PE cycle of 432 secs?

Noommuk

w7ay Do you know what’s the possible cause of this? Any change in reduction ratio for newer version?

w7ay

Noommuk Do you know what’s the possible cause of this? Any change in reduction ratio for newer version?

No, I have no idea why a manufacturer would change the internal specifications of a mount less a year after releasing the product; and on top of that don't change the model or version number.

Caveat emptor.

Chen

mikenoname

They have indicated in past posts that they were shopping for a new strain wave gear manufacturer. Maybe this prompted the change?

Noommuk

mikenoname Umm, possible. Do you have the link for that post?

mikenoname

Noommuk

https://bbs.astronomy-imaging-camera.com/d/15894-huge-difference-in-ra-guiding-on-different-sides-of-flip/29

mikenoname

Noommuk Umm, possible. Do you have the link for that post?

https://bbs.astronomy-imaging-camera.com/d/15894-huge-difference-in-ra-guiding-on-different-sides-of-flip/29

IlyaZ

My AM5 has a similar cycle. Produced in February 2023

jalvarez_bcn

I got a new AM5 produced in Feb 1st, it seems to have a similar slope too.
I didn't still have the opportunity to test it because weather is being cloudly.
@w7ay : Do you mean that the batch of AM5 with this slopes will have more PE error that the previous batchs?

w7ay

jalvarez_bcn Do you mean that the batch of AM5 with this slopes will have more PE error that the previous batchs?

If all else were equal, reducing the period by a factor of x will also increase the slope by the same factor. We don't know if this batch of gears have better harmonic distortions. If the percentage harmonic distortion were the same, then yes, guiding will be tougher than it already is.

Your scan does show a high harmonic distortion content, but we can't depend on just looking; you need to measure it to find what the slope is like (and this time, using a period of 288 seconds instead of 430.82 seconds as the fundamental period).

Chen

isaakPr

I tested two AM5 mounts, one with period of 432s and another with 288s. My first AM5 mount wasn't really bad (usually around 0.6 rms), but sometimes it behaved strangely in AR. Suddenly, the guiding error would increase to 1 rms, randomly.

I changed it for another one that was more recently manufactured (February 2023). It has a period of 288 seconds, however, it guides better than the previous one (0.5 rms) and is more consistent. It easily goes below 0.5 rms, sometimes as low as 0.3.

The only complaint I have is perhaps the high backlash, although it's usually not a issue. Only in high DEC dithers (5-6 px, 7-8 arcsec) can it become a problem. I'm thinking that perhaps it would be better to limit the dither to only 2 or 3 pixels.

AstroHolger

I also have just bought one of the newer AM5 with a PE cycle of 288 sec (1.2 degrees). One thing I do not understand: Looking at the first diagram, counting the number of cycles from e.g. 0-30 degrees, it should be around 17 cycles for the "old" mounts with 1.8 degree/cycle and exactly 25 cycles for the newer mounts (= 30°/1.2°). However, the diagrams of the newer mounts all still show 17 cycles per 30 degrees. What am I missing?

w7ay

AstroHolger However, the diagrams of the newer mounts all still show 17 cycles per 30 degrees. What am I missing?

I would trust your own measurements. Turn on autoguiding but set the max RA pulse duration to zero, and watch the RA error graph trace the periodic error curve.

If you are one of the unfortunate souls who use an ASIAIR, the PE will be larger than the guide graph scale, so you just need to wait for it to reappear. You also cannot turn the max RA pulse duration to 0 with an ASIAIR, so set it to 1.

Use a stop watch to time the major zero crossings. You can also use the data that PHD2 has collected to do an FFT to find the fundamental period. It should be either 288 seconds, or 430 seconds. But even a simple watch should have enough resolution to distinguish between a 288 second period and a 430 second period.

Chen

AstroHolger

Thank you for this advice. I will try it, whenever there is a glimpse of clear sky (which hasn't been here for months). And no, I don't want to support proprietary products - I am using Nina/PHD2, so at least this should work.

Holger

AstroHolger

@w7ay : Yesterday I could now test my new AM5. PHD2 clearly shows 288s, both in Guiding Assistant and FFT:
Harmonics are not too bad. Nevertheless, there is still room for improvement regarding guiding:

Even reducing exposure time of the guidecam to 0.5 seconds did not help. I have to admit, that seeing was not great.

phd2-guidelog-2023-05-21-234929.txt

188kB

However, my main point seems to be proven now: the "Full Periodic Error" graph from ZWO is just wrong, as it shows 1.8 degrees per cycle instead of 1.2 degrees (or 17 cycles per 30 degrees instead of 25):

I guess the graph has wrong x-axis values and does NOT show full 360 degrees, only 240.
@ZWO: Any comment?

Holger

w7ay

AstroHolger I guess the graph has wrong x-axis values and does NOT show full 360 degrees, only 240.

Hi Holger,

Pieces of paper like that are useless, since ZWO does not even document what they mean. Reminds me of politicians who patch together some technical buzz words to show that they are technically literate.

I suspect that the 360º is the plot of the input angle of the strain wave gear, not their mount, which has other gears and belts and thus other ratios.

Further proof of that is notice that in the first chart on their piece of paper, the start and the end don't even match. I.e., it is not periodic with a period of 360º.

I suspect that what they are measuring is 360º of the input of the strain wave gear shaft. Perhaps ZWO is unaware of is that the full repetition of a strain wave gear (see Harmonic Drive's patent, and also that NASA paper that I have given links to) is two input rotations of the gear, not one rotation!

They are showing 360º. Why? They might as well show 271.3356º and it would be just as useful (or just as useless). But I would love to hear what their explanation is.

Sorry that you received a 288 second mount, Holger.

Chen

tempus

AstroHolger Harmonics are not too bad.
w7ay Sorry that you received a 288 second mount, Holger.

Is that graph the proper way to know if there are secondary harmonics?

Kevin_A

AstroHolger i was shocked when i saw the dec max duration of 8 seconds! Even Ra is way too much.

AstroHolger

So true...providing these graphs with each mount makes it look like ZWO having a good understanding of the underlying technical correlations and a considerable QC. But as you stated often enough in this forum, they are looking at the wrong figures. Now we have new mounts with lower PE error peak-to-peak, but worse slopes.
I am curious, to which guiding figures I can trim my 288 second mount.

Holger

w7ay

tempus Is that graph the proper way to know if there are secondary harmonics?

What is a "secondary harmonic?" Any Google page that shows the meaning of that so we can answer?

Chen

tempus

w7ay What is a "secondary harmonic?" Any Google page that shows the meaning of that so we can answer?

That wasn't my question at all. I was asking which tool is the best to visualize those secondary harmonics. But if you prefer to take a condescending tone, never mind.