285 replies to this topic

[photo444]

I had an ET-8 equatorial mount with a 6" F5 Newtonian reflector more than 10 years.  Using Arduino and stepper motors, I have been modifying the mount up until I found OnStep (http://www.stellarjo...uipment_onstep) on the internet.  Immediately jumping into OnStep I was able to come up with an WORKING GOTO system. I mean really working GOTO equatorial mount even with an android controller.  I would like to share my journey with the Stellar Journey (The Website of Howard Dutton) here.  I believe there are many owners of ET-8 mount who can be benefited by this thread.  The main idea is not limited to ET-8.  Anyone who has manual equatorial mount can easily modify to a fine functioning GOTO mount.

 

As you see in the photos, I have been playing with many motor-gear combinations evidenced by several holes on the mounting plates. I have been keeping the mounting scheme as simple and flexible as possible for trying many different combinations.  Soon the perforated mounting plates will be replaced with dedicated good looking (?) ones.

 

Mechanical information of the ET-8 mount is as follows: RA worm 144:1, DEC worm 90:1, no ball bearings, cast aluminum body with steel shafts, and so on.  I have added 60:15 pullies and belt (GT2) for RA, and 90:15 for DEC.  Those ratios are working fine with stepping motors (200 steps/rev, http://www.omc-stepp...1206s-p-20.html). Currently this combination provides me  ~2 deg/sec of slewing rate which is satisfactory to me.

 

The electronics is very simple: A Teensy 3.1 controller (or Arduino MEGA 2560 R3) and two DRV8825 stepstick drivers. An HC-05 bluetooth interface was used to control the mount with my android smartphone.  All the works were done by Mr. Howard Dutten, and I am deeply indebted by him to have this complete GOTO mount of my own.

 

I may note that the total cost is incredibly cheap if you already have an equatorial mount.

 

Paul

Attached Thumbnails

• 
• 
• 
• 

Edited by photo444, 16 December 2014 - 08:36 PM.

[photo444]

Before bringing the clumsy set up out side into dark and cold night, it seems better to neat up a little bit.  I have 2 x 2 7/8 perf-boards from ebay looking good in size for the purpose.

 

 

[photo444]

The board is finished last night rather this morning. All of ugliness is transferred to the backside. I have plugged and turned on, amazingly it worked at the first time, very unusual.  Now I can go out with peace in my mind after providing a proper box.

Attached Thumbnails

• 
• 

[photo444]

I have measured the slew rate of this set up to be 2.0 deg/sec which gives solid slewing with the load shown in the picture in the previous posts.  The motors got barely warmed. I believe the rate will be higher with 24 volts.  2.0 deg/sec both RA and DEC is equivalent to 480 times the sidereal rate.  Still waiting for clear skies.

Edited by photo444, 19 December 2014 - 11:05 AM.

[Falcon-]

Very cool looking project.  I will be interested to see how it responds to guiding input!

 

Makes me ponder this type of system for the Super Polaris I picked up cheap off craigslist locally recently.

[photo444]

I am looking for some clear night to try out guiding and the result will be posted here. The Super Polaris I, IMHO, is better than I am using.

[Falcon-]

I am looking for some clear night to try out guiding and the result will be posted here. 

 

Hope your weather is better then I have been getting!

 

 

The Super Polaris I, IMHO, is better than I am using.

We shall see.  Normally I am sure you are right, but this one was (as I was told by the person I purchased it from) rolled down a flight of stairs by a 4 year old.  Took enough impact to rip the RA axis lock out of it's threads so there is probably a non-zero chance of the worm or wheel gears being damaged.  (I should be be able to repair the RA lock without too much work)

 

Of course I only paid $30 for it and I expected to use it manual with no drives at all, so if I am disappointed by it for AP performance I will not weep.  Would be nice to have go-to regardless though!

 

I actually have a non-goto CI-700 that I do use for astrophotography that I am *hoping* to eventually turn into goto using an EQ6 synscan kit.  If I had not already purchased the synscan drives/controller I would be jumping on this as a solution.  At the very least it may be a good backup if my synscan conversion fails.

[photo444]

If you can fix the RA lock, it would be a steal at $30. I hope the fixing will work out.

Knowing that the gentleman who wrote the software has designed it for his Losmandy G11, it will be a fun project to turn any manual mount to GOTO.  

 

BTW, you have an incredible website: I really enjoyed visiting there.

[ccs_hello]

Paul,

 

That's very impressive.

Would you mind commenting the tracking speed adjustment range and how the rate was derived/frequency generated?

 

Clear Skies!

 

ccs_hello

[photo444]

Thank you for your interest. The software uses two independent timers for generating necessary interrupts for RA and DEC rates so the rate generation is flexible and covering wider ranges.  You may look at the source code which is posted in the GitHub.  I think the author, Howard, of the Stellar Journey should answer your question than me.

 

Paul

Edited by photo444, 20 December 2014 - 08:01 PM.

[photo444]

ccs_hello,

 

Here are the urls you may be interested in visiting. The program name is OnStep, btw.

 

http://www.stellarjo...quipment_onstep

 

And the software is posted at

 

https://github.com/hjd1964/OnStep

 

Paul

[hjd1964]

Paul,

 

That's very impressive.

Would you mind commenting the tracking speed adjustment range and how the rate was derived/frequency generated?

 

Clear Skies!

 

ccs_hello

OnStep (the latest Dev branch version) has a minimmum recommended tracking rate of 32 steps/sec (it can go lower,) and the maximum tracking rate is 100 steps/sec.  During sidereal tracking the motor step pulse jitter measures about a millisecond.  During gotos the jitter drops to a few microseconds.  The fastest the motor timers can run at is 32us/step, that's enough range to do nice fine steps during sidereal tracking and still get to a couple of deg/sec for gotos.  OnStep can switch the stepper driver's microstepping mode during gotos too now, which expands the range even further.  Do keep in mind that this project is still in beta and some areas need more testing and probably some fixes, especially where I've recently added features.

Edited by hjd1964, 21 December 2014 - 04:52 AM.

[photo444]

Howard,

 

Thanks for the explanation.

 

Paul

[photo444]

While having disappointing weather here I have been playing with the mount inside.  I tried to mount a 6"F5 OTA (11 lbs) plus an extra weight (5 + 5 lbs) for balance, total of 27 lbs.  When performed slewing from the pole to M42 and back error of more than 2 degrees was observed.  As I repeated similar swings, it was found that 27 lbs is too heavy to handle with this mount.  I am going to limit the load to ~10 lbs and a 5 lb weight, a total of 15 lbs.  Slewing test with this load seem OK so far.  My initial try will be a 400 mm lens with a tracking camera and a LN300 camera with 75mm f.l. /F1.9 Apollo Television Lens for wider field video/pictures.  I might add 200 mm lens with a mirror-less camera for additional imaging.

[photo444]

After nasty sky condition in NJ the sky finally was clear last two nights but I had very little time to play the new toy.

With expectation and excitement I have brought out the rig outside in the middle of Christmas lights, did rough polar alignment by eyeballing, the first target for 1 star alignment was the Betelgeuse in Orion.  When the scope was slewed, the star was not there. After struggling with this and that finally went to the M42, but it was behind tall branches. Anyway I opened the MetaGuide, selected a guiding star, then turn on the camera, and found the video finder and the camera were not aligned as shown in the pictures.  When I tried to calibrate, the MetaGuide crashed  and the computer was in a mess.  I have to go back to reading instructions and practice.  During all these the Android App OnStep and the Star Planetarium were running fine.  The last picture was unguided of course; You can see driving error from the trails of the stars (400mm F5.6 ISO3200 30sec).  When is next clear skies?

 

My intentions was Sky_Planetarium -> POTH -> OnStep_Driver, and METAGUIDE -> POTH -> OnStep.

 

Today I checked guiding with above set up and it seems working.  Please help me with your suggestions.

Attached Thumbnails

• 
• 
• 

Edited by photo444, 27 December 2014 - 02:16 PM.

[hjd1964]

Not sure about the larger 'scope and weight vs. what the mount is capable of, make sure you're well balanced.  I know you mentioned the clutches slipping before.

 

Polar alignment is very important, tracking and goto performance depend on it.  The mount's lack of a polar alignment 'scope makes this task more difficult, I can see that the star trails aren't lined up with RA so I strongly suspect that you were not polar aligned very well.  Try drift aligning the mount before imaging, then start a one (or three)-star align with OnStep again when done.

 

I experimented with Metaguide only once, a few weeks back, but it worked fine.  My small guide scope wasn't really good enough to get the frame-rates up high enough to make it an improvement over PHD2 though, so I didn't bother really using it much.  It was calibrated and guiding here though, no crashes (Lumenera camera.)

[pbunn]

I just bought a Arduino computer board and some stepper drivers a few days ago after you started this article. I have a very heavy duty mount that was built by a very talented high school student a few years ago. The mount was made for DS motors and controller, but these were underpowered in my opinion. He made about five mounts. They were called Gemini mounts. I hope to bring life to this beast with this project.

 

I also bought a Byers worm and clutch assembly on EBAY a month or so ago that needs some electronics. This would work great for running the Byers worm with a new stepper.

 

Pat

Edited by pbunn, 28 December 2014 - 10:04 AM.

[hjd1964]

I just bought a Arduino computer board and some stepper drivers a few days ago after you started this article. I have a very heavy duty mount that was built by a very talented high school student a few years ago. The mount was made for DS motors and controller, but these were underpowered in my opinion. He made about five mounts. They were called Gemini mounts. I hope to bring life to this beast with this project.

 

I also bought a Byers worm and clutch assembly on EBAY a month or so ago that needs some electronics. This would work great for running the Byers worm with a new stepper.

 

Pat

 

Sounds like fun, I'll be around if you have questions.

[pbunn]

Thanks. I was hoping that you might offer some assistance. I have some experience with electronics and some mechanical experience, but not really any with an Arduino or Teensy board. Have done quite a few PIC processor projects using assembly.

 

Pat

[photo444]

Not sure about the larger 'scope and weight vs. what the mount is capable of, make sure you're well balanced.  I know you mentioned the clutches slipping before.

 

Polar alignment is very important, tracking and goto performance depend on it.  The mount's lack of a polar alignment 'scope makes this task more difficult, I can see that the star trails aren't lined up with RA so I strongly suspect that you were not polar aligned very well.  Try drift aligning the mount before imaging, then start a one (or three)-star align with OnStep again when done.

I just want to find out what this cheapo mount can do, and I am kind of challenged by it.

 

It is very hard to do polar alignment without a scope, I am always eyeballing on the left side then right side and back and forth.  Unfortunately I have to remove the setup every time so I cannot keep the previous alignment.  This is why the accuracy of GOTO is not good enough.  The picture on the post #15 was unguided 30sec exposure.  I have noticed there are some jerky motion now and then which I hope guide out someday.  I am now totally counting on drift aligning. First rough align at the north pole, goto a star, then adjust the mount to make the star centered, goto another star and repeat. Then start drift align before calibration.  Hope I am doing right.

I experimented with Metaguide only once, a few weeks back, but it worked fine.  My small guide scope wasn't really good enough to get the frame-rates up high enough to make it an improvement over PHD2 though, so I didn't bother really using it much.  It was calibrated and guiding here though, no crashes (Lumenera camera.)

Metaguide seems working fine.  I am on a steep learning curve, I believe.  Trial is the best to learn. Last night I had a short session before the sky was covered by clouds. But I was able to calibrate and guide a short while.  I will post the result soon.  The crash I had the other night was due to my mistake.  I didn't run MetaGuide as admin.  Crash happened when I was making connection to scope in the MetaGuide. Last night I ran everything as admin. and it was working.

 

Thank you for your comments.

[photo444]

I just bought a Arduino computer board and some stepper drivers a few days ago after you started this article. I have a very heavy duty mount that was built by a very talented high school student a few years ago. The mount was made for DS motors and controller, but these were underpowered in my opinion. He made about five mounts. They were called Gemini mounts. I hope to bring life to this beast with this project.

 

I also bought a Byers worm and clutch assembly on EBAY a month or so ago that needs some electronics. This would work great for running the Byers worm with a new stepper.

 

Pat

 Playing with Arduino is easy and fun.  The worst part was don by Howard (hjd1964). Could you show your mount, I am interested in seeing it?  

Edited by photo444, 28 December 2014 - 07:29 PM.

[photo444]

I had a short session last night before it became cloudy.  I am getting a lot better than the first time in using MetaGuide but still need a lot more. Opening software as administrator made the run smoother. At first it was calibrated with Guide Rate of 0.5 of sidereal rate: Calibration took so long time so I increased it to 1.5 times then 'quick calibration' was done quickly. 

 

The first picture is the snap shot of MetaGuide when guide rate was increased to 1.5 times. The parameters I set was not right yet and I suspect the wrong data input might have caused poor performance.  I was taking a series of pictures with 400 mm lens and NX300m at ISO1600 and 30sec exposure to see how guiding/tracking is working. The 2nd picture was obtained with guiding rate of 1.5X. The third picture is the worst when the camera was guided with guiding rate of 0.5X.  It shows a terrible jitters in both RA and DEC, and it comes in every 10 minutes which coincides with the period of the worm gear (24 x 60 / 144 = 10 mins.).  I wonder it can be corrected by PEC, or I have to avoid the region for picture taking.  The inset shoes how terrible the jitters are.  I didn't get enough number of subs for stacking yet.

 

I love LN300.  At this time it is used as a guide camera with a 75 mm (focal length) lens. I will sens-up to 1024 (for guiding color is not necessary) and Elec lens of x4 so the guiding software can react much quicker less than a second to smooth out fast jitters of the mount.  My challenge now is to study and learn more how to use MetaGuide, and optimize the parameters for guiding, perhaps with a longer focal length guide scope.  Polar alignment surely is one of my home work to solve.  

Attached Thumbnails

• 
• 
• 

Edited by photo444, 28 December 2014 - 08:53 PM.

[hjd1964]

400mm FL is pushing it for a mount of this quality... but in the interest of seeing what's possible:

 

Picture #2 seems to show fairly good RA tracking, but Dec was drifting a little.

 

I suspect that the 1.5x rate isn't doing what you're thinking... OnStep's ASCOM driver doesn't implement RA/Dec rates so that parameter in MetaGuide doesn't do anything.

 

On the other hand OnStep defaults to 1x guiding unless set to something else (in Sky Planetarium's "Telescope Pad" for example.)  And, I wouldn't go above 1x for guiding either (Slow in the "Telescope Pad"), faster rates are for centering objects.  Setting the aggressiveness in MegaGuide might be in order though.  Since the RA guiding doesn't ever reverse direction, backlash isn't a problem there.  Declination on the other hand does change direction and the gear teeth will float in a gap during this.  OnStep has backlash compensation values to move quickly through this gap, but there's normally an unpredictable element involved in the mount mechanics which makes this imprecise.  Some folks guide Declination in just one direction or turn backlash compensation off entirely to prevent erratic movements when the direction is reversed.

 

As for PEC fixing the 10 minute problem, it's possible that it could help but I'd try to get the Autoguiding working first before adding another element to the puzzle.  And I should note here that this Development branch of OnStep has changes to PEC (for Index support) and I haven't had time as of yet to put it through it's paces.

 

Overall this still looks like improvement to me and I'd bet better results will be achieved with more tuning and better polar alignment.

[photo444]

400mm FL is pushing it for a mount of this quality... but in the interest of seeing what's possible:

I think so, too.  I may stay with shorter than 200mm, say 150mm.

 

Picture #2 seems to show fairly good RA tracking, but Dec was drifting a little.

I don't think polar align was good at all. 

 

I suspect that the 1.5x rate isn't doing what you're thinking... OnStep's ASCOM driver doesn't implement RA/Dec rates so that parameter in MetaGuide doesn't do anything.

When the guide rate was 0.5x, stars were sliding toward EAST as if tracking was slow. I first increased RA/DEC aggresiveness from 0.7 to 1.0, nothing happened, then further incresed to 1.5 and still nothing happened.  So I changed guide rate from 0.5x to 1.0x, some improvement was there, then to 1.5x, and it appeared better.  Of course I didn't know what was going on with the guide rate but in my simple mind I was satisfied. 

 

One more thing to report to you: First I went to a star in Gemini and found the scope is roughly tracking without guiding. Then I went to M42. Now stars on the MetaGuide screen were slipping to EAST as if it was not tracking. So I checked the turning of the RA pulley; sure enough it was turning to correct direction. At this time I calibrated and guided as I described above.  

 

All of these observation can mislead because the polar alignment was poorly done.

 

On the other hand OnStep defaults to 1x guiding unless set to something else (in Sky Planetarium's "Telescope Pad" for example.)  And, I wouldn't go above 1x for guiding either (Slow in the "Telescope Pad"), faster rates are for centering objects.  Setting the aggressiveness in MegaGuide might be in order though.  Since the RA guiding doesn't ever reverse direction, backlash isn't a problem there.  Declination on the other hand does change direction and the gear teeth will float in a gap during this.  OnStep has backlash compensation values to move quickly through this gap, but there's normally an unpredictable element involved in the mount mechanics which makes this imprecise.  Some folks guide Declination in just one direction or turn backlash compensation off entirely to prevent erratic movements when the direction is reversed.

 

As for PEC fixing the 10 minute problem, it's possible that it could help but I'd try to get the Autoguiding working first before adding another element to the puzzle.  And I should note here that this Development branch of OnStep has changes to PEC (for Index support) and I haven't had time as of yet to put it through it's paces.

As you have commented, I should stay with autoguiding and master it before going further.  PEC seems too complicated  for me for now. 

 

Overall this still looks like improvement to me and I'd bet better results will be achieved with more tuning and better polar alignment.

Thank you for elaborated comments.

 

Paul

[photo444]

I just did angular analysis on the size of elongation to measure angular movement in RA and DEC direction during the 30 sec of exposure.

The best sub frame shows RA~30" and DEC~30" and the worst sub shows RD~30" and DC~60".  These are very crude estimate of angular movements during the worm period.  I will measure again when the autoguiding improves.