|
One of the nice things about having an observatory is that it is a much nicer place to play around with the mechanical settings than when sitting out in the open on the driveway. Not having to spend time on polar alignment each time frees up time to look at other things. The wind is much less of a factor, too. Two nights ago, I had good guiding of 0.5" RMS, something that rarely happens so I decided to finally sit down and check how to tune my stuff, way overdue.
As part of the tuning, I spent a whole day figuring out my adapters that came from who knows where meant for who knows what. I made a table of possible configurations for running the Baader MPCC Mark III (55 mm backfocus), my ES field flattener (45 mm backfocus) or nothing (like my Mak-Newt that has a flat focal plane), in combination with the ASI2600MM, ASI2600MC, QHYOAG-M with ASI120MM mini autoguider and ZWO 7x36mm EFW. This was when I switched my configuration from 6" Mak-Newt to the 12" GSO astrograph. I now have better stars than ever before with my Newt, so the work paid off. I got better stars in my ASI120MM OAG as well.
My setup is a DIY motor-integrated one-piece worm assembly (blocks bolted on an aluminum edge with pivot tightened from below by a pair of Belleville washers, outer block unconstrained) with Abec-9 bearings, no preloading, 17HM15-0904S direct drive on the pivot side, TCM2130 stepper drivers on 24V with SPI, bicycle tube rubber for SLW springs.
1) One thing I noticed is that the drive was wobbling tangentially because I had always been too careful with tightening the Belleville washers on the pivot. Tightening them up to the point where the friction becomes noticeable then backing off a little, leads to substantially straighter calibration curves. I also changed the calibration step sizes to be larger than the Ekos defaults.
2) My RA balance has apparently always been off. The RA balancing at 34 degrees always has a bit more friction than DEC balancing where you can put the DEC axis horizontal. As a result, I may have misjudged the balance and the worm apparently had a lot more (one-sided) friction. Tuning the balance such that the worm assembly pivots with minimal resistance, leads to a quite different balance that makes the drive work better.
3) My ZWO ASI120MM mini (on an OAG) is a bit noisy. Reducing the noise by 2x2 binning helps. I usually run at 2 seconds and noticed that my calibration curves have the dots in pairs - small jump, large jump and repeat. Going to 1 second leads to equidistantially spaced dots. Is this an ASI120 firmware issue? I may be looking for a better mini autoguider.
Most of the other settings (autoguider gains, multi-star vs single star guiding, integrating control action, aggressiveness, SLW spring tightening) don't seem to matter a whole lot so long as the settings are reasonable. Curious to hear about similar experiences, suggestions are welcome.
--
Mount: G11S mount with HD tripod.
Worm assembly: DIY motor-integrated one-piece worm assembly (blocks bolted on an aluminum edge with pivot tightened from below by a pair of Belleville washers, outer block unconstrained).
Bearings: Abec-9, no preloading.
Motors: 17HM15-0904S (0.9 degree) direct drive on the pivot side, TCM2130 stepper drivers on 24V with SPI, tracking at 32 microsteps.
SLW: bicycle tube rubber.
Control: OnStep (Wemos/CNCv3), Ekos on a Pi4b (8GB) mounted on-scope.
Scopes:? 6" MN152 Mak-Newt, 12" GSO Newt astrograph, ES ED127CF, RC6, ES ED80, Redcat 51.
|
Nice investigation and experiments, Henk.
Every single part needs to be investigated, and better replacements
tested. Great improvements in PE and DEC hysteresis have been achieved by
the same innovative approaches you describe.
In the end, you stop when your images achieve your goals.
The optics spacers and adjustments are another headache. The field
flatteners require a nearly exact back focus point... another headache.
The ASI120 has seen its day. Try an ASI178 for autoguiding. Very low
noise and very high sensitivity. You can also use it for excellent
planetary imaging (and Firecapture allows the planetary camera to also
serve as an autoguider at the same time as during planet imaging).
Have fun,
Michael
toggle quoted message
Show quoted text
On Fri, Jul 28, 2023, 10:01 AM HenkSB <haling@...> wrote: One of the nice things about having an observatory is that it is a much
nicer place to play around with the mechanical settings than when sitting
out in the open on the driveway. Not having to spend time on polar
alignment each time frees up time to look at other things. The wind is
much less of a factor, too. Two nights ago, I had good guiding of 0.5"
RMS, something that rarely happens so I decided to finally sit down and
check how to tune my stuff, way overdue.
As part of the tuning, I spent a whole day figuring out my adapters that
came from who knows where meant for who knows what. I made a table of
possible configurations for running the Baader MPCC Mark III (55 mm
backfocus), my ES field flattener (45 mm backfocus) or nothing (like my
Mak-Newt that has a flat focal plane), in combination with the ASI2600MM,
ASI2600MC, QHYOAG-M with ASI120MM mini autoguider and ZWO 7x36mm EFW. This
was when I switched my configuration from 6" Mak-Newt to the 12" GSO
astrograph. I now have better stars than ever before with my Newt, so the
work paid off. I got better stars in my ASI120MM OAG as well.
My setup is a DIY motor-integrated one-piece worm assembly (blocks bolted
on an aluminum edge with pivot tightened from below by a pair of Belleville
washers, outer block unconstrained) with Abec-9 bearings, no preloading,
17HM15-0904S direct drive on the pivot side, TCM2130 stepper drivers on 24V
with SPI, bicycle tube rubber for SLW springs.
1) One thing I noticed is that the drive was wobbling tangentially
because I had always been too careful with tightening the Belleville
washers on the pivot. Tightening them up to the point where the friction
becomes noticeable then backing off a little, leads to substantially
straighter calibration curves. I also changed the calibration step sizes
to be larger than the Ekos defaults.
2) My RA balance has apparently always been off. The RA balancing at 34
degrees always has a bit more friction than DEC balancing where you can put
the DEC axis horizontal. As a result, I may have misjudged the balance and
the worm apparently had a lot more (one-sided) friction. Tuning the
balance such that the worm assembly pivots with minimal resistance, leads
to a quite different balance that makes the drive work better.
3) My ZWO ASI120MM mini (on an OAG) is a bit noisy. Reducing the noise by
2x2 binning helps. I usually run at 2 seconds and noticed that my
calibration curves have the dots in pairs - small jump, large jump and
repeat. Going to 1 second leads to equidistantially spaced dots. Is this
an ASI120 firmware issue? I may be looking for a better mini autoguider.
Most of the other settings (autoguider gains, multi-star vs single star
guiding, integrating control action, aggressiveness, SLW spring tightening)
don't seem to matter a whole lot so long as the settings are reasonable.
Curious to hear about similar experiences, suggestions are welcome.
--
Mount: G11S mount with HD tripod.
Worm assembly: DIY motor-integrated one-piece worm assembly (blocks bolted
on an aluminum edge with pivot tightened from below by a pair of Belleville
washers, outer block unconstrained).
Bearings: Abec-9, no preloading.
Motors: 17HM15-0904S (0.9 degree) direct drive on the pivot side, TCM2130
stepper drivers on 24V with SPI, tracking at 32 microsteps.
SLW: bicycle tube rubber.
Control: OnStep (Wemos/CNCv3), Ekos on a Pi4b (8GB) mounted on-scope.
Scopes: 6" MN152 Mak-Newt, 12" GSO Newt astrograph, ES ED127CF, RC6, ES
ED80, Redcat 51.
|
Henk,
If you attached any photos, they never came through (at least I cannot find
any attachments). If you set up a free Flickr account, you can put them
thete and send a link.
I agree that epoxy is a wonder material, but it can fail with moisture.
They so make waterproof epoxies. Usually, boiling water or room temp
alcohols will take epoxies apart.
Best,
Michael
toggle quoted message
Show quoted text
On Fri, Jul 28, 2023, 5:37 PM HenkSB <haling@...> wrote: PS: These are the one-piece worm assemblies that I was talking about
They are easy to make from an aluminum edge and 2 small aluminum plates,
drilling some holes and attaching the edge with metal epoxy. I was afraid
that the epoxy might fail but it held up just fine and seem to work great.
--
Mount: G11S mount with HD tripod.
Worm assembly: DIY motor-integrated one-piece worm assembly (blocks bolted
on an aluminum edge with pivot tightened from below by a pair of Belleville
washers, outer block unconstrained).
Bearings: Abec-9, no preloading.
Motors: 17HM15-0904S (0.9 degree) direct drive on the pivot side, TCM2130
stepper drivers on 24V with SPI, tracking at 32 microsteps.
SLW: bicycle tube rubber.
Control: OnStep (Wemos/CNCv3), Ekos on a Pi4b (8GB) mounted on-scope.
Scopes: 6" MN152 Mak-Newt, 12" GSO Newt astrograph, ES ED127CF, RC6, ES
ED80, Redcat 51.
|
Hi Michael,
I indeed made a failed attempt to reply and upload a photo. Something went wrong, and I deleted the half-written post. Here it is again.
Attached is what aforementioned worm assemblies look like. Each consist of an aluminum edge and 2 small aluminum plates. I glued them together with metal epoxy (also drilled some holes for extra surface just in case). Having used them for quite a while now, I am confident that the epoxy holds up. They are easy to put together. Aside from them working well and - because the use of epoxy makes it easy to line the axes up - they are also easy to take off, just one screw (the pivot axis, with 2 small Bellevilles). SLW is simply by pulling it to the ring gear on the other end - I use a rubber band. Credit to Peter for the idea, and to Losmandy who made these kinds of assemblies long ago. I wonder why they stopped producing them.
Michael I looked at the ASI178MM. It looks good, but it is a bit wider than a mini. I would like a mini because the filter wheel is right next to it. The ASI174MM mini looks super. It's more than I wanted to spend but maybe I'll just get it.
--
Mount: G11S mount with HD tripod.
Worm assembly: DIY motor-integrated one-piece worm assembly (blocks bolted on an aluminum edge with pivot tightened from below by a pair of Belleville washers, outer block unconstrained).
Bearings: Abec-9, no preloading.
Motors: 17HM15-0904S (0.9 degree) direct drive on the pivot side, TCM2130 stepper drivers on 24V with SPI, tracking at 32 microsteps.
SLW: bicycle tube rubber.
Control: OnStep (Wemos/CNCv3), Ekos on a Pi4b (8GB) mounted on-scope.
Scopes:? 6" MN152 Mak-Newt, 12" GSO Newt astrograph, ES ED127CF, RC6, ES ED80, Redcat 51.
|
Very nicely made, Henk.
I see you are using stepper motors...not the Losmandy servo motors. So
you must be using something like OnStep not Gemini.
How are your images looking?
Best,
Michael
toggle quoted message
Show quoted text
On Fri, Jul 28, 2023, 8:17 PM HenkSB <haling@...> wrote: Hi Michael,
I indeed made a failed attempt to reply and upload a photo. Something
went wrong, and I deleted the half-written post. Here it is again.
Attached is what aforementioned worm assemblies look like. Each consist
of an aluminum edge and 2 small aluminum plates. I glued them together
with metal epoxy (also drilled some holes for extra surface just in case).
Having used them for quite a while now, I am confident that the epoxy holds
up. They are easy to put together. Aside from them working well and -
because the use of epoxy makes it easy to line the axes up - they are also
easy to take off, just one screw (the pivot axis, with 2 small
Bellevilles). SLW is simply by pulling it to the ring gear on the other
end - I use a rubber band. Credit to Peter for the idea, and to Losmandy
who made these kinds of assemblies long ago. I wonder why they stopped
producing them.
Michael I looked at the ASI178MM. It looks good, but it is a bit wider
than a mini. I would like a mini because the filter wheel is right next to
it. The ASI174MM mini looks super. It's more than I wanted to spend but
maybe I'll just get it.
--
Mount: G11S mount with HD tripod.
Worm assembly: DIY motor-integrated one-piece worm assembly (blocks bolted
on an aluminum edge with pivot tightened from below by a pair of Belleville
washers, outer block unconstrained).
Bearings: Abec-9, no preloading.
Motors: 17HM15-0904S (0.9 degree) direct drive on the pivot side, TCM2130
stepper drivers on 24V with SPI, tracking at 32 microsteps.
SLW: bicycle tube rubber.
Control: OnStep (Wemos/CNCv3), Ekos on a Pi4b (8GB) mounted on-scope.
Scopes: 6" MN152 Mak-Newt, 12" GSO Newt astrograph, ES ED127CF, RC6, ES
ED80, Redcat 51.
|
On Sat, Jul 29, 2023 at 12:18 AM, Michael Herman wrote: I see you are using stepper motors...not the Losmandy servo motors. So
you must be using something like OnStep not Gemini. OnStep with a Wemos and CNCv3, TMC2130 step drivers with SPI enabled for mode switching. How are your images looking? Here's a recent image, . The stats are here, . So far I was very happy with 0.7" total RMS but often it's closer to 1". I recently swutched to the dark side using PixInSight and BlurXTerminator. Mediocre images can be made to look much better that way. In the linked TSS thread there's also a DSS/Gimp version for comparison. I think I will get to 0.5" with the changes mentioned here. -- G11S mount with HD tripod.? Modifications:? OnStep controller with stepper motors, DIY SLW, belt drive, DIY "Medusa" on-scope power/Pi4 rig with Ekos for high level control.? Scopes:? 12" GSO Newt astrograph, Z12 Dob, ES ED127CF, RC6, MN152. ES ED80, Skymaster 20x80.
|
Thanks for the nice comments about DeepSkyStacker
David (DSS lead developer)
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] < [email protected]> On Behalf Of HenkSB Sent: 29 July 2023 11:32 To: [email protected]Subject: Re: [Losmandy-Hackers] Tuning On Sat, Jul 29, 2023 at 12:18 AM, Michael Herman wrote: I see you are using stepper motors...not the Losmandy servo motors. So
you must be using something like OnStep not Gemini. OnStep with a Wemos and CNCv3, TMC2130 step drivers with SPI enabled for mode switching. How are your images looking? Here's a recent image, . The stats are here, . So far I was very happy with 0.7" total RMS but often it's closer to 1". I recently swutched to the dark side using PixInSight and BlurXTerminator. Mediocre images can be made to look much better that way. In the linked TSS thread there's also a DSS/Gimp version for comparison. I think I will get to 0.5" with the changes mentioned here. -- G11S mount with HD tripod. Modifications: OnStep controller with stepper motors, DIY SLW, belt drive, DIY "Medusa" on-scope power/Pi4 rig with Ekos for high level control. Scopes: 12" GSO Newt astrograph, Z12 Dob, ES ED127CF, RC6, MN152. ES ED80, Skymaster 20x80.
|
Henk,
There is also the QHY 5-III 178M guide camera. This is what I use.
Peter
|
On Sat, Jul 29, 2023 at 11:58 AM, Peter Boreland wrote: There is also the QHY 5-III 178M guide camera. This is what I use.
Peter Thanks Peter, but unfortunately it is sold out and discontinued. At Agena Astro I just went with the QHY 5-III 200M that claims IR sensitivity and has large 4 mu pixels, always a good thing to reduce noise. It also has USB3 support whereas my ASI120MM only has USB2. The price is right, too. I hope my YI low light camera does not interfere too much with it. Its light emission appears to be in the same range. I can always mount my YI lower, so it does not shine down on it. Henk -- G11S mount with HD tripod.? Modifications:? OnStep controller with stepper motors, DIY SLW, belt drive, DIY "Medusa" on-scope power/Pi4 rig with Ekos for high level control.? Scopes:? 12" GSO Newt astrograph, Z12 Dob, ES ED127CF, RC6, MN152. ES ED80, Skymaster 20x80.
|
I am currently in Ekos source code hell as my installation did not support the QHY 5-III 200M (Ver. 2) camera. The camera works fine on Windows in SharpCap so there is nothing wrong with it. My best shot is to build Ekos and its 3rd party drivers from the latest source code, something that is useful to know anyway.
As it turns out, there are no clear instructions for the Pi4b specifically. I've seen threads in the Indi forum about building 64-bit astroberry but it does not look like anyone figured out how to do it. I am following the github instructions. So far, with the thin repository copy, I ended up with link missing symbol errors about pthread_detach/join/create, while the pthread .so library has those symbols, and is found for pthread_mutex_lock/unlock so what the heck. I am redoing it with the non-thin build and am puttering on. I will figure it out eventually.
Henk
--
Mount: G11S mount with HD tripod, Worm assembly: Spring-loaded direct drive one-piece worm assembly with Abec-9 bearings, Motors: 17HM15-0904S, TCM2130 stepper drivers (24V, 32 microsteps, SPI), Control: OnStep (Wemos/CNCv3), Ekos on a Pi4b (8GB), Scopes: 6" MN152, 12" GSO astrograph, ED127CF, RC6, ED80, Redcat 51
|
I use the scripts from to build the whole kstars /Ekos /Indi system on a 64-bit raspberry pi setup. There are some dependencies but cmake will tell you if there's anything missing.
|
On Thu, Aug 3, 2023 at 02:27 PM, Ed wrote:
I use the scripts from to
build the whole kstars /Ekos /Indi system on a 64-bit raspberry pi setup.
There are some dependencies but cmake will tell you if there's anything
missing.
Thanks Ed! I just discovered that there is a 64-bit astroberry release named Bullseye. On the Indi forum I got a reply from someone who was on Bullseye and had no problem building Ekos with the instructions on the indi and indi3rdparty github repositories. That may be all I need because it is a fairly recent release. Back to the drawing board. It will be easier than the rathole that I went down with the 32-bit Buster release. Henk -- Mount: G11S mount with HD tripod, Worm assembly: Spring-loaded direct drive one-piece worm assembly with Abec-9 bearings, Motors: 17HM15-0904S, TCM2130 stepper drivers (24V, 32 microsteps, SPI), Control: OnStep (Wemos/CNCv3), Ekos on a Pi4b (8GB), Scopes: 6" MN152, 12" GSO astrograph, ED127CF, RC6, ED80, Redcat 51
|
Finally, I think I found the problem, and just sent a support ticket to QHYCCD. Their camera is called QHY 5-III 200M (Ver. 2). To me, this suggests that there has been a version 1. The 85-qhyccd.rules file has a PID of 0200 for a camera named QHYIII200. When connecting my camera, both in Windows and Linux, it reports a PID of 0201. That suggests to me that this camera requires a separate driver and a rules file update. I was able to get the camera to test OK in Linux using a test utility that scans for any QHYCCD camera. But Ekos cannot find it because it uses the rules file that has a PID of 0200 instead of 0201. Just changing the number doesn't work of course.
To get here, I spent 5 days testing, and building KStars and Ekos from scratch. That is something that developers do a lot, and once you know how to, it is easy. Just to get to that point is a whole different matter. It is very useful though; now I can build a new version any time I need to. Maybe I can even fix minor problems. In order to get it working I had to wipe my Pi4b to switch from 32-bit Raspbian Buster to 64-bit Raspbian Bullseye. I guess the latest OS versions are supported, older ones perhaps not, so flashing a new OS may be needed.
QHYCCD is supposed to get back to me within 24 hours. I am curious what they have to say, but I think I hit the nail on the head and that it will be easy for them to provide me with an SDK update soon.
@Ed: Thank you very much again! That script is the only reasonable way of building KStars/Ekos, considering all the requisites. It also had the fxload update that was probably the only thing that was wrong, looking back. It is good to learn how to build everything from source, so I don't mind the extra work. At least I (think I) solved it.
--
Mount: G11S mount with HD tripod, Worm assembly: Spring-loaded direct drive one-piece worm assembly with Abec-9 bearings, Motors: 17HM15-0904S, TCM2130 stepper drivers (24V, 32 microsteps, SPI), Control: OnStep (Wemos/CNCv3), Ekos on a Pi4b (8GB), Scopes: 6" MN152, 12" GSO astrograph, ED127CF, RC6, ED80, Redcat 51
|
|
|
I am going to claim victory. When the seeing is good I will get 0.5" to 0.6" RMS total. I was on my way tonight when the clouds moved in.
The QHYIII5200M guide camera started working after a clean source code build. I recall that it got interrupted the first time, or maybe I did not clean up the old QHY SDK. The camera works very fine. I get lots of stars in my OAG. That IR capability works nicely. I use 3x3 binning and sample at 1 second with multiple stars and GPG activated (same as PPEC with PHD2).
Good balance is very important as well with the SLWs, and there has to be enough spring pressure. My calibration plots are spot on now, nice 90-degree angle and all in and out markers on top of each other. Before I start, I check the balance by pulling the SLWs and letting them snap back, then I set the pivot Belleville pressure such that the friction sets in then back off slightly.
BTW I added a Pegasus box, the "Pegasus Astro Pocket Powerbox Advance Gen 2 Power Supply ". This after I got repeated low power alarms with the new guide camera. I knew I was at the limit of what the Pi could handle (ASI2600MC + ASI120MM, now ASI2600MC + QHYIII5200M) so I needed to add a powered USB3 hub. I have one but it is large and has lights, and I cannot make it look as slick as that little Pegasus box so what the heck. It is about the same size as the Pi, and together I have 6 USB ports. No more power problems.
--
Mount: G11S mount with HD tripod, Worm assembly: Spring-loaded direct drive one-piece worm assembly with Abec-9 bearings, Motors: 17HM15-0904S, TCM2130 stepper drivers (24V, 32 microsteps, SPI), Control: OnStep (Wemos/CNCv3), Ekos on a Pi4b (8GB), Scopes: 6" MN152, 12" GSO astrograph, ED127CF, RC6, ED80, Redcat 51
|
Michael Herman
David C. Partridge
Peter Boreland