Getting the best performance from my AM5: Page 15

Getting the best performance from my AM5

AstroBen60

here is my analysis of my test report:

As far as I didn't make any mistake, the max first derivative is 0,21 arcsec/sec, what seems quite good anyway.
But the truth is in the pudding, so let's wait for clear ang long nights to come back.

Ben

w7ay

AstroBen60 I mean having a general 0.3asec/" derivative and peaks like 0,8"/sec

Hi Ben,

You would have to set guide parameters (such as guide rate, max pulse duration, etc) based upon the worse of the two (actually worse possible, which is probably not shown in the piece of paper, since it emphasize the peak amplitude [mostly dependent on fundamental] and not the region of worst case slope [highly dependent on the higher harmonics] -- this is why I'd recommend doubling the number that you can see on the piece of paper, to be safe).

The 0.8 arcsec/sec does not bode well (read on further down this post, since you also mentioned a 0.21 arcsec/sec number). The reason is that if you were to use a 1 second guide exposure, a guide star (or stars) would have moved by 0.8 arc seconds within that exposure. You now have to measure the centroid of that small streak instead of the centroid of a fixed star.

Using a 0.5 second guide exposure would reduce that streak to a more acceptable 0.4 arc seconds. I would gently suggest that you try a shorter guide exposure, as long as your guide instrument is up to it. To use 0.5 second exposure, you will definitely need to use multiple-star centroiding (averaging the centroids of multiple stars reduces the variance of the "seeing" error by two for each doubling of stars, if they all have the same weight in centroid computation). You will need a sensor with a large FOV so that it maximizes your chance to see more stars. And you need a larger aperture since you need a larger light bucket.

As an aside, I had for a couple of years been using a Borg 55FL objective as my guide scope (200mm focal length or 250mm f.l. depending on whether the reducer is used; thus f/3.6 or f/4.5) and an ASI178MM as the guide camera that has a decent sized FOV compared to what others typically use as their guide camera. For the past month, I have been experimenting with using a 40mm aperture, 180mm focal length as a guide scope (with same ASI178MM) to see if I can use a camera with a smaller "light bucket," but with a larger FOV to try to find as many guide stars -- that FMA180 appears to work quite well -- ASIAIR could always find 12 stars even in star-starved regions near M81/M82 with a camera gain of 250 (analog gain of 25 dB, which appears excessive, but is fine). So, a smaller and lighter guide instrument that works just as well as my earlier setup. I had to drill some holes in the EAF to mount one on the FMA180(pro), but just a 30 minute exercise (helical focusers are impossible to get pinpoint guide stars without some kind of aid).

Just remember that guide exposures never produce pin-point stars. The star is moving (due to the slope of the PE curve) while exposure is occurring. The best you can do is to focus really well so at least you reduce other centroid measuring errors. This is that "sawtooth" that I keep referring to.

Now, your 0.21 arcsec/sec is much more encouraging, and lines up with what others are seeing with their 430 second periods.

With that kind of slope, you should be able to use short max pulses of around 150 ms even with a guide rate of 0.25x sidereal rate, and a 0.5 second exposure (2 FPS). After discovering that sawtooth graph, it occurred to me that reducing the guide rate will also reduce the sawtooth amplitude (and thus centroid estimation error), so I have been using 0.25x sidereal guiding since. It has been a rather major factor in guiding for me. Whether you can also use a slower guide rate will depend on the worse case slope of the PE curve, and it looks like your mount should be able to handle it (although I think Kevin has reported that ASIAIR cannot choose 0.25x sidereal guide rate with ZWO's own mount; so you may be out of luck there if you use ASIAIR -- try INDIGO Sky instead of ASIAIR if you want to stick with a small microprocessor as your controller).

Good luck.

Chen

AstroBen60

w7ay

Hi Chen

Thanks a lot for your reply

The 0.8arcsec/sec I was talking about was just an example. The max value I have on that tiny part of the whole measurement is 0,21,what lies within the average slope most people have with these mounts (0.17 to 0.35 from what I read ), maybe a little bit on the good side. The only difference I see is that we often see large Max PE amplitude on previous mounts with 432 period, and this one is quite reduced, like the one posted above, also a 288 period one. But as you say, we don't really care as long as we are guiding... And looking at the graph, we can see that for let's say at least 85% of the time, it's under 0.1. Some frames might be devoted to dustbin, but that's not catastrophic neither.

I asked zwo the complete dataset. As it has been measured, it must be stored somewhere. Will they acces to my request...? As you say, a conservative approach would be to double the max value we get from our measurement.

Anyway, IMHO, these mounts are really devoted to travel, easy setup, and short focal lengths. So the resolution of these short FL setups, associated with typical pixel sizes of 3.7u, will be forgiving.
But it's a nonsense to me when I see C11s on it 😳.
As long as you can't take your scope in a bag pack, or cabin luggage, what is the point to have a compact mount to put a huge tube on it? When buying these mounts, we pay for compactness, easy setup, not for accuracy. I have a feeling that, with typical sensor chips, anything above 600/700 mm is not adequate with these harmonic mounts, they are not made for that.

I actually guide with an OagL and a 178mm also. As my focal length is quite short (400mm), I don't really have difficulties to get several guide stars, so I should be able to test your recommendations. ( as soon as the clouds will get away, and the real night will be back..).

And Finally, I don't use ASIAIR. I always put minipc's on my mounts ( mele Q3), because I can't live without NINA. So this shouldn't be an issue to test 0,25x guiding rates.

Thanks a lot for all the extraordinary stuff you posted here, I learned a lot.

Ben

w7ay

AstroBen60 As it has been measured, it must be stored somewhere. Will they acces to my request...?

I agree, Ben. That is what allows the "reference microphones" to be built relatively cheaply nowadays, for example -- their serial number lets you access a per-microphone database for use with compatible software, making the microphone be the equivalent of much more expensive ones.

But it's a nonsense to me when I see C11s on it

Agree 100%. The strain wave geared mounts can swing very heavy payloads like a rag doll. But I am not sure if they can guide heavy payloads with long focal lengths well enough. The heaviest payload I have on mine is an FSQ-85. I will be testing the RST-135 with a Mewlon 180C when it is scheduled to arrive this Autumn -- if it struggles, I would need to go to a RST-300, since the Pryxis NYX-101 (who also never built a mount before) has problems of its own. But forget travelling with an RST-300.

I always put minipc's on my mounts ( mele Q3), because I can't live without NINA.

Hah, I have been writing my own software to use INDIGO on macOS. The latest fun exercise is to construct a new All-Sky Camera. My earlier camera had use a discarded first gen ASIAIR on WiFi, with a color ASI178 and a CCTV lens. The new camera uses a Raspberry Pi 4 running INDIGO Sky, Ethernet PoE for a single wire connection, an uncooled ASI294MC, and a Samyang E-mount 8mm fisheye lens. I am especially happy with ditching the ASIAIR, which is severely limited in what it can do.

Chen

Kevin_A

w7ay just a quick question for you Chen… is there a reason you guide with a 178 vs the standard 174 guide cam? Just bigger sensor or usb3 or ? I am thinking of getting a Askar 180pro and maybe a new guide cam with bigger sensor.

ASIMount@ZWO

Hi Chen, Kevin and Ben
The changeable guiding rate for AM5 is ready for the next release version of ASIAIR APP since we announced the AM5 firmware V1.3.0(public firmware that supports custom guiding rate) in late May.
It will be soon.

Kevin_A

ASIMount@ZWO thank you! That is excellent news and is greatly appreciated!

KC_Astro_Mutt

ASIMount@ZWO Thank you!

Agepi7o

ASIMount@ZWO Awesome!

w7ay

Kevin_A is there a reason you guide with a 178 vs the standard 174 guide cam?

The 174 pixels are tooo huge for my guide scope plate scales. I would use an ASI174 if I use an OAG, but not on the short focal lengths of my guide scopes. And, I do not use OAGs (I still have a number of them in my closet -- Celestrons, QHYs, Orions, etc; always came back to dedicated guide scopes).

I am thinking of getting a Askar 180pro and maybe a new guide cam with bigger sensor.

I was quite suprised by the Askar 180, Kevin. I was hesitant to try it (got the 180pro to test as wide FOV scope) because it is one full f-stop dimmer than the Borg 55FL that I had been using for ages. But I wanted to see if I can go lighter weight, so I went ahead. The dimmer objective was countermeasured by a larger FOV, and that actually produced a better distribution of star magnitudes for multi-star guiding. With an ASI178MM at 25 dB gain, it never seem to lack guide stars. I probably could dial back to 22 or 23 dB of gain and redo the guide dark frames.

The nice thing about the 180p (unlike the original 180) is that the front knurled focusing ring is already notched for an MXL belt teeth (and it also does not move in and out):

The EAF is mounted directly on the two tapped M4 holes on the 180p's body, so I didn't have to drill the 180p itself, just the EAF shell:

The shell is mounted with some short brass standoffs:

It is solid enough that I use the EAF as a "handle" when mounting the dovetail plate :-).

Here is a size comparison (the Chinese optics is about 1/2 the price of a Borg 55FL with reducer :-):

As usual, helical focusers are close to impossible to focus by hand (and the 180p's focuser is quite stiff too).

Chen

Kevin_A

w7ay yes, now i remember about the smaller pixels of the 178! Awesome! Like a 290mm mini on steroids! Haha! Nice setup! Did you notice any tilt or coma that some people have reported? Maybe they were on the non pro version…. I can’t remember!

w7ay

Kevin_A Did you notice any tilt or coma that some people have reported?

Not on such a smalll sensor as the ASI178.

But, given Askar's typical "quality," I wouldn't doubt one bit that the scope comes with collimation problems even just for an APS sized sensor (remember that I have a tilt on my FRA300?) .

By the way, I rechecked the prices... an Askar FMA300 "pro" comes in at almost exactly 25% (not 1/2) the cost of a Borg 55FL with a reducer :-). I expect teenagers are assembing the Askars, while Borg, Vixen and Takahashi are using some gray haired person who had been with their company for decades :-). You get what you pay for.

Chen

Kevin_A

w7ay haha… you are right about the teenager assembly line. I will look into the 178 as it looks like a 14 bit mini 183. I am not sure why zwo phased out the 290 and replaced it with a 4um 220 but it really does not matter here in canada because every astro product here is out of stock and backordered… every camera, mount and scope. Big supply issues.

w7ay

Kevin_A I will look into the 178 as it looks like a 14 bit mini 183

The 183 would have been even better if it were fast enough. But, at least with ASIAIR guiding, you cannot get 2 FPS from an ASI183MM; that makes it not practical.

I am tempted by a QHY5III-678M with its 2 µm sensor. Should be useful for near-IR guiding too. Using INDIGO of course, instead of ASIAIR. ZWO does not sell a monochome version of the Sony IMX678.

Chen

w7ay

Kevin_A it really does not matter here in canada because every astro product here is out of stock and backordered

Is it a big problem (except for exchange rate) to order from the US? HighPoint is closer to you than they are to me!

If you were in western Canada, I would have recommended CloudBreak, since they are just across the border in Seattle. I usually use CloudBreak or Agena. US Postal System takes less than a week from Agena, and UPS ground is practically overnight from CloudBreak -- order in the morning, and get the part the next morning :-).

Chen

Kevin_A

w7ay how do you figure out if the cameras will actually do 2fps when they all say they do much higher than 2? That has me perplexed! Haha

Kevin_A

w7ay my only issue these days is quality and returns so I don’t usually buy from the US due to the fact that 50% of the gear I have bought recently has to be exchanged due to teenager assembled defects! Haha

w7ay

Kevin_A ow do you figure out if the cameras will actually do 2fps when they all say they do much higher than 2? That has me perplexed!

Just connect to ASIAIR and switch on guide looping with 0.5 second exposure. The ASIAIR struggles to even do 1 FPS with ASI183MM.

Chen

Kevin_A

w7ay at 1s exposures mine jump around from 0.9 to 1.3 and on 0.5s i get 1.9 to 2.3 so it seems my 290 and 120 minis are ok… ish! Haha

Kevin_A

w7ay Chen, it is great news tha zwo will finally add guide rate adjustability in asiair for their own am5 mount soon. My question for you would be… (my slope is 0.16 arcsec/sec) If mine guides on a good night at 0.5 rms using 1.0s guide exposures, 40-45% aggressions and 300-350ms max durations… what would my duration settings be if I reduced my guide rate to 0.25 from the standard 0.5? Would it smooth out the sawtooth more if I also dropped my guide exposure down to 0.5s then adjust my max pulse size?
All using the very limited asiair when they implement the guide rate adjustability.

w7ay

Kevin_A what would my duration settings be if I reduced my guide rate to 0.25 from the standard 0.5?

Double the max pulse durations when you half the guiding rate, of course. But I think your current max pulse durations are way too pessimistic. Even at 0.25x sidereal rate, my RST135 max durations are set to around 150ms to 180ms.

Just think of it as having to drive your car a distance X km (== adjustment in arc seconds) while running at 50 km/hour instead of 100 km/hour (guide rate of 0.25x sidereal instead of 0.5x sidereal). In both cases, you need to go a distance X, so if you travel slower (slower guide rate), it takes a longer time to travel that X km.

One thing you need to make sure is that the pulses are shorter than the exposure time. You don't want the pulses to be so long (when you decrease the guide rate) that it is longer than the exposure time -- the centroid corrections become fubar in that case. I suspect a lot of people who have set the max durations (especially mounts that have to send a declination pulse after an RA pulse, since they are accumulative, to more than the exposure time -- e.g., exposure time of 0.5 seconds, and max pulse rates of 0.25+0.25 seconds -- is suffering from this problem. This is why I have mentioned that not all ZWO mounts will be able to use 0.25x sideral guide rate or slower. And also a primary reason why you want to absolutely limit the pulse durations.

This is why I have maintained that allowing pulse durations to be 2000 ms is utter nonsense when you are guiding at slower than 0.5 second frame rates -- guiding becomes a real mess. And if both RA and declination corrections are sequential (the guide graph shows blue and red pulses for each guide exposure), you really need to limit the max pulses so they don't add to over the time to capture one frame.

I really think that companies are doing a real disservice selling strain wave geared mounts to novices since you need to factor in multiple things to guide well. But, if the mounts are anywhere close to an RST135, they can be successfully guided. You just need to factor in the individual mount's characteristics, not just the brand of the mount. Would you sell a F1 race car to a high school kid? I am sure there are companies that will, just to make money.

Chen

Kevin_A

w7ay i tried setting my mount at 100ms and all i got was the dreaded comb looking pulse pattern when my tracking goes erratic.. I am guiding currently with those numbers only due to fatigue of trying everything without it helping get rid of large sawtooth patterns followed by smooth guiding… rinse and repeat on both axis. So while I know 300ms is overkill for when it is smooth… it helps when it goes crazy and tames it enough to keep my guiding under 0.7rms. No kidding, my guiding is 0.3 for a long time then it goes crazy for 15 seconds then back down to 0.3.

Kevin_A

w7ay i decided on 300ms as it currently is 3x less than my 1s guide exposure.
I tried guiding at 0.5s and even 2.0s and all the same results…. Guides well for 80% of the time then it jumps all around going from 0.25 to 2.0 then back to 0.3 rinse repeat. So i settled on 300.
My calculated pulse requirement is 44ms so i know 300ms is overkill.

Kevin_A

w7ay when they implement guide rate adjustment I will try using a guide rate of 0.25X with a guide exposure of 0.5s, agressions at 40% using 100-200ms pulses as a starting point. That is what I believe it should be but my mount thinks differently! Haha

« Previous Page Next Page »