EQ8R poor guiding and periodic error

[matija]

Hello!

 

For the last few months I've been having a lot of trouble with my EQ8R-Pro. The thing is moving up and down when guiding by 2.5"-3", making it useless for astrophotography. I already sent it back to get repaired but the problem persists. They don't want to refund the money so all I'm left with is to try and find an answer here.

Like I said, the mount guides terribly(I will attach 3 examples). I tried everything from better PA, better balancing, making sure its level, making sure seeing is good, making sure there is no wind, making sure its on solid ground, making sure there is no gable snag and so on but nothing seems to fix the issue. I am using an ASIAIR but the same thing happens on PHD2. I unfortunately do not have guide logs but my PA is 30".

I tried pretty much all settings, guide rates, aggressions and exposure times. Same thing all the time.

I am using a 12.8V battery to power the mount. SW recommends 13.7V but I'm not sure if that's the problem... or can it be?

I'm guiding with an OAG on a 2000mm 10"RC with 5s exposures to prevent seeing effects on guiding.

On a few of the graphs I drew the worm rotation period(189s)

This error repeats every worm cycle.

 

What else is there to do, check or fix?

I'm lost here. I spent all my savings on this rig that I now cannot use and I feel miserable.

 

Thank you!

 

Matija

[vlaiv]

Skywatcher mounts are hit and miss affair, I'm afraid.

Almost no works very well out of the box for such a demanding application and if you want good performance - you need to do two things:

1. mechanically improve the mount

2. tweak guide settings to best suit your particular mount.

I had to do a lot to my HEQ5 to make it behave in satisfactory way. First order of business is stripping down mount, changing all the bearings for high quality SKF ones and cleaning and re lubing the thing. Next to that, I did belt mod on my mount, changed saddle plate, added Berlebach planet tripod, but you probably don't need to do that as your mount already has belts and decent saddle plate?

Make sure that you also "tune" the mount when assembling it back together. Tune out any backlash / play in gears by adjusting tension where it needs to be adjusted. Maybe lookup EQ8 stripping / tuning videos on youtube. I'm not sure if there are any, but I've seen such tutorials for smaller mounts and they are helpful.

Next is guide settings. I've found that skywatcher mounts need to "accelerate" slowly when doing corrections. They are not very stiff / solid mounts. This means slow correction speed - like x0.25 sidereal. You'll also need to play with aggressiveness settings and other parameters. Min motion parameter (I'm here talking about PHD2 params - not sure how ASIAir handles those things) is also very important - and is often misunderstood one as it is expressed in pixels and not arc seconds. It therefore depends on guide system.

As far as guide exposures are concerned - I'm firm believer in longer guide exposures - but this can be used only if mount is smooth enough, so mechanical tweaking is important.

One thing that you can and should do to tame your periodic error is to do periodic error correction. This will reduce total deviation and will make guide system work less to correct for errors - which is a good thing. Ideally guide system should be minimally employed.

From above guide graphs - I think that you have significant amount of backlash in the mount - and that is something that you need to tune / tighten up while re greasing / changing bearings.

whenever you see pattern like that - it might be sign of backlash issues. In ideal world - one correction is enough to return mount to where it should be. In reality it often takes 2-3 corrections of varying intensity as there is a lot of noise in the data, but as soon as you see that you need a lot of corrections - this means that mount is not responding to correction properly. Either correction is wrong, or mount "responds" to it - without really moving. This is what backlash does. Instead of correction moving the mount it is "spent" clearing the backlash before teeth properly engage again. It can also be sign that there is a lot of lag / flexibility in the setup. To move large mass - you need to accelerate it and it takes time for that to happen. If everything is not rigid and there is flexibility - then this time goes up (think of pulling something with inelastic and elastic rope - with regular rope it will start moving right away as you pull it, but with elastic rope it will take some time for it to start moving).

Hope this helps

[windjammer]

Hi

I use PHD2 so not familiar with ASIair.  But, those aggression values look high compared to what I am used to.  My mount is a diy gem, so it is a not precision bit of kit at all. I guide at 10-20ms exposures and 100ms interval, so a 5sec gap between exposures seems v high - you might try changing that.  All that stuff about guiding to the seeing is not right IMHO - what you have to avoid is resonance between guide impulse and how fast the mount moves, so the delay between guide exposures is the important thing.  The mount has to settle before the next guide exposure and impulse.

Your EQ8 is meant to be a precision bit of kit, 5 seconds accurate tracking should be business as usual, so a lot of what vlaiv says is quite right!  But before you get the spanners out and take it to bits try some more guiding experiments.

Simon

[symmetal]

Your RA looks like mainly a backlash issue. You say you're well balanced but that's not ideal for gear driven mounts. It's normal to make the balance 'East side heavy' so that the side of the mount facing East is trying to swing down under gravity so the RA drive gears are always engaged and the RA drive is always having to push the East side up. In the northern hemisphere anyway. This way, RA backlash shouldn't cause a problem, as the RA is only driving in one direction or stopped, never in reverse.

I have an EQ8R and I balance such that when the bar is horizontal and the clutches off,  it just begins to fall on its own under gravity. If the bar is nearer vertical then this gravity inbalance effect  is reduced so if I have poor guiding when the bar is near vertical I attach a weight pulling horizontally on the bar to simulate the gravity inbalance as shown below. The pulley the cord goes over needs smooth bearings or it can cause sudden jerks leading to large RA spikes every minute or so.

Dec backlash can be more problematic as Dec needs to drive in both directions. Making the scope balance back heavy helps for the same reason as above. Again when the scope is near vertical this has limited effect so a weight added as before trying to pull the scope to one side should help. Dec backlash is then only really noticeable when a dither command has driven the dec across the backlash area and it can then take 30 secs or more for the Dec to settle.

You don't say what altitude your scope is pointing at when you have your issues. If pointing closer to the horizon guiding will invariably be significantly worse than when pointing higher in the sky. When assessing the mount guiding you need to be pointing at around 50 to 60 degrees in altitude, to minimize atmospheric effects spoiling your results.

With PHD2 you can run the guiding assistant for 10 minutes to see what the periodic error is like. As long as there are no sudden large jumps in the curve and it tends to follow a fairly smooth curve then guiding should be able to correct it without problems, even if the total periodic error is around 20" or so.

Your guide star profile doesn't look too good, and is rather bloated. OAGs often have poorly shaped stars but as long as you choose a star which isn't too overexposed they seem to manage OK.

I haven't done any backlash adjustment, or anything else, on my EQ8R and am happy with its performance. 🙂

Alan 

[matija]

20 hours ago, vlaiv said:

Skywatcher mounts are hit and miss affair, I'm afraid.

Almost no works very well out of the box for such a demanding application and if you want good performance - you need to do two things:

1. mechanically improve the mount

2. tweak guide settings to best suit your particular mount.

I had to do a lot to my HEQ5 to make it behave in satisfactory way. First order of business is stripping down mount, changing all the bearings for high quality SKF ones and cleaning and re lubing the thing. Next to that, I did belt mod on my mount, changed saddle plate, added Berlebach planet tripod, but you probably don't need to do that as your mount already has belts and decent saddle plate?

Make sure that you also "tune" the mount when assembling it back together. Tune out any backlash / play in gears by adjusting tension where it needs to be adjusted. Maybe lookup EQ8 stripping / tuning videos on youtube. I'm not sure if there are any, but I've seen such tutorials for smaller mounts and they are helpful.

Next is guide settings. I've found that skywatcher mounts need to "accelerate" slowly when doing corrections. They are not very stiff / solid mounts. This means slow correction speed - like x0.25 sidereal. You'll also need to play with aggressiveness settings and other parameters. Min motion parameter (I'm here talking about PHD2 params - not sure how ASIAir handles those things) is also very important - and is often misunderstood one as it is expressed in pixels and not arc seconds. It therefore depends on guide system.

As far as guide exposures are concerned - I'm firm believer in longer guide exposures - but this can be used only if mount is smooth enough, so mechanical tweaking is important.

One thing that you can and should do to tame your periodic error is to do periodic error correction. This will reduce total deviation and will make guide system work less to correct for errors - which is a good thing. Ideally guide system should be minimally employed.

From above guide graphs - I think that you have significant amount of backlash in the mount - and that is something that you need to tune / tighten up while re greasing / changing bearings.

whenever you see pattern like that - it might be sign of backlash issues. In ideal world - one correction is enough to return mount to where it should be. In reality it often takes 2-3 corrections of varying intensity as there is a lot of noise in the data, but as soon as you see that you need a lot of corrections - this means that mount is not responding to correction properly. Either correction is wrong, or mount "responds" to it - without really moving. This is what backlash does. Instead of correction moving the mount it is "spent" clearing the backlash before teeth properly engage again. It can also be sign that there is a lot of lag / flexibility in the setup. To move large mass - you need to accelerate it and it takes time for that to happen. If everything is not rigid and there is flexibility - then this time goes up (think of pulling something with inelastic and elastic rope - with regular rope it will start moving right away as you pull it, but with elastic rope it will take some time for it to start moving).

Hope this helps

@vlaiv

The mount has been tuned and is in the best possible working order.

I tried all possible settings but it still has +-2.5" movements.

A thing I noticed is that it guides ok (+-1") near NCP and very bad everywhere else. Especially the south.

Here is the guide log: https://drive.google.com/file/d/1y-N9JJ4s8g2SWP23eYG_GLFiTeztDZ4A/view?usp=sharing

I cannot do PEC on asiair.

Thanks!

[vlaiv]

40 minutes ago, matija said:

A thing I noticed is that it guides ok (+-1") near NCP and very bad everywhere else. Especially the south.

That is normal - it is due to projection on guide sensor.

You can image Polaris without tracking as it moves only very little in a small circle over 24h - that is just a handful of pixels for 24h, so rate of movement in pixels per seconds is very small. This means that relative error also gets smaller (how much mount trails or leads actual position).

Most severe guiding requirement is near equator at DEC 0.

How stiff is your scope and connection of scope to the mount?

When you take it by hand and shake it - the scope that is, can you wobble it by hand or is it firmly sitting on the mount?

[vlaiv]

Also, it would not be bad idea to try running PHD2 on some computer just for diagnostic purposes - to see how everything behaves.

Run guiding assistant - it will tell you amount of backlash as well.

[windjammer]

I had a look at your guide log with PHD2 Log Viewer - a handy tool to look at log data if you don't have it (its a free download). 

It shows two guiding periods of 25mins and 19mins at time points 19:51 and 20:16 with sub arcsecond RMS errors.  That is a vey good starting point.  Nothing wrong with with the mount on this data.

For the other periods the log shows dithering and mount settling starting and failing - the mount is bouncing around.  What are you trying to accomplish with dithering ?  Turn it off !  Every time the mount settles you kick it with a dither !

There are a lot of experts on PHD's user forum who might be able to help you with these more sophisticated PHD settings like dither.  My understanding of dither is that it is an effort to reduce the effect of camera defects like fixed pattern noise.  Are you sure need it ?

I agree with Vlaiv 's recommendation on PHD guiding assistant - very helpful.  Also run the PHD star cross test - this will point up RA and DEC backlash issues, if indeed you have them.  A high end mount like EQ8, well tuned as you say, should be OK in that regard.

Let us know how you get on

Simon

 

[matija]

@windjammer My problem is the two up spikes and then a down dip of 2.5" that repeats every 189s or 1 worm rotation.

[windjammer]

OK - here is log and frequency analysis at 19:51

log - looks OK'ish:

 

 

Here is frequency analysis at 19:51

 

 

 

Here is log at 20:40 - note dither commands

 

 

Here is frequency analysis at 20:40

 

 

I agree you have a period at ca 200 seconds, but worse when you dither!  

You have guide exposures of 2 seconds and no time delay between guide exposures.  It would be interesting to see guide exposures shorter (0.5 or 0.1 sec) and guide interval 0.5s (mount completes move before you re-measure).  With a higher time resolution perhaps the excursions would not be so spikey - and thus more controllable.

Simon

 

[vlaiv]

On 17/06/2023 at 15:15, matija said:

I'm guiding with an OAG on a 2000mm 10"RC with 5s exposures to prevent seeing effects on guiding.

Something is off here.

What guide camera are you using?

From this data it looks like you are guiding at about 5.2"/px which would mean that you are using 52 micron pixel size?

For the sake of argument - 3.75um pixel size at 2000mm focal length will give you 0.39"/px, so error of 0.52px translates into ~0.2" RMS - which should be considered superb guiding performance.

Can you confirm that you have FL and pixel size correctly entered into guiding app?

[matija]

@vlaiv I have a 10" RC at f8 so 2000mm and am ASI485MC with 2.9um pixels that ive set to bin2.

 

[vlaiv]

34 minutes ago, matija said:

@vlaiv I have a 10" RC at f8 so 2000mm and am ASI485MC with 2.9um pixels that ive set to bin2.

 

This is from PHD2 log file that you posted on google drive:

If above is correct and you have 240mm entered as focal length (maybe you used 60mm F/4 guide scope with 240mm prior to switching to OAG and forgot to change focal length?) - then your error is exaggerated by x8 and in reality it is much much less.

Try checking what FL you have entered in ASIAir and correct if needed and then see what sort of guide RMS you get (it will still behave the same - but will probably report much smaller numbers - like 0.5" peaks instead of 4" or whatever the value was).

[matija]

4 minutes ago, vlaiv said:

This is from PHD2 log file that you posted on google drive:

If above is correct and you have 240mm entered as focal length (maybe you used 60mm F/4 guide scope with 240mm prior to switching to OAG and forgot to change focal length?) - then your error is exaggerated by x8 and in reality it is much much less.

Try checking what FL you have entered in ASIAir and correct if needed and then see what sort of guide RMS you get (it will still behave the same - but will probably report much smaller numbers - like 0.5" peaks instead of 4" or whatever the value was).

Ah yes, I forgot. this graph is from when I used a guide scope. 240mm is right

 

[vlaiv]

1 minute ago, matija said:

Ah yes, I forgot. this graph is from when I used a guide scope. 240mm is right

 

So mount behaved the same as it is behaving now (more or less)?

[vlaiv]

@matija

I just saw that EQ8R-Pro has PPEC, or permanent error correction. You should be able to train that even with ASIAir - as it is feature of synscan controller.

Can you check that?

[david_taurus83]

For such a demanding application I would suggest going windows based PC over an ASIAir. That way you can get full use of PHD2. I find the ASIAir guide graphs are very optimistic. You say the mount had been tuned. Did you strip it and rebuild yourself? You tuned backlash out yourself?

[dan_adi]

On 17/06/2023 at 16:15, matija said:

Hello!

 

For the last few months I've been having a lot of trouble with my EQ8R-Pro. The thing is moving up and down when guiding by 2.5"-3", making it useless for astrophotography. I already sent it back to get repaired but the problem persists. They don't want to refund the money so all I'm left with is to try and find an answer here.

Like I said, the mount guides terribly(I will attach 3 examples). I tried everything from better PA, better balancing, making sure its level, making sure seeing is good, making sure there is no wind, making sure its on solid ground, making sure there is no gable snag and so on but nothing seems to fix the issue. I am using an ASIAIR but the same thing happens on PHD2. I unfortunately do not have guide logs but my PA is 30".

I tried pretty much all settings, guide rates, aggressions and exposure times. Same thing all the time.

I am using a 12.8V battery to power the mount. SW recommends 13.7V but I'm not sure if that's the problem... or can it be?

I'm guiding with an OAG on a 2000mm 10"RC with 5s exposures to prevent seeing effects on guiding.

On a few of the graphs I drew the worm rotation period(189s)

This error repeats every worm cycle.

 

What else is there to do, check or fix?

I'm lost here. I spent all my savings on this rig that I now cannot use and I feel miserable.

 

Thank you!

 

Matija

Get a refund and buy a Mesu mount.

[matija]

2 hours ago, dan_adi said:

Get a refund and buy a Mesu mount.

@dan_adi Im trying to get one but the seller refuses to accept the mount is faulty. He says its 100% perfect and that the error is coming from outside the mount.

[Stuart1971]

On 18/06/2023 at 16:43, matija said:

@windjammer My problem is the two up spikes and then a down dip of 2.5" that repeats every 189s or 1 worm rotation.

So what happens when you run the periodic error tool and then have that turned on, as you mount Phd graph, looks perfectly normal for one that is not using PPEC….mine was the same before I used….🤔

Edited June 20, 2023 by Stuart1971

[Stuart1971]

I see you use the ASIair and so can’t use PPEC, in which case you are screwed, as it needs the PPEC to be used, there is nothing wrong with your mount it is normal, as you can’t use PPEC, dump the ASIair and get a mini PC with EQMOD one even better GS Server, and all your issues will go away….simples….👍🏻

[matija]

@vlaiv Is there any way to prove that the error is caused by the mount / that its a mechanical problem?

[vlaiv]

3 minutes ago, matija said:

@vlaiv Is there any way to prove that the error is caused by the mount / that its a mechanical problem?

I guess that you would need to make sure that mount is operating out of specs - which might not be the case.

Do you have specification for the mount anywhere - like max P2P error, or maximum RA drift error?

First you need to see what is specified as "normal" for that mount and then to show that one of parameters is outside of those normal ranges. You would need to show that without additional gear on the mount, so there are several ways to do it, but I'm not sure which would be considered "definite" proof by your retailer.

1. Using just guide scope and guide camera with PHD2 to create star motion log without actually guiding. This is usually done as a part of PEC correction file preparation and is a way to characterize periodic error. It will provide you with information on peak to peak periodic error and max RA drift rate among other things

2. Using high precision encoder coupled to RA shaft. This is best way to do such analysis as it will give you very precise information without all the noise from seeing or other sources. Problem is that you need at least ~20 bit absolute rotary encoder (you need to be able to at least measure 1" of deviation if not more precise than that) but ideally 24bit one.

3. Using some other means of measuring how precisely mount works - there are several ways to do it, but none are "out of the box" and will require some "software hacking" (finding what software can be used to get results without writing dedicated software).

3.1 Using laser on a white screen in a distance. With this method you can record with camera and lens position of laser pointer mark on a screen and then with help of some software that does motion tracking extract position and in spreadsheet make a plot of how the mount tracks (you'll need some trigonometry to get actual angle from point position)

3.2 DEC tracking vs RA tracking. This one can be done with single exposure or with a video film. In either case you'll need a way to measure a line in an image or point position in a video. You turn the mount to "its side" so that RA actually tracks vertically, and you record star at meridian and equator (due south at celestial equator).

Star will naturally drift in RA due to earth's rotation and mount will track in DEC with sidereal speed. This will create line at 45 degrees in long exposure (or dot that drifts along 45 degree line in video). If your mount tracks perfectly - the line will be straight, but since there is periodic error - line will be bent in the shape of periodic error. From the shape of line - you can calculate all the needed info on periodic error

3.3 Using two scopes and artificial star indoors. You place artificial star at focus of one of the scopes (will need some adapter to hold it there - it can be DIY / 3d printed - whatever works) - and this will project artificial star image at infinity (effectively create collimated beam from artificial star - like it is placed at infinity) - you aim that beam at guide scope mounted on your mount and set the mount tracking.

Again use software like PHD2 to create log of star position or any other software that is capable of motion tracking and producing coordinates from frames.

This is like method 1 above - but it removes:

- need for clear sky and night to perform the operation - it can be done indoors in darkened room or basement during the day

- removes seeing influence

Downside is that it requires artificial star (which you can DIY from a strand of optical cable and LED - look it up online) and way to mount that artificial star to a telescope. Another thing is that it is limited in how long you can record periodic error for. You need to take focal length of your guide scope and pixel size and calculate "/px - or pixel scale. Then combine that with pixel resolution of your guide sensor.

Say that you have 5"/px guide resolution and 1920px across. At sidereal which is ~15"/s - mount will move 3 pixels per second and since you have 1920px - that will be 1920 / 3 = 640 - or about 10 minutes of periodic error recording. You can gain a bit of time if you orient your guide camera so that star image moves diagonally - that way you'll get longest distance.

That would be about all I can think of. Probably best way to start would be to do point 1.

You have the gear for it - just take mount and guide scope that we mentioned (one with 240mm FL) and laptop with EQMod and PHD2 and produce guide log with guiding output disabled. This will "trace" your periodic error with mount alone - no gear attached so no problems with wind or sag or cable drag.

Once we analyze that - we will see how much out of "specs" the mount really is (and it would be good idea to find those specs for your mount - I failed to find anything after a quick search).

[Stuart1971]

6 hours ago, matija said:

@vlaiv Is there any way to prove that the error is caused by the mount / that its a mechanical problem?

You can not prove or disprove anything until you use your mount correctly, if you try EQMOD and then use PPEC, all your issues will go away, this proves the mount is mechanically fine, and the ASIair driver is very poor for this mount….the seller is correct, the mount is fine and works just like any other EQ8, which I have, so ditch the ASIair…

[licho52]

Yeah I'd say EQ8/RC10 and AsiAir is just a mismatch.