Skywatcher AZ GOTO alignment and accuracy

[DrRobin]

Hi,

I decided to start a new thread to discuss AZ Goto alignment and accuracy as I have been having problems with my Dob using either Goto or Auto-track.

The following is what I have learnt reading the threads on this forum, from the manual and what I have been able to piece together using guess work.

When I first got a Goto mount (Mak 127) I had assumed that alignment counted the number of steps between two stars on both axis and worked out how many steps you need for 1 degree/arc minute/arc second and also set the absolute position.

I now believe this is a false assumption and this is based on practical experience, careful studying of the manual and from what others report.

When a two star alignment is performed all I think the Synscan does is to work out an average absolute position for both AZ and Alt based on the two stars selected, the angular distance and number of steps on the motor is ignored. What really gives the game away is under two star alignment it says choose two stars at least 60 degrees apart and two stars with the same Alt is good.

If the Synscan was calibrating the step size to angular movement, you would need two stars about 90 degrees apart, preferably in both axis.

So I think the main purpose of calibration is to set the AZ absolute position and to a lesser extent the Alt absolute position. If this is the case an accurate initial North and level set up would be nearly as good. This is confirmed by the fact that if you are using tracking mode (or the auto-track) then that's all you need to do.

To make AZ Goto work correctly then you should have the scope base as level as possible, the location and date/time as accurate as possible. Some date/errors and small position errors will be taken care of by the alignment, but it will be no substitute for getting the basics right.

Alignment isn't close on the second star or tracking/position is not very good.

Both of these will occur even with good alignment if the motor moves but the scope doesn't, since the encoder used for position is on the end of the motor shaft, not on the end of the scope axis.

The Dobs have friction nuts which need very precise setting, too slack and the motor will move (especially true at slow speeds) but the scope stays in the same place. Too tight and the motor won't be able to move the scope against the friction. The same happens if the scope is not balanced very well.

The Dobs have dual encoders, but the motor encoder provides a minimum resolution of 0.8 arc seconds whereas the main axis encoder only provides a resolution of 109 arc seconds (taken from the manual by dividing 360 degrees by the number of steps. The main axis encoder is mounted on the shaft, so always moves with the scope regardless of friction nut tension.

Since you can manually push the Dobs and then it will track with the Goto, I assume that when the main encoder moves the mount updates it's position, but then only to a resolution of 109 arc seconds.

I am guessing that if you are tracking with the Goto and the motor moves, but the scope doesn't once it gets 109 arc seconds out, the Goto sees the error between the motor and main shaft encoders and probably jumps to correct? Not sure if this actually happens, I might be seeing this on some of my 2 minute exposures.

Anyone any thoughts? I would like to try and get to the bottom of this.

[ronin]

If you mean literally 2 star alignment and no start position then you are really only really doing a single star alignment. The first one is probably just a half defined start position.

The movement from the first to second star will have to determine how much drive in boith axis is required also it has to determine how well you have set up the mount and work out the compensation needed for this error in 2 axis, say +- 5 degrees either way as your possible error at start. Add in that there is a small error for Lat/Long as although you are in the GMT timezone you are probably not on the 0 Longitude so "time" errors for that to be determined.

I have a Meade and they "cheat". The start point is level and north, in effect a virtual star at that position. So the Meade 2 star alignment could be considered to be a 3 star alignment.

Effectively the first "star" is no more then some starting point from which the various measurements are taken but in isolation does not contribute.

I think they should all have the option of starting with the scope pointing at and with Polaris in the centre of the view.

[DrRobin]

Hi Ronin,

For the auto-track you start at north and level, there is no other set up. For the Goto, it doesn't seem to make any difference if you start at North and level, in fact the manual says point at a bright star and then switch the mount on.

From tests I have made, I don't think the Skywatcher does calibrate out the errors in moving between two stars, I just think it sets the absolute positions for AZ and Alt. The Meade and for that matter Celestron might be different.

The fact that they only use two stars for alignment and that the tip from the manual says that you get a better alignment if both stars are at the same Alt does mean that they are only really doing a one star align and assuming that the scope base is level.

I used to design closed loop control circuits many years ago, later some work on GPS and if it was calibrating out the movement required to move between two or three stars to get any accuracy at all you would need three stars (especially if it's going to compensate for scope level) and then ideally each star would be 90 degrees apart (in three dimensional space).

[SeedyF]

Robin - I only received my 250PX Goto on Friday so I may not have anything to add, but the goto accuracy certainly seemed wide of the mark. I followed many of your posts and others on here before plumping for the scope so thought I had the set up steps nailed. I spent a frustrating few hours on Saturday trying to get it working the way I thought it ought to with little joy.

After a first test on Friday night I realised I needed to have the scope levelled more accurately. My patio slopes slightly to allow drainage and is made of weathered sandstone so it's hopeless for getting level - my old tripod was easier obviously to level. So I build a platform on Saturday with adjustable legs that is sturdy and easy to level.

Once level, the scope horizontal and pointing north I followed the set up steps to the letter. I used Altair as star 1 and Scheat in pegasus as star 2 for allignment. The second star was in view in the 25mm eyepiece but only just so I centered it with the handset and confirmed allignment. Then I asked the scope to go back to Altair and it missed, the star was right on the edge of the field of view, now that can't be right, can it? I repeated all steps a few times then gave up. When I selected the park option from the handset the scope slewed to approximately North East, not North as it should have. At one point I did select a double star near Vega from the goto list and the scope went very close - possibly more luck than judgement.

I'm sure I am doing something wrong but can't figure out what. I have been doing the following (not sure of the sequence now that I am nowhere near the scope);

1. Level the scope
   
2. point it north
   
3. set altitude to zero
   
4. turn on handset
   
5. set date in mm/dd/yyyy
   
6. set the time
   
7. timezone to 00.00
   
8. daylight saving on
   
9. set location - checked in google earth, compass on my phone, google maps (converting decimal to minutes)
   
10. 2 star align - handset control to slew to first & centre in 25mm then finer set using 10mm eyepiece
    
11. Select 2nd star from the list and goto, same centering process as above.
    
12. Confirm allignment
    
13. Select an object to goto
    
14. Sigh deeply and shake head
    

I did read in the manual since then that when you centre the allignment stars you need to make sure your last adjustment is with the up or right key to ensure accuracy. I didn't do that and it's possible that made a difference.

Now there is the suggestion above that you should choose stars more than 60 degrees apart - well my two were nowhere near that far apart so that could be an issue, I couldn't see anything in my manual that gave this advice but I'll certainly try it. I may also take the scope somewhere that I can see more of the sky (the house is in the way) to select better spaced stars if that really does make a difference.

But for now, it's daylight, worse still it's cloudy and I'm at work. So this will have to wait for more favourable conditions.

[DrRobin]

Hi Chris,

Thanks for your reply, I think you are finding exactly what I found and your experience looks like it confirms my experience. I have had mine since April, but what with the light nights and poor weather still haven't used it that much. It's frustrating I know, they must be better than this, this is why I started this thread to get to the bottom of it.

Like you I am at work at the moment, so will go through your post in more detail later, but I offer the following:

As well as levelling your scope, make sure it is reasonably well balanced and make sure that the friction nuts on the drives are set to the correct tension. The friction nuts are under the black plastic covers. I have been trying to determine the correct tension with some success, see towards the bottom of the following post

http://stargazerslounge.com/discussions-scopes-whole-setups/155804-skyliner-250px-flex-tube-auto-tracking-dobsonian-2.html#post1978676

Go forwards and backwards to see my symptons and testing.

I actually think I have the friction nut too slack, so will try and do some more testing when time permits.

Regarding the advice in the Synscan manual there is a tip under two star alignment which says pick two stars at least 60 degrees apart and it works best if they are at the same Alt (Page 14 of the manual). I have seen others suggest close to the horizon is best.

This appears to suggest to me that the calibration is really setting the AZ absolute position as an average of the two stars and to minimise levelling issues, it's suggesting only moving one axis. Certainly I have never got a good alignment when a high altidue star is choosen.

I am guessing you haven't set the backlash yet? I only did mine last week, not sure how much difference it makes to GOTO accuracy, I might try clearing mine to see if it makes it better or worse for any given alignment.

The manual also states the GOTO accuracy as 5 arc minutes. A 25mm eyepiece should have a field of view of 1.5 degrees, a 10mm eyepiece should be 0.5 degrees (30 arc minutes) so it should centre objects pretty well.

To get round the problems of poor alignment I use PAE and whilst this centres the scope quite well it doesn't do anything for the tracking accuracy, more of a problem for me at the moment.

[DrRobin]

Hi,

Right it was blowing old boots up here tonight, but there was some clear sky so I decided to make some tests using the Synscan handset and aligning to North and level, I.E. No GOTO, just auto-tracking, using a spirit level to get it spot on.

This enabled me to set up the scope inside my summerhouse pointing out of the doors and point it towards a bright star, not sure which one but it didn't really matter.

Okay, aligned the star with a cross hair 25mm eyepiece and then watched it for a few 10s of seconds. It wandered a little bit, then stopped and then after that looked to be fairly stable.

I replaced the eyepiece with a camera and started taking 60 second exposures. The first couple weren't very good, either erratic or long star trails, but after that they had fairly consistent star trails, indicating incorrect movement in both axis.

I then proceeded to tighten both friction nuts, one at a time repeating the 60 second exposure, using about 1/4 of a turn at first and then fine adjustment of 1/8th of a turn.

I found that when the friction nuts are too slack the mount under tracks and gives fairly smooth straight line star trails. When they are too tight the star trails become more jumpy and erratic. This confirms what I thought, too slack and the motor turns without moving the scope correctly, too tight and it seems to jump.

I found that both friction nuts had to be set to within 1/8th of a turn to get the mount to track correctly. I also found that after moving the scope you need to wait for up to 2 minutes for it to settle and start tracking smoothly. I think the mass of the scope is such that it gently oscillates for a minute or so after a large move.

The scope is reasonably well balanced, certainly without the balance right it's even harder to get it right. I balanced it by slackening the Alt friction nut right off and then add weights mounted on a strong magnet. Moving the weights forwards and backwards to get it right, or at least nearly right for most angles.

In summary, not doing a 2 star align and using the auto-track facility of the Synscan seems to track correctly once the scope is balanced and friction nuts set. You can of course use the read out to know where you are pointing, but if it's aligned the Goto should work just as well, if not better, that is the next test.

[Space Cowboy]

Good stuff Robin! I was getting backlash in alt the other night when imaging Jupiter with the auto-track handset. Just wondering if this means the tension nut is too slack or the scope needs a counterbalance?

[DrRobin]

Hi Stuart,

I would say that it definitely needs a counter balance and then it might need the tension nut fiddling with.

I had great problems finding the balance point as the weight isn't distributed around the Alt pivot axis. The two truss tubes at the bottom outweigh the finder and eyepiece and mean that you really need to add weight towards the bottom and back of the tube.

When I slackened the tension nut I needed to use 1 or 2 1kg weights mounted on magnets to balance the tube, quite a lot and more than I had expected.

I know you don't use the Goto anymore, but if you still have it you can use it as an Auto-track, the benefit is the ability to program backlash and have a read out of where you are pointing.

[Space Cowboy]

Thanks Robin. I still intend using the GOTO but do find the auto-track handset easier/quicker to use when planetary imaging. You say the GOTO control will auto-track but isn't that only if it's been star aligned during a previous session in the same location?

Yep i will try counterweights again when i tried one in the past it was only the same weight as the cam and barlow I was using. Clearly from what you are saying there needs some extra "beef". I reckon this backlash in alt has been my main problem all along with the scope being nose heavy and why its only noticeable when following an object that is rising in the east. The scope wants to nose dive and the planet is going in the opposite direction where as in the West the Planet is falling so the nose heavy scope appears to be tracking perfectly (if that makes any sense!). I've never had any problem with the AZ tracking its always pretty much on the money.

[DrRobin]

Hi Stuart,

Yes the Synscan can be used as an auto-track without having to go through the 2 star alignment. Just align the scope to North and level the tube, then switch the power on. Go through the set up menu until you get to 'alignment' and then press 2 to say no to alignment.

After that you get most of the functions, but not GOTO apparently. To make the scope track select the tracking mode by pressing the SETUP button and scroll until you find the tracking option.

See page 11 of the current Dob Synscan manual (http://www.skywatcher.com/swtsupport/Instruction_Manuals/Dob%20SynScan-180610V6-3.08_lowrez.pdf)

[Space Cowboy]

Cheers Robin! That info was not in the GOTO manual i got with the skymax 127. I will definitely give that a try.

[Space Cowboy]

Went out last night and setup the GOTO handset for auto-track which worked fine (apart from the alt backlash/drift). Added a couple of welders clamp magnets weighing in total just under 1kg to top of scope about 3 inches from primary. Still drifting so added a 1lb weight to magnet and this seemed to do the trick (was using a 3x barlow with zoom eyepiece set at 7mm so about 650x mag) cloud covered Jupiter for 90% of the time and sometimes for several minutes but each time Jupiter appeared it was still bang in the eyepiece (this was about 11.30pm).

Decided to try and take some images with DFK and 5x powermate (this is approx 1500x mag) sadly clouds allowed me only 2 mins clear sky and during that time Jupiter again drifted south so I'm hoping only because the DFK added extra nose weight and its simply a case of adding more beef to the counterweight.

Very frustrating the limited time we have to experiment because of the blxxdy cloud!

[DrRobin]

Hi Stuart,

Great that you are getting there with the balance and the clutch setting.

I have been looking back through my long exposures (with star trails) and the one thing I have noticed is that in a lot of them the galaxy behind hasn't moved very much.

In fact if I compare pictures taken over 30 minutes the galaxy is almost central in the first and last, but many of the 120 - 180 frames show definite star trails. This therefore says to me that over a reasonable period of time the scope is tracking but it is moving in jumps (or jerks) rather than smooth movement. It's mostly in Alt but to a small extent in AZ.

Following my session of testing I have now decided on the following:

If the clutch/friction nut is too slack and the scope isn't absolutely in balance, the motor probably fails to move the scope, or it moves and then falls back. Once the motor has turned the shaft far enough for the low resolution shaft encoder to come out of sync (109 arc seconds), I think the motor (and hence fine encoder) spins faster to bring the scope back into position. This probably explains the jumps or jerks in the movement and hence the star trails.

If the clutch/friction nut is too tight then there is probably to much friction for the motor to move or the slip plate doesn't move freely enough and hence we get jerky movement again.

Examining the mechanism, we have a shaft connected to the scope axis (Alt and AZ are the same), this connects to the motor through a gear. The gear and shaft are connected by a clutch and the same bolt which tightens the clutch also increases the friction against the side of the mount, hence the precise setting required.

Skywatcher has come to this arrangement so that the scope can either be driven or moved by hand and you then need a clutch to stop damage to the motor. I am actually thinking we would better with the drive gear and scope shaft connected without a clutch, then the only thing the bolt would do is set the friction.

If you scope was really well balanced then you could have the nut quite slack and get very free movement. The level of friction could be adjusted to suit without the clutch slipping when the motors are trying to turn the shaft.

[DrRobin]

Hi Stuart,

I finally managed to get out and test my scope last night on Jupiter, now that both clutch nuts are set and the scope is balanced.

I must say, quite a difference. I used a simple polar alignment (no GOTO) and slewed the scope on to Jupiter. I then used a variety of lenses and a modified MS Lifecam to take videos of Jupiter.

Typically I took 1000 frames at 10 frames per second (100 seconds) using just the camera and the camera plus a TV 3x barlow and a TV 5x Power mate. The red spot and other detail was clearly visible on the screen whilst shooting and even zooming in (Sharpcap) Jupiter stayed in the centre of the frame for enough time to see detail.

I had to make minor manual adjustments after about 40 seconds with the 5x and generally adjustments between shots, but then I was adjusting focus and camera settings.

I would say I am just about spot on now, at least for plantary imaging after 2 hours last night I was too cold to try the Atik on the DSO's even though it was still clear.

Now all I need to do is work out how to use Registax 6 and I will post the best picture of Jupiter! I am going to have to look for a good tutorial as I don't think I am getting the best out of it.

[Space Cowboy]

Thats good to hear Robin! Try these tutorials :

http://www.sunspot51.com/Registax6/Registax6.htm

http://stargazerslounge.com/primers-tutorials/110593-registax5-1-dbk21-colour-cam-tutorial.html

[DrRobin]

Hi Stuart,

Thanks for the links. I just don't seem to be able to get Reg 6 to work, something goes horribly wrong during stacking. After spending about an hour, I loaded Reg 5 and did a quick process of one of the Lifecam images with 5x Power Mate. I haven't done any other processing so I should be able to get a better image out of all of the AVIs I took.

The image is actual size from the Lifecam, no scaling being done so you can see the tracking is good.

[Space Cowboy]

Good effort Robin! Just needs auto RGB balance selecting on the wavelet page of registax. Can your lifecam do 30fps as that would be better especially when seeing is unstable as it was for me last night. I would go for 3 min avis or even 4 mins.

Very pleased with how the 2kg+ counter weight has helped my tracking. Last night it was so good I actually went in the house while the dfk cam was shooting and Jup was still dead center when I got back.

[DrRobin]

Hi Stuart,

I tried the auto RGB, sometimes it seems to work and other times it just seems to mess it up completely. The MS Lifecam is capable of 30 frames per second, but only in 640x480. In 1200x800 it can only manage 10 fps.

You can use a window feature (zoom) to output 640x480 over a smaller area of the chip and it will then give you 30fps, but I could only use that with the 3x Barlow otherwise Jupiter would more than fill the frame.

It seems very sensitive, I have turn down the exposure and brightness quite a long way otherwise it's over exposed. I think I had the saturation up a bit high for most of the AVIs, most are a little on the red side.

Glad to hear the weights are working for you. I use 2kgs with my Atik, but the Lifecam and SPC900 only needs 1kg. I reckon my scope will now track Jupiter on a 8mm eyepiece for several minutes, perhaps 10-15 and certainly tracks for around 2-3 minutes with the webcam.

[DrRobin]

Hi,

The attached image is with a Lifecam, 10" flex tube, TV 3x barlow and 1800 frames (1700 stacked) taken in 1280x800 at 10 fps on the same night as the other image about 30 minutes before.

It took about 10 times as long to stack as it did to take. The small dot to the left is one of Jupiter's moons.

[DrRobin]

Hi Stuart,

Just wanted to ask you what you thought of your DFK? I am going to see if I can find a DFK or DBK on the second hand market, but might consider splashing out and get a new one.

I know the 21's do 60 frames per second, and the 31 and 41 are higher resolution and hence slower frame rate. Do you know if these support zoom/window?

E.g. A 31 run at 640x480 outputs at higher than 30 fps?

TIA

[mindburner]

hi DrRobin very interesting post. Upon taking the side cover off the motors/ecoders, I could see the motor shaft encoder moving when I moved the scope by hand. I don't think this is right? Oy should be still unless moved by the hand controller

I loosened the nut off a bit and the encode stopped but the tube OTA would not move up and down properly. Seems very fine adjustment is needed. Just wondering if you have any more tips on setting the tension of this clutch nut

picture below..

[DrRobin]

Hi Mindburner,

Please excuse the first couple of line in case I am teaching you to suck eggs...

There are two encoders, one is attached to the end of the motor (shown in your picture with the black arrow) and the other is attached to the scope Alt shaft. (Actually the AZ is exactly the same).

The motor encoder is about 120 times more accurate than the encoder on the scope shaft and it is the motor encoder which is used for fine tracking. The encoder on the scope shaft is used if you move the scope by hand.

The motor encoder is fixed to the motor shaft and the other encoder is fixed to the scope Alt shaft. The clutch connects the motor shaft (and gearbox) to the scope Alt shaft. The clutch nut also tights the scope Alt shaft against the fixed side of the mount, so it has a dual purpose.

If the clutch nut is too slack, the motor will move but if the scope isn't perfectly balanced the scope doesn't move or lags behind. What seems to happen is the faster the slew rate the more the scope keeps up, but when tracking the slew rate is small and the scope just lags behind all of the time.

If the clutch nut is way too slack and the scope isn't perfectly balanced, it drops at low altitudes.

If the clutch nut is too tight, there is too much friction between the Alt shaft and the fixed side of the mount for the motor to either move or move smoothly and again the scope lags behind or has jerky movement.

To get the clutch nut right, I first balanced my scope as best as I could (I need between 1kg and 2kg for the 10") and then during daylight with tracking turned off I centered a distannt object, slewed the scope away in Alt and then slewed back using the position on the handset to bring it back to datum. I started with a fairly slack clutch nut and of course there was lag, I.E the scope didn't return to the start point, even though the handset had the same coordinates.

I then tightened the clutch nut until it did return for all slew speeds. I marked the clutch nut setting. I then continued to tighten the clutch nut until the scope didn't return to the same place, again marking the spot. I then set the clutch nut mid way.

I repeated the exercise for AZ as both suffer the same problem. It is important to use a fixed object with tracking turned off for this test and since you do it in the day, it doesn't require clear skies or wastes observing time.

I actually found the he clutch nut could only be turned about of 1/4 to an 1/8 of a turn before the scope didn't track correctly, quite a small difference and I suspect that it will drift over a period of time so the exercise might have to be repeated from time to time.

I hope this is useful to you, let me know how you get on.

[mindburner]

many thanks for such a great reply. It was very helpful. I had a spare hour or two last night. Where I am there are really no decent terresterial targets so I used my Hotech collimator secuerly taped onto the telescope tube to point at a small precise target mark I made on the wall, got this inspiration from watching a program on spitfires lining up their guns-lol

Seemed to work quite well. I centered the laser on the mark then noted the postion on the handset. Slewed the scope on speed 9 then returned. Im managed to get it to return on most occasions. Tried it on the lower speeds too, again mostly successful. Sometimes it was out quite a bit though . I think getting the scope balanced better would make a difference.

I found that putting weights on the front section of the flextube helped as the 12 inch seems a bit backheavy. The weight I used was only about .5kg. I really need to get some decent scope weights.

The scope does seem to 'buck' when slewing still. But I can't really sem to get the alligment better than 10 arc minutes. The amount of play in the gears is shocking. And the backlash. I tried entering values in the GOTO handset but little difference. Also tightening up the mount screws on the actual OTA. The whole mount is a bit ropey though. I love the optics on this beast but I would love to replace the mount, especially to an EQ but the sheer size probably will not allow.

Hopefully tonight will be clear and I can try a 2 star alligment.

I appreciate your input and previous efforts with this, as these dobs are so good and if the mount could be improved, they would be so much better again.

[DrRobin]

Hi Mindburner,

Good idea to use the laser. Slewed my scope away on rate 9 and then back on a lower setting, finishing off with rate 1 as I found that the clutch slips more at lower slew rates.

I agree with you that a great scope is a little let down by a poor mount.

Interesting that the 12" is back heavy, I hadn't really though about it before, but I guess the focuser and finder are the same on the 10" as the 12 so it probably makes sense that the 12" is better balanced.

I actually think the real problem lies in the clutch friction material and the bearing. As far as I know the bearing is a Teflon sheet between the moving part and the clutch. The clutch must be some sort of friction plate.

As far as I am concerned, the clutch should offer high friction and the bearing plate low friction smooth operation for clutch nut settings. I am going to strip mine apart sometime to find out what is in there, but if the clutch connected the scope to the drive train with lower tension on the friction nut and if the bearing plate moved more smoothly it would give much better operation and tracking.

Regarding the backlash, I centred my fixed object and slewed in one direction to it. I then slewed the other way on a slow slew rate and noted the point the scope moved on the handset. That was my starting point for the backlash. I then fine adjusted the backlash by centred the alignment mark and press the slow slew rate until the scope moves straight away.

When set correctly, if the scope is centred using the up, pressing the down the motor spins quickly to take up the backlash and then starts to move at the slew rate set. Getting it right makes tracking much easier.

There are separate backlash settings for each axis, just repeat the exercise in both axis until it's right.

[mindburner]

hi I made up a basic counter weight with a magnet ans some pliers. Not very permanent but acted as a good guide. I discovered that the teeth were jumping on the ALT gears, when the tube was hand moved, which worried me a bit in case of stripping. I guess the clutch was too tight.

I moved the motor mount up slightly but releasing one of the black hex screws until the motor drive was closer to the axis & gear. Then loosened up the clutch a bit. I noticed that generally when the scope would not return to the park position I had set, it was off by one gear tooth or the play in the ALT. If these gears were of a higher resolution, the backlash would be lower perhaps.

I will have a go at the backlash maybe tomorrow or Tuesday and post back. Thanks again for the detailed help.

There is definately a mod that could be done on the clutch, maybe better materials of washers.