Jump to content

Banner.jpg.5ed196c1e70861ebc79109e023c96067.jpg

Automation project


NickK
 Share

Recommended Posts

Balance testing with the two boxes stuck using hot glue into position. The scope is sat on a small wooden square that represents the mount point of the NEQ6. I've also put he focuser out enough for focus as well as put the TV 2x on to give the full weight at the back.

IMG_2587.jpg

I will also put a netgear wifi on there too. If you look by the stepper there's the section of aluminium U cross section sticking out that will take the limit switches.

 

IMG_2588.jpg

I found one issue - the 1mm thick steel with the boxes in place takes a second or two to finish wobbling so additional structural integrity is needed, or put the section with the boxes between the two scope rings and on the axis of the NEQ6.

 

I need to check out the mounting on the NEQ6 itself before it's final places are known.

 

Link to comment
Share on other sites

Full on test... connected with the lens cap on.

* NEQ6 mount
* ATIK 383L+ CCD
* ATIK Titan CCD|
* YARF focuser (moonlite driver)

Took a capture and auto-stretched each:

Screen Sharing Picture 15 May 2016 at 19.39.00 BST.png

  • Like 1
Link to comment
Share on other sites

One tool I can really recommend to anyone attempting this on linux is "filelight" (just apt-get it). It shows disk usage in graphic form, so I now know where 2GB of my 16GB eMMC has disappeared to!

 

Link to comment
Share on other sites

Interesting to know INDI supports GPSd directly - it's a little hidden away on the EQMOD mount menus but it will find the date/time and the long/lat etc from the linux GPSd which in turn supports most GPS systems.

 

Link to comment
Share on other sites

Managed to do create one of the small USB cables inside the box last night, just the ethernet connection and the standoffs/spacing of the components and the box can run closed with everything inside.

 

Link to comment
Share on other sites

Tonight will be ethernet cable and chassis mount :D  I'll probably put in a temporary spacer/insuation/hold - that way I can close the box for the first time :D

I think the usb cable for the SSD is a little too large but there are ways to cope with that - such as cutting a slot in the case and hot gluing.

Todo

* second usb cable & power - at the moment that's blanked off

* I have to sort out the mounting positions & final bracket with the wiring - mainly will it fit when the scope moves.

* reduce the length of the internal wiring

Focuser

* still need the limit switches.

* focuser coupling is still in the post.. work out why I'm not seeing the focuser V curve system (perhaps that's removed on the current dev build)

GPS
* FTDI connection and that's it.. (linux support the serial chip and INDI supports GPSd)

 

EDIT - successfully wired the ethernet but too late to start drilling things :/

Edited by NickK
Link to comment
Share on other sites

Fitted the external network connection into the box so now I can completely close up the box. 

I've noted a little pausing on the board so I'm not sure if that's the updating, the network cable too close to the power (interference) or the network connection. However the very loud beeps when the UBEC starts up are now muted :D I think the USB cable I had in there to connect the show2 screen has died so I need another small USB cable.. 

With having the focuser working, now I just need to mount the bits and mount the limit switches with the final wiring and it's ready for it's first test :)

Link to comment
Share on other sites

I have an old Garmin Geko 301 that has a serial connection... add a FTDI RS232 board, attach & power via USB and then add the linux gpsd todo the connection work... INDI can use gpsd to help setup the mount..

IMG_2590 (1).jpg

So this is nice little neat solution to finding the long/lat at SGL star parties.

 

I also found that the issue with the slow pausing was actually due to the ethernet cables dying and my mbp ethernet socket.. I've made some smaller cables and the link is far better. Currently the odroid internet access is bridged via my MBP's wifi link. When on the system's own wifi I can't bridge so hence the MBP wired connection.

 

Next up let's see the performance with astrometry.net.....

Edited by NickK
  • Like 1
Link to comment
Share on other sites

Hmm astrometry.net is being a pain.. well actually not astrometry but python (scripting language).

In short, python was born in modern days when people didn't have to worry much about what they're doing... 

 

TypeError: Cannot cast ufunc add output from dtype('int32') to dtype('uint16') with casting rule 'same_kind'

 

It looks like a dynamic type casting when attempting to parse the fits file.. so probably the numbers in the FITS file are being identified as the wrong type..

 

Slightly more annoying is that the mac formatted SSD is read only by ubuntu.. so it looks like I'll have to reformat that in linux.. which then will not be readable in osx..

Edited by NickK
  • Like 1
Link to comment
Share on other sites

38 minutes ago, NickK said:

Hmm astrometry.net is being a pain.. well actually not astrometry but python (scripting language).

In short, python was born in modern days when people didn't have to worry much about what they're doing... 

 


TypeError: Cannot cast ufunc add output from dtype('int32') to dtype('uint16') with casting rule 'same_kind'

 

It looks like a dynamic type casting when attempting to parse the fits file.. so probably the numbers in the FITS file are being identified as the wrong type..

 

Slightly more annoying is that the mac formatted SSD is read only by ubuntu.. so it looks like I'll have to reformat that in linux.. which then will not be readable in osx..

Kids today got no discipline! lol ;)

Link to comment
Share on other sites

Hehe easy option - write the image as a PNG rather than a FITS.. then run using that instead.

Ok, after first accidentally putting the wrong arc sec/pix scale in (using my 2x it's 0.88"/pix and using it native for this image it's 1.66" with the 383L's 5.4um pixels) it's quite quick for a 383L 17MB image. Actually it takes longer to pre-process then create the annotated output images (which are optional).

Top image solve in the picture is with annotated plotted outputs - ~4m

Bottom image solve in the picture without annotated plotted outputs .... 14 seconds. I can live with that :D And that's with (a) non-parallel index search, (b) 1 core being used and (c) SSD attached via USB2.0.

Screen Sharing Picture 22 May 2016 at 10.45.01 BST.png

 

Edited by NickK
Link to comment
Share on other sites

  • 2 weeks later...

So it's finalisation day :D with a predicted clear spot tonight I hope to get some polar alignment and a couple of guided, in focus long exposures.

Focus today is really push to complete the bits to a functional level rather than create pretty pictures but looking forward to seeing it working.

 

Link to comment
Share on other sites

So not quite as complete as I wanted - first setup and play on the pier! My trusty butane soldering iron leaks flames and brimfire out of the handle valve so no more soldering for me until it's replaced.

IMG_2598_crop.jpg

 

There's still a lot of work todo with the cables and hard screwing things that are currently stuck with hot glue..

The good news is that the focuser has a lot of torque and happily pulls the entire AP camera load by using the debug movement commands direct to the arduino's serial interface. The bad news is that some interaction with the indi moonlite driver seems to reduce torque and I think this is down to the speed between steps so it looks like I'll have to write something that works on the indi side... happily to report the wifi, the mount and the cameras all work (and once the focuser is sorted - that will work nicely too).

 

  • Like 4
Link to comment
Share on other sites

Got down to the bottom of the focuser issue - some intermittent wiring.. it even works at 1/32 now.

Video of the focusing using the INDI/C2/YARF: 

 

So to quote the Emperor... it is a fully functioning Death Star.. as all the bits work now - capture, focusing, plate resolving, mount control, external wifi desktop access.

Todo, once I have a new soldering iron:

* secure mountings of boxes and limit switches with bolts rather than hot glue (this works really well to find any issues of position before drilling holes)

* limit switch calibration upgrade to the focuser providing an absolute focusing system that can cope with the resolution ( log10(1600000)/log(2) = 21 bits required, i.e. 24bit 0xffffff) instead of the 65K 16bit resolution at the moment.

* replace the USB cables with shorter cables

* fit the SSD inside the main box... 

* fit external mount power connector

* make a stowage box so that the scope can be kept assembled.

 

Link to comment
Share on other sites

  • 2 weeks later...

Finally got around to bolting he boxes down to the bracket. Just have some USB cables and some power cable changes and everything is done. Should get myself a new soldering iron after the other one became a death trap!

Link to comment
Share on other sites

  • 1 month later...
  • 2 months later...
  • 10 months later...

Nick this is a fantastic imaging automation project. I am just coming round to researching this for myself, assumed a Windows NUC box would be an appropriate platform, then discovered INDI.

Have you looked at porting your system to the Odroid XU4 i wonder? The interesting thing about the XU4 is USB3 ports i.e. more usable for planetary imaging at high frame rates. I'd prefer to put some effort into integrating an INDI-based automation running on a low power Ubuntu platform than going the Windows stick computer or NUC route...

Tony

Link to comment
Share on other sites

  • 6 months later...
On 21/09/2017 at 19:17, tonyowens_uk said:

Nick this is a fantastic imaging automation project. I am just coming round to researching this for myself, assumed a Windows NUC box would be an appropriate platform, then discovered INDI.

Have you looked at porting your system to the Odroid XU4 i wonder? The interesting thing about the XU4 is USB3 ports i.e. more usable for planetary imaging at high frame rates. I'd prefer to put some effort into integrating an INDI-based automation running on a low power Ubuntu platform than going the Windows stick computer or NUC route...

Tony

Probably a little late in responding :)

I deliberately kept with the C2 because it's passively cooled, thus no cooling fan vibration on the mount/scope.

I don't see why this wouldn't simply move straight over to the XU4 or the N.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.