F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 7 months ago

Hi Frank,

I am sorry for not getting you more information sooner. I think my will-power to fight these battles has sagged, but I will do my best to get fired up again. I will redo my test with the new update, and will be sure to record a log.

Thanks for looking at it,

Scott

Read More...

F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

Though I promised to record one, I forgot. Fortunately, this behavior is highly repeatable, so I will redo the measurements this weekend, and record a log.

Thanks for your support,

Scott

Read More...

F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

5) As measurement 4 made me suspicious of my OTA offset, I include a picture of measuring it. It looks close to 14.5 to 15 inches, or 0.38 m, to me. See picture below. However, INDI seems to work better if I use 0.5 m. Then the dome rotates far enough in the E and W positions to not occlude the scope. Unfortunately, I have not been able to find and E-W offset that will center it. It seems to want a very big number, like 1m, that makes no sense, and only works on one side of the dome.

So what am I doing wrong?

As always, my sincere thanks,

Scott



Read More...

F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

4) Next, I repeated this test in the west. Note I changed the meridian setting in ICP to West. In this position, the scope was not occluded, but only just barely. The spreadsheet thought the angle should be -2.6°, INDI called for 341.8°, and I would estimate that it actually needed to be ~339°. Pictures, and spreadsheet below.

I interpret this to suggest that the the length of the OTA offset needs to be bigger, and the pier offset E-W larger as well.

File Attachment:

File Name: DomeSlitWest.xlsx.tar.gz
File Size: 9 KB


Read More...

F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

Continued from above:

3) Next, I decided to test the dome positioning at two positions that I think simplify the issues. I first put the scope on the East side of the mount dead flat, looking due north (see figure and picture). I believe this will remove any issues of offsets in pier position in the N-S direction. Thus, the only measurement problems I can have are E-W offsets, Up-Down offsets, and OTA length offsets.

In this position, the spreadsheet says the dome should be at 0.2°, INDI places the dome should be at 10°, and I estimate the dome actually needs to be at about 16° to be centered. About 1/2 of my scope view is occluded by the dome in this position.

The software settings are shown in the attached image. Sorry for the crazy moire patterns.



Read More...

F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

Hello all,

First, to fcasarra: Thanks for the spreadsheet! That I could do easily, and the results were interesting.

Here are the new tests from last night. The high level summary: The values calculated in the spreadsheet are different than those from INDI (Server, ICP, AT, etc), and appear to be even less correct than INDI.

Here is a detailed description:

1) I polar aligned the scope as an absolute positional reference.

2) I shown a flashlight through the polar scope port of my CGEM DX. It's wide illumination filled the port, and illuminated a point on the dome that is presumably within 1 degree of north. No amount of flashlight tipping would move the spot (it isn't a laser, it is a broad; see the attached picture named DefiningNorth). I defined the positioning of the dome to be at 0 degrees when the spot was centered on the shutter.

Read More...

F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

This looks straightforward! Thanks! I will get going on it this afternoon!

Scott

Read More...

F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

Dear Wildi - I think I will start by getting some numbers from the debug side of INDI - if I am still confused, I will go to the work of installing the code you mentioned!

My sincere thanks,

Scott

Read More...

F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

Thanks Wildi, it is great to have your help!

I looked up RTS2 - I didn't know about it, and it looks pretty neat. I see that it wants Ubuntu to compile. I am running a late model OS X on a Jurassic era Mac Mini (2010). Is that going to be a problem, or do you think it will work out of the box?

Thanks,

Scott

Read More...

F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

Hi Ferran,

Thanks for taking a look. In my previous experiments, all measures were in meters, yet that didn't end up working very well. I am working on redoing the test I described with debugging turned on, so I can send you a more detailed log.

Sincerely,

Scott

Read More...

F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

A few more comments:

1) The dome seems centered on the scope when it is vertical and pointing north. So I don't think I have a rotational error exceeding 1 degree or so.

2) I understand that in theory, this calculation is usually unit-less. I am creating hypotheses for the results I am seeing. Looking at the code for the calculation, it looks reasonable, but I can't be sure without working a detailed example.

Read More...

F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

Hello Jasem and INDI guru's,

Peter Polakovic got a new version of the INDI Server to me tonight, and indeed, the dome now frequently rotates! Solved - thank you!

However, telescope pointing is off everywhere but the South East quadrant. I suspect something may be wrong with the units in the Dome Parameters. My carefully measured values are here:

Inches Meters
Radius 45 1.14
Width 25 0.64
N -2.75 -0.07
E -0.75 -0.02
Up - 3.75 -0.10
OTA 15 0.38

To test the parameters, I set the scope to pointing due north, and rotated it so the GEM weight bar was flat, with the weights in the East, then later in the West. I set the meridian appropriately for each case. With the scope in the East, the scope optical axis was aligned with the east-side dome slot edge, partially occluding it's view (it looked like my 11" scope could move at least 7" to the West to get it closer to centered). With the scope in the West, the scope top was aligned with the west edge, enabling a view, but was still very close to the dome slot (it looked like it should move 3" towards the East).

This suggested to me that OTA offset was too small (even though it was carefully and easily measured from the intersection of RA and DEC to the optical axis), as well as there being an E-W offset, in the W direction, of about 4". I increased the OTA offset to 0.5m, which resulted in the western side scope being nearly centered in the dome slot, and the eastern side scope position being just inside the east edge of the dome. I don't understand how this could be the case though, as the size of the offset is clearly smaller than 0.5m.

Next I tried offsetting the E-W dome parameters 4" to the east (-0.13m total). The dome didn't move at all with this change. I then tried -0.5m. This led to a small positional change, but not enough to center the scope in the slot. Finally, I tried a -1m change. The dome moved noticeably, but still not far enough. This doesn't make much sense, as a 1m move of the pier in a small 8' dome is huge!

Is it possible that the units are not all in meters? For example, these observations make sense if the Radius and width are in meters, but the N, E, and Up are in centimeters. I don't yet understand the OTA offset value, unless it is from the DEC-RA to the far side/top of the scope. 0.38 to 0.5 made a noticeable difference in dome position, in the right direction, and I think adding a little more would help. It is almost like it needs to be 1.5X to 2X the distance between the RA-DEC intersection and the optical axis.

Does any of this sound like a known issue?

As always, big thanks,

Scott

Read More...

F. Scott Anderson replied to the topic 'Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

Thanks Jasem!

Read More...

F. Scott Anderson created a new topic ' Dome Sync Gateway not producing timely GOTO commands' in the forum. 9 months ago

Hi all,

I am using Cloudmakers software, including INDI Server (IS), INDI Control Panel (ICP), and AstroTelescope (AT) with my CGEM DX and home motorized 8' dome. I have been using the Dome Sync Gateway (DSG) in ICP to produce GOTO python commands which drive my dome. The problem is that when DSG is slaved to my dome, only the first (or rarely the first few) dome move commands (all of which are GOTO's) are actually given by the INDI driver.

0) I can reliably get AT to convince ICP to throw a GOTO.py by turning off the DSG, then turning off the INDI link to the telescope, then turning DSG back on, and finally turning the link to the telescope back on. However, this is a very clumsy way to track an object all night!

1) I have tried a range of tolerances from 0.02 degrees to 10 degrees. All result in the same infrequent calling of GOTO's. All other dome measurements are entered in ICP normally.

2) My dome has a tolerance of 0.1 degrees. Meaning, if a goto asks for 165°, the dome might stop anywhere from 164.9° to 165.1°. However, once stopped, it tells INDI via a DSG Status command that it is at 165°, either way. So I don't think it is a tolerance error, or an issue with INDI thinking the dome is still moving (I think - there could still be rounding differences, and possibly output formatting differences that make INDI think the dome is close, but not entirely there).

Does anyone know what might be happening?

Peter Polakovic suggested contacting Jasem - with this post, I am attempting to do just that, as well as crowd-source your wisdom!

Sincerely,

Scott

Read More...

SOLVED!! The real clue is the addresses with "btmm" in them.

It looks like the cause of the problem is a service under OS X called “Back to my Mac”. I use this service, but only intermittently, so I can turn it off. It looks like the INDI networking approach is incompatible with it. It appears to cause two network addresses for each Mac - one “.local” and another that is some apple tracked dyndns type something or other.

Heads up other mac users on this one!

S-

Read More...

OK - more data. The DSL router is a C1000A, which the web indicates has a notorious history with bonjour. Hmm. So I figured out how to set up my DSL modem as a transparent bridge, so that my airport extreme, the only item now on the network, could handle all traffic. I have endeavored to remove extra DNS addresses everywhere I could find them - in the airport setup, and in the network settings panel. And Astrometry still reports two bonjour services. ARRG!

I am still soliciting opinions on just what the heck is happening here.

:-)

S-

Read More...

Login



3rd Party

Choose from the numerous 3rd party INDI drivers to suit your needs!

Got Problem?

Check out the FAQ, the forum, and the bug tracking system to resolve any issues you might have!
You can also subscribe to INDI newsletter and development mailing lists to get the latest updates on INDI!