Getting the best performance from my AM5: Page 17

Getting the best performance from my AM5

Kevin_A

w7ay do you see the zigzag pattern in dec and ra in my photo that covers the full period length? Would a lower guide rate help with that or do you think that this is the best it can do without getting a better guide camera! Looks like it just can’t control the overall periodic error well on top of the random spikes.

w7ay

Kevin_A Would a lower guide rate help with that or do you think that this is the best it can do without getting a better guide camera!

Can't tell until we can see the correction pulses. See if you can expand the scale of the graph and see the individual correction pulses.

Problem is I don't see any region where the error is less than 0.5" (uncorrelated 0.5" RMS error in RA and declination will lead to a total RMS error of 0.707").

Chen

w7ay

Kevin_A do you see anything in this chart that i could try to lower the zigzag over the period length?

Kevin, you appear to have a mount that consistently gives large errors (up to 2"). Notice that when the error grows for RA, they often also grow for declination. But not always, and not always with the same signs.

That leads me to think it is just a mechanical imprecision thing (that is why you buy mounts from companies who treat it is an engineering project instead of an MBA project). But it could also be bad centroid computation too -- I notice that you are using a rather small guide scope with 162mm focal length, with a plate scale of 3.69 arcsecond per pixel. Try exactly the same setup (don't change anything else but the guide scope) with a 250mm to 350mm focal length scope and see if the graph changes.

Now, 3" of error may well be good enough for your main OTA (assuming you chose a small guide scope because you also have a small main OTA). So, it really depends on what you expect from a cheap mount.

Do you still have the data? If so, could you expand the x-scale so that we can see what leads to the large errors? Like, expand the region around 22:20, so we can see the actual correction pulses (turn the correction pulses on) by every second (I notice that is your guide rate).

You must have a lot of patience. By now, I would have ditched the mount and not throw more good money after bad.

Chen

Kevin_A

w7ay patience is my middle name.

w7ay

Kevin_A

OK, it looks like this is NOT the case of insufficient max pulse durations. Notice that the correction pulses do not consist of a constant amplitude comb. If there is insufficent max pulse, all of the correction pulses would be maxed out, producing a constant comb (think a crew cut :-). But we see correction pulses of different amplitudes.

So, the next question is why is the guide graph (which plots the centroid of the guide stars) swinging so badly. There are two prime candidates: (1) you have hit some spot on the mount gears where all of a sudden the errors became huge (like 2" huge), or (2) the centroid computation is fubar.

I tend to favor case (2) [I am an "occam's razor" type guy] since the likelyhood that you hit bad spots of the gears on the RA motor and the declination motor at the same time is small, unless you have something loose that can couple the RA axis to the declination axis (but I assume you have tightened every dovetail plates there are; if not, check how securely the guide scope is held down, relative to the mount itself, not relative to the main OTA -- this is why I use a dual saddle plate with an independent guide scope mount).

Try a much larger guide scope with a flat field, and make sure you have no fewer than 12 stars running in multi-star guiding (don't be shy with the guide camera gain setting). And make sure that the stars which are selected all have similar brightness. Don't let a single bright star dominate the mean of the centroid, since ASIAIR uses SNR as a weighting function -- a very bright star in multi-star guiding becomes a single star guiding, and yu will need guide exposures that are long to counter "seeing." Increase the gain of the guide camera sufficiently so that the first couple of guide stars get rejected because they saturate.

If it is a case of bad centroid estimation (my suspicion), then nothing with the guide parameters will help. I.e., max pulse won't change anything, guide rate won't change anything, etc, (even changing to a 10Micron mount won't change anything) since the measured centroid is bogus, and causing those excursions in the guide graph. You are attempting to move the mount based on bad data (centroid). At the same time, you are needlessly (and erronously) issuing guide pulses that should not be issued. What you need is better centroid estimation.

As I mentioned earlier, your guide scope may well be appropriate for your main OTA. 3 arcsec may be good enough.

Chen

Kevin_A

w7ay that is why I have been asking you a lot of guide camera related questions.
On my tiny Rokinon setup, i can not use much more of a guide scope size increase but i will try bumping up the gain even further since i did notice that 1 to 2 stars of the 12 were much bigger than the rest. Is worth me trying to bin2 the guide camera to get more star definition? I have both 290mm mini and 120mm mini cameras and my only concern will be my biggest scopes as my tiny rokinon setup is a 3.6 image resolution so sub arcsec guiding wont be as important. My 162mm is a f3.2 50mm dia doublet guide scope and it is fairly decent… but my stars on all my guide scopes seem very blobby so maybe i should try a fma180pro when they become available. Do you think that a 0.25 guide rate will help also when asiair impliments it? Sorry for so many questions but it helps my patience! Haha thanks Chen!

w7ay

Kevin_A Is worth me trying to bin2 the guide camera to get more star definition?

With your guide plate scale, that would be disastrous!

You are already at 3.6"/pixel. Bining by 2 would make it over 7" per pixel! You need to go the other way (i.e., by using a longer focal length).

Do you think that a 0.25 guide rate will help also when asiair impliments it?

Not when the guiding system is given bogus centroids to start with.

it helps my patience!

Not mine.

Chen

Kevin_A

w7ay haha. I think I will just have to try a better guide scope like the 180 or 230 with flatteners. I thought that binning while worse might increase the centroid size just like a lot of people say that out of focus guide stars are better for phd2 to find the centroid. That was my point. I think I am getting to the point of trying anything unconventional as I have tried and used everything that makes sense and what should actually work. I often wonder if it is a asiair issue or bad guide scopes since all 3 of my guide scopes (240, 162, 120mm) and guide cameras seem to have low quality guide stars or I may just have 3 bad guide scopes.

w7ay

Kevin_A I often wonder if it is a asiair issue or bad guide scopes since all 3 of my guide scopes (240, 162, 120mm) and guide cameras seem to have low quality guide stars or I may just have 3 bad guide scopes.

I have had no problems with ASIAIR guiding once they upped the guding frame rates to 2 FPS, and even better recently after dropping my guide rate to 0.25x sidereal.

But you need good centroids first -- I don't use tiny guide scopes or small sensors. The smallest guide scope that I used in the past was a Borg 36ED (focal length 200mm). And since then, Borg 55FL, and in the past month or so, an Askar 180p. I also abandoned the ASI290MM class cameras, and went to an ASI178MM. A QHY678M arrived today from Agena, and I will start some time experimenting with that (writing my own software on INDIGO), replacing the ASI178MM.

FWIW, I also use Bahtinov masks to focus my guide scope (lately, with EAFs) as best as I can.

Chen

Kevin_A

w7ay i think i might buy a smaller pixel 178 as i am not fond of the 290 either as I expected better images. I think I will see if I can find a better guide scpoe too. Borg 55fl may be a bit too much for my liking but maybe a better 60mm william optics 61 doublet or SW evoguide 50DX. My cheap Chinese 60mm is garbage so maybe it is just poor star images that is the core issue as my past 2 mounts had similar guiding roughness issues and now there seems to be a commonality. Thanks.

w7ay

Kevin_A SW evoguide 50DX

I do have an Evoguide 50ED with field flattener, but the construction is amazingly rough and heavy, I didn't bother to attach an EAF to it, and thus have not tested it so far. The threads on it look like they were machined in a high scool machine shop.

If the 50DX is like the 50ED, the focuser is rather odd -- the focuser roughly turns one way to get rough focus, and from there get a small region to use as fine focus in the reverse direction. I have not tried to tap a set of holes to mount it, and I am not terribly fond of tube rings of the 50ED.

EVOguide and FMA180p side by side:

Whatever you use, make sure that the spot digram of the OTA is good all the way to the image circle of the ASI178, and preferably wider, to plan for future better cameras.

Also, if you don't already have an ASI178MM and are not stuck with the ASIAIR, you might as well go to a monochrome IMX678 sensor (with 2µm pixels), like the QHY678M camera. A larger sensor is even better.

Chen

Kevin_A

w7ay have tested the star quality of both? These are my two contenders as the WO 61 is just too slow. I am leaning towards the evo50dx due to it being 242mm and i could use all the extra rez i can get but star quality matters most to me. Not sure i would need the flattener on my 290 but maybe on the 178. Machining aside i could maybe put a helical fine focuser on the back end to help focus better.

Gernot

w7ay

Hello Chen!

As promised, I'll show you my first test results here. Since the nights are very short at the moment I was only able to record one session with 32 minutes. I used the default settings of the ASIair Plus. I set the aggressiveness to 50:50 after 15 minutes. Can you use this data to evaluate the quality of the mount?

CS Gernot!

phd2-guidelog-2023-06-24-233417.txt

163kB

w7ay

Gernot Can you do something with the data?

I don't think I have the patience to keep repeating what has already been written. You just have to go back and read the old posts.

As I mentioned before, I don't have a dog in this hunt. I own two strain wave mounts, but they are not manufactured by ZWO.

Perhaps you can ask ZWO to help you improve the guiding results of their mount.

Chen

Kevin_A

w7ay

Gernot I agree with Chen, ZWO should be giving all the advice to its users oreven a little bit. But without them, it looks very typical… similar my mount too. The weakest link I am finding out is that most of these mounts can benefit from smaller pixel cameras on bigger focal length guide scopes, especially if you see just marginal numbers with a rough guiding pattern. Also a benefit is bumping up the guide camera gain so you can get a minimum 12 even size stars for multiguiding. Ensure that the max durations are below 450ms on both axis if running 1s guide exposures and below 50% aggressions. This can be fine tuned with lower guide pulses but it is a good start for you to start experimenting with. Your mount is a typical mount and anything under 0.7 should be considered very good considering it is marketed as a portable amateur mount.

Gernot

w7ay

Hi Chen!

I understood that so that you look at my measurement data again... but no problem 🙂 I will continue to read up on the topic!

Thank you for your efforts!

CS Gernot

Gernot

Kevin_A

Hi Kevin!

Thank you for your assessment and your suggestions.

I am happy with my mount 🙂

CS Gernot

w7ay

Kevin_A I am leaning towards the evo50dx due to it being 242mm and i could use all the extra rez i can get but star quality matters most to me.

IMHO, the build quality of the EVOguide is very "Chinese." I.e., heavy for no reason, and poorly machined -- worse than a typical Askar even (IMO, the typical Askar is about the same as a WO). Just compare the threads to the ones on Takahashis and PreciseParts.com.

The Askar 180p is a pleasant surprise. If I didn't know it was designed in China, I would have thought it was designed in Japan, Korea or the West. (If you want to split hairs, the Samyangs don't have the build quality of the Sigma ARTs or the Canon Ls, but the build quality of an RST-135 is nothing short of exquisite). The build quality and attention to engineering details (what ZWO really lacks, for example) is definitely a couple of notches up from a typical Askar -- even the focusing ring is notched to directly take an MXL belt.

As I had earlier mentioned, I haven't looked at the stars of the EVOguide, but from various postings on Cloudy Nights, the flattener is definitely needed. Like the FSQ-85, the EVOguide's flattener does not change the focal length. (To be pedantic, the flattener on the FSQ is actually 1.01x and not 1.0x).

The 180p optics is identical to the non-p version. Quite decent, at least for the FOV of an ASI178. The Bahtinov diffraction spikes are quite clean (I find that the straighter and thinner the spikes on a monochrome camera, the more likely it is more APO-like, as the spikes are really made up of segments (dashed lines) of color components, which you can see on a color camera). But like all Askars that I have (I also have the ACL200, FMA230, FRA300), they all have a chromatic bloat. The spot diagram can be found here at FLO's site:

https://www.firstlightoptics.com/askar-telescopes/askar-fma180-f45-ed-apo-v2-astrograph-lens-reducer.html

Notice that the spot's GEO radius (which includes chromatic) is significantly larger than the RMS radius. Even at the corner of an ASI178 (the center to diagonal of an ASI178 is about 4.5mm), the blue bloat is quite bad (their spot diagrams have distances of 3.999 mm and 9.999 mm from the optical axis, so you have to do some guessing of what 4.5 mm is like). And the spot diagrams are paper-simulations. By the time their assemblers are done, the real-world is probably much worse. The GEO radiuses are much larger than the Dawes limit of the optics.

But I doubt the EVOguide is any better, and more likely, knowing the manufacturer, is much worse. SkyWatcher does not even publish spot diagrams. Do you trust manufacturers of technical equipment that do not publish technical specs? We have lots of examples of that lately.

Attaching an electronic focuser to the EVOguide would be a complete nightmare. I have looked at it and short of drilling into the OTA tube itself, there is no neat way to do it. If you remember, I had posted images on how I attached an EAF to the two tapped screw holes of the FMA180p finder mount, by only having to drill the EAF -- no modifications to the stock FMA180p.

Chen

jpoulette

Being a complete newbie to astro-imaging (01/2023), I was shocked at how easy ASIAIR+ was to set up and how accurate it was at tracking. After working a few bugs out (mostly of my own design) I routinely get RA and DEC each at 0.25" for the entire session. I suspect tripod/mount level, stability, and payload mass are critical factors. My payload is ₅ kg and I have 3 markers on a mature asphalt driveway where I set up. I level like a fool (sometimes for 1/2 an hour) and check level both before and after the mount is attached. Good luck to everyone!

Kevin_A

jpoulette good results are usually a sign you got a good mount. It seems it is a 50/50 chance these days. Mine guides all day long at 0.66 and once it guided well at 0.3 until i realized my scope focal length was double. Haha mine often guides at 0.2 but still averages 0.66 over the entire night. As long as your stars are round anything under 0.7 will look great!

Kevin_A

w7ay thanks Chen, the evo had at least specified the glass as fairly decent but if the tube is poorly manufactured who knows how the lens cell is held in place or if the camera end sags and has tilt. I am thinking it is a pass on that for me. I might look at the fma230 and the 180pro but nothing is in stock anywhere so it gives me more time to look at more reviews, diagrams etc. Thanks again as I appreciate your reviews and info. One thing I dont want to do is throw too much money at it to just compensate for a crappy mount. But with a 180pro, I can at least try it as a imaging camera for my 183mc pro if I ever want to.

« Previous Page Next Page »