do a little bit of east heavy balance and more often that not I would image 10 degrees past the meridian without noticing any image degration in 10mn exposures. So I don't think it is a balance issue.
I am really thinking flexure also. That's pretty much the only thing left. Have you noticed if DEC was affecting the angle at which it starts messing up? I am not sure how your lenses are attached to your mount but I can imagine problems happening when the lenses are pointing straight up if they're only held by 1 screw.
Posted 22 September 2012 - 06:25 PM
Posted 22 September 2012 - 06:48 PM
In the interim, thanks all for the replies.
Posted 22 September 2012 - 06:49 PM
Posted 23 September 2012 - 12:06 AM
Posted 24 September 2012 - 02:11 PM
The meridian presents a different set of variables and this became evident with my older ST-4 guider. It has a calibration routine - and at the meridian being east heavy or otherwise has very little effect due to the position of everything. The counter weights are almost hanging straight down so how do you effect east heavy?
I found that loading the Dec axis east heavy did the trick. I'm a side by side set up and hung a large C clamp on the outside edge of the Dec plate. This caused the RA to become east heavy - improved tracking, but also loaded the Dec axis to one side. Otherwise with a neutral balance you are subject to the play in the worm/pinion. The guider will command a movement but the slack in the worm prevents any movement. Next cycle the guider says - wow way off - and does another bigger move. Finally the slack catches up - and it's way overshoot. Next cycle - guider tries to bring it back, but the slack in Dec won't take up.
I hope this is not confusing. I also solved the problem by making the worm mesh in Dec very snug to where the hand controller would message Dec lag off and on. The tighter mesh would then respond quickly to guider commands. But you need to constantly adjust mesh - and release it before each slew. So the C clamp became the answer for me. It's a 4" clamp. Can't hurt by trying it. It's offsetting the weight away from the counter weight shaft.
Posted 17 October 2012 - 02:12 PM
I said I would post images of the gear when I got back so here they are.
The problem persists. I have taken the G11 apart pretty much. Adjusted the RA backlash until it is so small I can no longer measure it by any deflection of the CW shaft. I have not touched dec backlash but it seems pretty minimal. Both axes rotate smoothly with no binding in evidence. Balance seems fine but I've played with it too, to no avail.
I don't see how alignment can be an issue. Twenty minute subs were fine off the meridian last night, just as they always are. As I entered within 10 degrees of the meridian, they fell off a cliff, just as they always do, and I had to cut back to 5 minutes.
Despite all my Herculean efforts to eliminate flex and my adjusting the RA backlash, quite successfully I think too, nothing has made any difference.
Guiding seems fine with nothing untoward showing up on the PHD graphs.
Len retains sharp focus which, in my mind, rules out any "barrel creep" in the lens. It is a pretty light lens in any event. It happens with other lenses too. Likely all of them but I have exhaustively determined it to be so with at least two.
I just can not image near the meridian with this gear and remain perplexed at which component is causing the problem.
Posted 17 October 2012 - 03:45 PM
With only tracking, the elongation is in RA only or is it RA/DEC? Since in tracking only RA axis is moving, if elongation is in RA or other direction, then that would be good data point to investigate further.
Posted 18 October 2012 - 10:31 AM
First off, OAG is not an option. No OAG solution is available for EOS lenses due lack of back focus.
The direction of elongation appears to involve both RA and dec axes. I attach an image showing a portion of a ruined sub along with superimposed test trails.
Posted 18 October 2012 - 11:22 AM
An OAG shouldn't be necessary at this FL.
Do you have any telescopes or just lenses? As a test, you could guide through the longest FL scope you can find and do your imaging piggyback. I'd try and rule out any lens creep.
Posted 18 October 2012 - 12:24 PM
You may try shimming between the band and the lens to get more clamping force and washers under the heads of the wood screws to give the screws a better grip on the band. See if any of that changes things and if so work towards some improvements based on the results.
Longer term, I would suggest changing the wood block to an aluminum bar and replacing the washer shims with a bar also to get more contact area between the camera and the dovetail.
Posted 18 October 2012 - 10:17 PM
If you are only tracking (not autoguiding), then it has to be flexure since (1) only happens around meridian - so not polar alignment issue (2) tracking only moves RA and flex will affect both - this is what you are seeing.
One thing you can try is to stack your images *without alignment* and see if the stars gradually move from frame to frame. This is indicative of flexure if the stars move in some semi-random way from frame to frame. You can make these stacks and compare star movement for frames taken around meridian vs. before/after it.
Looking at your setup, it seems like flexure is your issue. You can try making the connections more solid as others have suggested.
Posted 18 October 2012 - 11:37 PM
1) turn off dithering
2) take maybe 5-8 guided subs
3) load them into DSS
4) set the first image as the reference
5) align only
and 6) compute offsets.
if you see a steady X/Y drift, you're looking at flexure.
for the life of me i could not eliminate flexure on my G11/AT6RC/finderscope guider combo. i had to move to an OAG and now... round stars. but it was a big hassle and as you say, it's impossible with camera lenses.
my flexure problems definitely changed depending on the orientation of the telescope. when the telescope was more horizontal, the problem was worse. in your case it sounds like you have the opposite problem; flexure when the tube is pointing up.
i agree it seems insane that you'd have problems with DF at such a short focal length, but other than maybe your clutches slipping, i can't see what else would cause the problem.
what happens if you take unguided images?