I am noticing that when I look at the guiding graph in asiair, that while guiding is acceptable, there is very little small guide pulses to keep it in check and it only seems to have a few large pulses when my graph gets spikes? Anyone else notice this? Guiding generally is in the 0.4 to 0.8 range but no corrections in either axis are issued unless a 1s spike occurs. Maybe asiair just is not showing smaller pulse corrections?
Getting the best performance from my AM5
- Edited
Kevin_A it only seems to have a few large pulses when my graph gets spikes?
OK, we are finally getting down to the gist of it all.
Did you see an especially large guide pulse before the large guide graph deviation?
If you are seeing large number of pulses before the guide graph has gone, up, PHD2 is doing something weird (issuing guide pulses without any stimulus).
However, if you are seeing large number of pulses after the guide graph has gone up, this is "normal." However, you need to investigate it the rise in the guide graph is cause by (1) wind (I see the effect of even 3 mph gusts (measured on NetAtmo 15 feet away from the tri-pier); but the graph would deviate for a couple of seconds, and settle back), or (2) by centroid estimation error (i.e., the guide scope "lying" to PHD2 that the sky has moved even though it has not), or (3) by the mount' mechanism not being able to keep sub-arc-second precision.
(1) and (2) are likely to affect both axis at the same time. (3) tends to happen to just one axis at a time. So that could be a clue.
I was just going through 6 mph wind gusts last night, and autoguiding deteriorated to the 0.5" range from the 0.35" range the night before, when the gusts peaked at no more than 2 mph.
Chen
Revised: I had noticed that when I look at the guiding graph in asiair, that while guiding is acceptable, there is very little small guide pulses to keep it in check and it only seemed to have a few large pulses when my graph gets spikes? Guiding generally is in the 0.4 to 0.8 range but no corrections in either axis are issued unless a 1s spike occurs but asiair just is not showing smaller pulse corrections it seems. Actually in phd2 log viewer they are there… just not in asiair app. Another thing I just noticed is in the latest firmware 10.74 for asiair plus the minmo is 0.1 but in the same firmware version for the older asiair pro it is 0.2. No wonder my 2 older asiair pro guide much worse on my am5.
Zwo can you fix this as my guiding is 50% worse on my asiair pro setups using a minmo of 0.2.
- Edited
w7ay the pulses in log viewer are there and happen right at the middle of the peak so it is working it seems… they are just not being shown on the asiair graph. I did find out why 2 of my older asiair pros are crappy… their minmo is 0.2 vs my asiair plus at 0.1. Same firmware update.
Weird but now I know why 2 rigs are good and 2 are bad.
w7ay I guess I could go back to using phd2 but I think I prefer imaging wirelessly from my couch. Haha. But when I build another house with a real observatory then I will go back to where I can decide and use my own settings. Winter is coming soon and I just can’t do cold anymore!
I hope they fix their errors in the next update. I moved both asiair pro units to my widefield scopes so higher guiding numbers are less critical . I did try an OAG on my Askar FRA300 and since the scope delivers crisp stars at F5 and with my 533mc pro having a tiny sensor, i could get it deep towards the centre… it worked quite well when I pumped up the gain. I imaged all night at 0.5rms and under.
Star quality definitely matters with this ecosystem.
- Edited
Kevin_A I guess I could go back to using phd2 but I think I prefer imaging wirelessly from my couch.
You can do the same using this with a Raspberry Pi 4 (guess how I know :-):
https://www.indigo-astronomy.org/indigo-sky.html
From indoors, you can do rudimentary stuff (planetarium, mount control, camera control/scheduling, PHD2 guiding) just from any web browser (including on tablets), or use a full blown laptop on your couch with full blown components:
https://www.indigo-astronomy.org/software.html
The components are all packaged as a turn-key INDIGO "All-in-one" A1 if you are lucky enough to own a Mac:
https://apps.apple.com/us/app/indigo-a1/id1617125771?mt=12
https://www.cloudmakers.eu/a1/
The second link has videos to how to do things like polar alignment, etc.
(Notice that it collects no user data. Even if it does, it would go to a NATO and EU country, instead of to China.)
If you prefer YouTube (comic book) documentation, Rumen Bogdanovski (one of the two principals) even has a YouTube video:
https://www.youtube.com/watch?v=ko1Qjs9Zdes
Rumen [https://bg.linkedin.com/in/rumenbogdanovski] owns one of the earliest AM5, by the way. Peter Polakovic, the other principal, is a MacOS nut -- he ordered his M1 Mac even before I did :-) For day work, one of them is an astrophysicist, and the other is a computer IT professional.
Best of all, you can use better cameras, filter wheels, and image rotators than you can on an ASIAIR.
Even better for a geek, since the INDIGO bus is completely open sourced, and documented (in fact, the only place I even found the documentation of communication protocol for the ZWO mounts was in INDIGO's documentation), you can add any of your own software (that is how I run my All-Sky camera nowadays, though a Raspberry Pi 4, and my own macOS program, using a no-longer used ASI294MC).
https://github.com/indigo-astronomy/indigo
Everything looks like devices on a networked bus. Add as many servers as you want to the bus, and access them from multiple clients.
I have already 12-volt-modified a M1 Mac Mini (retired it as a desktop computer when I replaced it by the Mac Studio) to use as my outdoor server. Even if you buy an M1 Mini new from Apple, you can get it for $600, and it is orders of magnitude faster than an ASIAIR.
So, you can go from just a single Raspberry Pi 4 and a tablet running a web browser, to a serious system where the sky is the limit. The upgrade path is already there today. Just add more processing power (more microcomputers, or simply a faster microcomputer) when you feel the need. (Oh, INDIGO can run any device that is supported by INDI -- you can even find humidity sensors. :-) All of your current ZWO devices will still work: I have tested ZWO EFW, EAF, cameras, etc. Just get a Raspberry Pi 4, download and burn an microSD IndigoSky Image, and start playing from just a web browser.
Chen
w7ay I will keep it simple for now as i am getting consistant 0.5rms now that I tweaked a few things and put some better feet on my tc40 tripod, bigger guide scope, bumped up gain etc… but I think that is the best this mount can do and I am fine with that. It will never be an RST and as long as zwo stop making bonehead moves like increasing minmo for us strainwave dudes… then I will keep it simple for now. I just need the next update to reduce the minmo of the asiair pro for my older setups and add am5 guide rate adjustability and I will be happy.
Cheers,
Kevin
Also, one note, some progress on my side too. When I use BIN=1x1 on the guide cam (instead of the previously used BIN=2x2), the guiding results are in the realm of 0.4"-0.6" instead of 0.8"-1.0"+
This is true both with a guide scope (60mm, 254mm FL, ASI ZWO 120MM) and with an OAG (ASI ZWO 220MM mini).
Also, ZWO recommended that I open the AM5 case and check the belt tension; they asked me to adjust it by pushing the RA motor away to fasten the belt. I didn't do this because the method is quite rudimentary (unscrew, push, tighten - no easy way to measure the adjust movement)) and has no easy fallback.
With 1x1 binning the performance is OK for my FL (1628 mm) so I think I'll just stick with it for now.
w7ay Awesome links, thank you. I was getting desperate after multiple unsuccessful attempts to make my gear work with a Linux laptop and just tried INDIGO today - it works!
For everyone, here's a very interesting article written by the authors of INDIGO, featuring the ZWO AM5:
https://github.com/indigo-astronomy/indigo/blob/master/indigo_docs/GUIDING_MISCONCEPTIONS.md
I don't know if I will accomplish anything by posting this, but I have recently seen the same large RA oscillations in my AM5. Autorun (on M27) is where it cropped up. In complete desperation I went back to Preview and forced a calibration, and it stopped. This may have been a coincidence, but I thought that I would add the data point. I assume that there is a subtle (or not-so-subtle :/ ) bug in the guiding routine or the mount firmware.
I've had the AM5 for almost a year now, and other than this recent development I've been completely satisfied with it.
-Jon
5 days later
BrineyEye Would you please contact our support and provide your guiding log. It would help us to deal with your issuse. Thank you.
5 days later
BrineyEye and ASIMount@ZWO The guiding with ASIAIR appears to be sub-optimal to me because of the number of stars ASIAIR chooses to use for guiding. While I can visually see a number of stars, ASIAIR typically picks only 1 or 2, or starts by picking 5-6 then a few minutes later it's down to 1-2 again. With the AM5 requiring fast corrections, e.g. 0.5 s guide exposures, this means that you're chasing the seeing instead of correcting the mount tracking errors. Averaging the centroids over 10+ stars would more or less compensate for the seeing induced movements of every individual star. I've been having the hardest time trying to keep ASIAIR track more than 2 stars. This is where you can go from 0.4-0.6" guiding to 1.2"+ guiding with the same mount...
Jan75 this isn't likely a function of the ASIAir. More than likely, it's the camera. I use an ASI178MM, and consistently get 10 to 12 guide stars on the Air. I also have an ASI290MM that gets about the same. My old 120MM would barely get 3 to 5.
- Edited
I agree about the camera - I upgraded to an ASI220MM and there was a significant difference. I also found that a larger aperture guidescope helps - I went from a WO 32mm to WO 50mm and saw a further improvement.
A larger guide scope does wonders. Also, having a guide camera that has clean higher gains help. My 290 and 220 are great but my 120 is gain limited. Going from 30mm to 60mm guide scopes is a game changer. Another thing I have noticed with the AM5 is that the heavier scope setups guide much better than a very light scope. Maybe it is the bigger guide scope! My guiding is 0.7 with my Rokinon 135mm rig but down at 0.35 average with my big 648mm triplet.
- Edited
What is fundamental with multi-star guiding as far as guide scopes, is to have at least 20 guidable stars available in a guide frame.
What this means is
a good guide scale. Oversampling will reduce the intensity of a star (and not be recognized), while undersampling will case centroids to be completely bogus. So it is more advantageous to oversample, as long as you make up the deficit.
a large sensor. A larger sensor will allow more area of the sky to be imaged, as long as the scope produces a flat enough field to handle the sensor. Just a larger sensor is not useful since any poorly shaped star that are away from the optical axis will do more harm than good. By the way, this has nothing to do with whether the scope is APO or not -- an APO objective converges three wavelengths of light to the same point; but with a refractor, those points will not be on the same sensor plane. You need a field flattener. (The is the biggest joke about the SeeStar. Notice that ZWO claims that it has an APO objective, but there is no flattener! None of the YouTube shills have pointed this out, by the way. ZWO always like making boneheaded design decisions like this -- just like putting emphasis with their mounts have small PE amplitude, while the real problem is the slope of the PE curve.)
at the same time, the sensor must not be so large that the guiding frame rate falls below 2 FPS.
the plate scale plus a large sensor may require a larger objective to get enough guidable stars (the 20+ that I mentioned earlier). You need enough photons.
good focus is imperative (my guide scopes all have an electronic focuser). A large blob of a star will spread the photons over too many pixels. The better focus, the more stars you will see in the guide frame.
Why 20 guidable stars? This has to do with the distrrbution of star intensities, and the fact that ASIAIR uses SNR as the weighting function; including the brightest stars will cause the averaged centroid to just depend on the brightest one or two stars. Even when guiding with 12 stars, you will effectively be guiding using only 2 or 3 stars. So don't get fooled by ASIAIR. You need to increase the gain of the camera so that the first two or three stars are saturated. That will cause ASIAIR to use the remaining stars, which have a more even intensity distribution (thus more even SNR).
Typically, an 10 dB increase in camera gain after attaining 12 stars will allow the brghtest stars to be rejected, but sometimes, you will need an additional 10 dB. ZWO uses a gain scale of 0.1 dB per gain unit, so be prepared to add a gain of 100 to 200 after being able to capture the first 12 stars.
I cannot overemphasize this aspect since picking the first 12 brightest stars may only give you the equivalent of 2 or 3 stars guiding (definitely insufficient for typical "seeing").
Getting a good centroid is only one aspect of the guding. It does make the difference between guiding to 0.35" (total RMS) vs guiding to 0.5". However, if your mount can only guide to 1" to start with, unless you are using ZWO's guide scope/ASI120 combo (really poor choice, I'm sorry for those who fell for it just because it is cheap), then improving your guide scope may not help.
Chen
.
13 days later
- Edited
w7ay Hello Chen - Reading this post on how to calculate the min-move settings for the AM5:
https://bbs.astronomy-imaging-camera.com/d/15989-getting-the-best-performance-from-my-am5/186
I'm curious - I typically run the PHD2 Guiding Assistant for 3 or 4 minutes after calibrating PHD2 - and I let it suggest the min-move settings for my current mount (AP Mach-1 not the AM5), and I use those because they are from data gathered at the time I'm imaging. Curious if the Guiding Assistant will give good values for the AM5 (assuming they do for my Mach-1, which appears to be the case).
Thanks,
JD
- Edited
There are too many anomalies with this mount that I have given up giving advices on it. It is not just poor and noisy PE, but random anomalies.
If you can afford it, there are Strain Wave gear mounts that actually work consistently. While it will never guide as well as your Mach-1, my RST-135E can consistenly guide on the long term at 0.35" or better even with an ASIAIR, if there is no wind.
Because of the weight and age issues, I started a couple of years ago with two RST-135. I added an RST-135E recently, and liked it so much that I added yet another RST-135E. For shorter focal lengths (and planetary), I just let the Renishaw encoder do its thing, without autoguiding.
Sorry, but I am out of the business of figuring out all the ZWO mount anomalies.
Chen