Big problems with Pointings and/or MLPT

[markusblauensteiner]

Hi to all!

 

First of all: i'm using my DDM60pro since nearly a year now and i'm very satisfied with it. I was able to do unguided imaging with MLPT for exposure times up to 20 minutes, (1000mm FL) but since a few weeks i've big problems with it.

 

I don't know exactly when and how the problems began any more. The result of my troubles is, - i'm getting star trails in RA-direction when using MLPT.

This means:

- i did a pointingfile as it is written in the manual, i did it per hand. I did a MLPT and i was able to do unguided images with 8 minutes exposure time. I haven't tried longer exposures. So, 25x8 minutes where done with perfect round stars!

- the next night, i thought that a bigger pointing would be better and so i made a bigger one with 30 stars per hand, before i deleted the old pointing. I did a MLPT again, on the same target as the night before and - i got those trails. They are 3-5 times the star diameter in 2 minutes exposures!

- i thought "ok, lets delete that pointing and do an autopointing". So i did, all images where solved. As result, i got trails again.

- i delted the autopointing file and loaded the pointing of the first night (which then worked!). I did a MLPT again and got trails again...

 

The next step, i deleted the pointing and just used MLPT - same result!

 

Then, i ran the monut without a pointing, without MLPT, without guiding. After 4 minutes exposure time, stars where nearly round. Two minutes - no problem, round stars.

So, i don't think my telescope is too weak, because if, uncorrected imaging wouldn't be possible? Also. at the first night, pointing + MLPT where working perfect..

 

I had this troubles in the past and i thought my telescope, or rather my mounting plate where too weak, so it was built new and very stiff. The first night after modification was the night when everything worked.

 

I also deinstalled Sequence and deleted all files, same problems after re-installing it.

 

So, i think, by doing a pointing (per hand or auto), setting home-position and so on, anything is saved anywhere and is making troubles when a new pointing or MLPT is done? Curves of MLPT are looking normal, as in times when everything was fine.

 

When using MLPT in Autoslew, the coordinates aren't changing (they do, but only in 1/10 " in RA, up and down), but the stars are drifting through the image.... ;-)

 

Maby somebody can help me,

 

BR, markus

[lukepower]

Hi Markus,

 

maybe the mount thinks he's on the wrong side of the pier?

Well, I occasionally see a similar situation when I do a pointing in the area near the meridian where I could go from either side of the pier. I kind of solved this by doing the pointing away from that area, or forcing it to stay on one side. Maybe this helps...

[MarkS]

Hi Markus,

 

Could the polar alignment have shifted?  What is the PA error?

 

Mark

[markusblauensteiner]

Hi Lukas, Mark!

 

Thanks for your answers!

Wrong side of pier: Autoslew asked "East or west" and i told the right side -> east.

Also, the pointing was done all over the east side of the meridian, because at the west side my scope looks into a roof.

MLPT was done, when the object was far away the meridian.

 

Mark, my PA error is less than 1' in both axes. I don't think, that my mount was misaligned in one day (first night, everything worked, next night not), because it stands in an observatory. Also, if it had shifted, only using MLPT should work?

 

Best regards, Markus

[lukepower]

HiMarkus,

 

just to be sure, does Autoslew shows you the correct SideofPier in the lower left corner?

[MarkS]

Markus, this is very strange indeed.

 

Have you tried returning Autoslew to factory settings, and then setting up all the parameters again from the beginning?

 

Regards,

Mark

[markusblauensteiner]

Hello Mark, Lukas!

 

Lukas, i've checked it this moment remote. AS says, east side of pier, this is right.

 

So, a little summary as i see it:

MLPT + Autoslew, are "measuring" a "movement" (drift,..) that doesn't exist. Accordingly, AS is correcting this "movement", and the result are those trails.

Why does the movement not exist? Because unguided, uncorrected imaging is possible. If a movement, flexure,..exist, it is small enought to allow 4min uncorrected imaging. However, with MLPT the trails are there in between 2 minutes.

This "movement" is also measured, when no pointing file is used.

 

But, the first night, before starting experiments with a new pointing (i shouldn't have done that... ), MLPT + Pointing file worked well.

So, i ask myself, what "thing" was saved and is considered by AS when changing the pointing file? I think this is the key.

 

I will try it with my laptop, only using MLPT. We will see. Next step will be to deinstall AS and start at the beginning.

 

Regards, markus

[GeorgeCarey]

When you made the new pointing file, did you clear the old configuration before you started?

[LucMurphy]

I had a similar problem. In the process of trying solving it, I went throught the same process as you on the software side with no real results. I then did several timeflex measurements with sequence to finally find out that the cables were introducing a drift that was orientation-dependant and that got worst in cold weather. The solution was to do a better cable management. I am not saying your problem is cable management (mainly because you have the pro version) but if you  do a couple of timeflex, you could eliminate any mechanical/setup problem. Very important though, do a couple of timeflex in different orientations, not just pointing to the zenith, e.g. low north, intersection of celestial equator and meridian, etc.

I hope this help

Luc

[markusblauensteiner]

Hi!

 

Thanks again for all your ideas!

George, i cleared the old file before started a new one.

Luc, althought i've the pro-version, the cables of the power supply are hanging down. I've tied them to the base of the mount, so that there is a bow. How did you solve that? I've got many cables there - two focusers, two ccd, filter carussel..

 

Yesterday, i tried several things:

1. Go over Homepoistion, go to an object, synch, MLPT - star trails like before; no pointing file;

2. Seperate the second CCD and filter caroussel from the PC. A friend had problems like me, his PC was to weak on the USB for a second CCD camera. But, no changes, trails like before;

3. Try 300 seconds without any correction. Not perfect round stars, but, stars are elongated in both axes, should mean diagonal; And they are less elongated than using MLPT;

4. Re-Installing Autoslew and MLPT; I also deleted all files; After doing all standard settings like motor parameter, i tried only using MLPT - same result;

5. Then i noticed that the location in Autoslew was wrong; but correcting it didn't imrove the result;

6. Then i improved the polar alignment, now its 30" in both axes. No changes using MLPT;

 

But: before i improved polar alignment, the curves of MLPT looked like that: something around 5" in RA and 1.8' in DEC, a big difference; After polar alignment, i had about 1.2' in RA and 1.5' in DEC, which seems more realistic? The result wasn't better...

 

Then, i also noticed something strange, maby it has to do with my problems: when centering a star with the joystick, there was nothing strange. But, when centering the star with the arrows in autoslew (0.06° per second), the star went left when i clicked left, but went back right, left again, right again, for several times before it stayed on its position. Same in DEC. One can see it at the coorinates in AS.

But, AS says balance is fine...and autoguiding brings RMS-data of 0.0xx in maxim for RA and DEC over hours.

 

BR, Markus

[robertp]

Hi Markus,

 

very strange indeed

When you do an MLPT, how does the graph look like that interpolates the dots that represent the errors in RA and DEC? Are these smooth curves (no s-shaped curves)?

After the MLPT-calculation run, did you press "send to autoslew" and did you check whether in autoslew the LPT-icon is orange?

 

Do you have a gps installed and connected? Autoslew relies on a very accurate timestamp.

 

Best regards,

 

Robert

[markusblauensteiner]

Hi Robert,

 

nice to meet you here!

MLPT "curves" are straight lines in both axes , this is so when using a pointing file and without one.

I pressed that button, yes, and in AS that icon is orange. In Sequence Beta, the data should be sent to AS automatically, but also pressing that button there didn't help.

 

GPS is installed and connected, everythings fine there.

 

It seems MLPT does some correction, but the wrong one. It measures an error that i don't have.

So, if i would do that for an hour or two, the target would be out of the image and AS would still think it is in the middle.

I wanted to try it with my laptop yesterday, but  there was somthing wrong with the installation and at 03:00 AM i was a bit angry...

 

Markus

[Waldemar]

I had some weird behavior, because in AS the park positions were 180° off.

As soon as I told the mount it was in West position, while in reality it was in East, problems were gone.

It seems in your case the mount is getting the opposite signals from what it should get. Maybe the above error?

 

This happened with the beta version.

 

Waldemar

[markusblauensteiner]

Hi Waldemar,

 

i have no problems with park position, AS gets there like i want. But, my behaviour is also in "old" AS and Sequence.

 

How should those yellow curves look like when turning motors on? In DEC they are symmetric, but not in RA. But in Servo Setup - Balance - everything looks fine.

 

Markus

[NorbertG]

Hi Markus,

 

did you measure the settle time in sequence ? Sometimes it takes quite long depending on wind, PID and Balance. When you do the pointing manually you won´t see it, but MLPT will do. If your exposure for MLPT is short you might t take the Image at the wrong position. Did you look at the pointing pics ? If settle time is a problem you might see trails there.   

 

Norbert

[markusblauensteiner]

Hi Norbert!

 

This is another brilliant idea! I did this measurement in the past, but i'm nor sure, if i've changed to Sequence Beta before getting those troubles. So it could be, that this wasn't right.

As i remeber, i got a value of 5 seconds when i measured it.

I never looked at the pointing pics, i must confess....

 

Markus

[markusblauensteiner]

Hi to all,

 

i checked the following the last nights:

- cables: no difference if hanging down or beeing bended;

- slew settle time: measured it with sequence, no difference;

- pointing pics: round stars there;

 

I tried to expose 4 minutes without any correction - stars are getting elongated a bit, but diagonal; 4 minutes with MLPT are worse; 

 

I wanted to try with another PC, but there i wasn't able to connect Sequence to the other programs for some reason.

 

BR, Markus

[si6pack]

Hi Markus,

 

I would like to share my observations with you...

 

I use my DDM85 in a mobile setup on a wheely bar - therefore I always do polar alignment, pointing files and MLPT before I can start my sessions...

I have a similar observation with my setup - polar alignment is around 15" in both axis, pointing file 24 stars (12W, 12E) brings the pointing error down to something like 5".

Yesterday I decided to image M81 using SEQ. The exposure settings in SEQ included MLPT. I was very surprised to see an RA error of 14' while the DEC error was 1.4". However, after interpolation (standard settings) I gave it a try. A 600sec 1x1 light frame resulted in an egg like shape of stars...I stopped SEQ, reloaded SEQ and my object and exposure files, started the sequence run, MLPT was performed, same result...

 

Then I decided to substitute the MLPT-based SEQ exposure settings with non-MLPT-based exposure settings...perfect round stars with a 600sec exposure! I imaged the whole night (nearly) no problems - when M81 was crossing the meridian, it was in the middle of 600sec light frame, this frame got long star trails...after meridian flip, no further issues...

 

This is not the first time observation of this issue, sometimes MLPT works perfectly "together" with the loaded pointing file in AS, sometimes NOT.

I do not have a clue, why? The only thing I can say, if I have achieved a

 

Did you ever try imaging with a pointing file in place but without MLPT?

Tonight I will give it a try the other way round, no pointing file, but MLPT...

 

Sometimes I suspect the "pier side" readout from AS being the issue...however, so far I couldn't nail it down...

 

Best,

Christian

[si6pack]

Hi,

yesterdays experiments were a disaster: after having started over, the mount did not track, despite AS told me it was...switching back and forth between "tracking with siderial speed" and "tracking off" yielded in star trails from "right to left" (siderial) and star trails "from left to right" (tracking off). I could not identify the reason why. After 2hrs of trying (reboot, switch off power completely, cable checking etc.), I removed AS (the beta version) completely, and installed the latest official build (5.1.1.6) - no improvement! Thereafter I completely deinstalled all ASA drivers, SEQ and AS and re-installed all programs again, everything worked fine again!!!

I was really pissed to loose all this valuable imaging time for discovering NO real reason for this wired behaviour...Any thoughts to this finding? Or similar experiences out there?

 

Christian

[si6pack]

I forgot to mention, slewing was no issue at all, moving with a gamepad or via AS was no issue, only tracking...

[Jay]

Haven't (luckily) experienced similar problems but some smaller mysteries anyway...

 

E.g I'm using a GPS module in connection with AS and most of the time it works fine. However, sometime when starting the system there is no connection to the GPS module, what ever I do. Then suddenly the connection can reappear and everything is functioning as usual. Haven't been able to find any reason for that behavior, and I'm doing everything exactly the same every single time when starting the system. The irritating factor is the lack of reliability and no way to fix it.  

[prefetch]

christian, that sounds super frustrating.

 

it sounds like it would benefit everyone if ASA had an option to enable verbose logging of AS and sequence.  software is complicated, and sifting through logs can be difficult, but at least logs could be sent to ASA and their developers should be able to quickly identify what causes these strange anomalies.

[si6pack]

jay, prefetch,

 

thank you! next day everything was up and running without any problem...I will have an eye on this and keep you posted in case of.

must have been something with AS...

clear skies,

christian

[robertp]

Hi,

 

I can't help on Cristians issue, but one question to Markus with regard to his original problem:

Markus, when you experienced the elongated stars with MLPT, did you check the first image in a session or did you also see the problem after several images of a session?

The reason I ask is that in the past, I basically always experienced the problem, that the first and last images of an MLPT run had elongated stars, while the images inbetween were perfect. I did some tests and had an email exchange with Philipp about that issue. My guess was, that maybe the starting and end-points of the fitting curve  had a problem leading to not so perfect tracking. Philipp is quite sure that there is no problem with the calculation of the fitting curve. At that time, he fixed the problem with doing a MLPT across the meridian and that issue is gone, but I'm not sure whether the other issue still exists. It's not too big a problem for me, but I usually loose two images of a session.

 

Could you please check whether your stars get perfect after a while?

 

Best regards,

 

Robert

[si6pack]

Hi Robert,

This a very interesting observation! I always stopped the SEQ-run w MLPT after the first shot with elongated stars! To be honest, I never continued to the second shot. However, I an right now running a sequence file w/o MLPT but a really good aligned mount (12 stars, 6W, 6 E) with a pointing file (24 stars, 12W, 12 E) with an ultimate error of 0.04 (RA) and 0.07 (DEC) arc min, respectively. 900sec on M106 perfect round stars!

I am aware that the polar alignment should use 3 stars on W or E, but the 12 star polar alignment file gave the best results for my pointing only (24 stars, w/o MLPT) approach.

Since my SEQ files will run the whole night, I will try to reproduce your observation tomorrow, after intentionally worsening my polar alignment. I am really exited to see whether or not I can reproduce your observation!

Clear skies,

Christian