Jan75 min_move at 0.2

That simply means that ASIAIR will not issue any correction pulses until the error has exceeded 0.2" (really a bad idea with mounts that have large PE slope to start with).

I can't see offhand how it is going to help the 2.9s spike case, where your spike's amplitude is already larger than 0.2". Are they really studying your problem, or simply conflating your problem with some one else's problem?

I also bought a counter-weight shaft and will try to attach a small CW to the mount to see if that alleviates the oscillations.

Don't completely balance it, though. You will find that a slight imbalance with actually help the meshing of the gears; even though a strain wave gear already have lots of teeth that are meshed -- that is why they have so much torque and so little backlash.

But certainly does not hurt to try.

Good luck.

Chen

Kevin_A
Hi,

Long time huh!!..
I already calculate it, Its = 0.141"/sec.
So, Could you tell me please what to do with this number? Rather than I search all this long thread
I am using PHD2. WO scope FL=200, ASI290MM mini.
Thanks

MrAstro each mount is different. That graph that everybody gets with their mount is not representative of the complete full period length, it is just the max PE, not max slope necessarily, so all calculations made will not be exact… so you just use it numbers to calculate as a baseline settings average. From there you try less aggressions and less durations. If that does not help then start at the baseline and try more durration etc. That is why the baseline is a good guess to start at and a place to start based on the part of the period graph you do have numbers to calculate from.

@MrAstro Kevin_A From there you try less aggressions and less durations.

Actually longer "max durations," since the true worst case slope is probably steeper than the one found from the expanded graph on that piece of paper.

Just increase the max duration if you see a string of flat-topped correction pulses in a row on the guide graph (indicating that the pulses are not long enough to correct), if the flat topping is not caused by wind gusts.

You probably don't want to correct for wind gusts, since the guiding will undershoot the opposite direction after the gust. And if aggressiveness is set too high, leads to oscillations.

The worst case slope in that graph is simply the best that you can expect from your particular sample of the mount.

Chen

CHriss what did the star mass look like? Was it repeating like that? Sooo many variables can cause issues. Hard to say. Looks to me that the max durations were not enough to recover as I see a flat comb shape but huge over corrections from something too.

Kevin_A ... I see a flat comb shape but huge over corrections from something too.

Hmmm, where did you see that, Kevin? I was looking for correction pulses, and did not find any (my browser may not be showing blue that prominently either) -- I was about to ask him to turn on the correction pulses the next time he see this oscillation. What I had been calling a "comb" refers to a comb of saturated correction pulses (i.e., max duration pulses sent in a long sequence and still not sufficient to correct the mount's slope error), and not to a constant amplitude oscillation of the RA error itself.

@CHriss

When you ask for recommendations from folks who are not at your computer, please include all autoguiding information that you can. I suspect, for example that your problem comes from an RA Aggressiveness setting (loop gain) that is set too high, and that caused the RA pulses to alternate in sign (i.e., oscillation caused by the autoguider software, and not the mount). However, you did not show the aggressiveness setting for RA, nor turned on correction pulses on your chart -- so there is no way we can tell with even 1% certainty what is happening, except that we see abnormal rapid oscillations (that is usually associated with applying too long a pulse to the mount, together with aggressive numbers -- basically the loop gain, if you are familiar with feedback loops -- that are too large). Actually, I am also surmising that the oscillations are rapid -- I was looking for the abscissa scale and didn't find that either.

The other important thing -- you will find that in general, strain wave gear mount autoguides about as well as a worm gear mount that is 1/3 the price of the strain wave mount. It is just the nature of the beast -- the market for strain wave gears are the people who are willing to pay the extra for portability (no need to lug around heavy counterweights, or for that matter the weight of the mount), and no need to worry about balancing (especially not having to worry about "third axis" balancing). Part of the market is for old folks like me who can't handle weights, even though I don't take any telescope out of the home anymore.

If you had bought these mounts on the premise (or promise) that they can autoguide as well as worm geared mounts, that is false advertising, and you should remedy that.

I have been using strain wave geared mounts for some 4 years now and still slowing trying to get them to work to my satisfaction (part of it due to my OCD -- most people would be happy with a 0.5" error). At least I have gotten mine to now guide to the 0.35" region by applying everything that I know about the behavior of these mounts, and the autoguiding software.

So, patience is required to learn your mount (each mount is so different from another one from the same manufacturer) that you can't just depend on what other people use to tune their autoguiding parameters.

There is a large variance in the sample you get (luck of the draw). Just because I can get 0.35" from my RST-135 does not mean that a different owner can get the same performance -- his/hers might be better, or it might be worse. But it appears like ZWO mounts have much larger variance (bigger lottery) from one sample to another.

You need to be patient (just like other aspects of this hobby) and slowly learn your own mount. Or simply buy a nice traditional mount. An Avalon M-zero for example will wipe the floor with these strain wave mounts when it comes to configurability and autoguiding performance.

Chen

Kevin_A #p69123 Hmmm, where did you see that, Kevin?

Oh, I see it now! My apologies. (I had to magnify and altered colors. 75 year od eyes that went through cataract surgery :-)

The pulses are indeed maxed out, but only after the oscillation has started. Since the amplitude of the oscillation are large (12 arcsecond region), any pulse durations would naturally be maxed out when trying to correct it.

But there is a latency too. Notice that it takes a couple of pulses before the oscillation changes sign again.

(With digital feedback loops, latency is a killer when it comes to introducing oscillations -- kinda like phase error in analog feedback loops.)

Looks like it takes three pulses to correct 12 arcseconds. Assuming those are 1 second pulses, each pulse is some 4 or so arcseconds (0.5x sidereal rate would be 7.5 arc seconds). Thus Max RA durations appears to be large.

@CHriss

Check your RA aggressiveness setting. Start at 0.15 (15% in ASIAIR) and only increase it slowly (and try not to increase it to more than 0.5) only when there is insufficient correction.

Are you using a Max RA duration of more than 1000 ms or a guide exposure that is longer than 1 second? The pulses appears to indicate that you perhaps did.

If so, start at a Max RA Duration of 100 ms. Increase it only if you see long combs of the correction pulses, and the RA is slowly creeping off the chart.

Then make sure that you are using guide exposures of 0.5 seconds (and checking that your guide camera can keep up with a 2 FPS frame rate). That will make corrections more rapid. That implies that you will need to use multi-star guiding (or the "seeing" will eat you alive).

However, a 0.5 second guide exposure may mean that you need to upgrade your guide scope to get solidly a dozen stars when picking the multiple stars.

If you are happy with a 1" type guiding, a 1 second guide exposure should be sufficient -- but that will still require you to use multi-star guiding with at least 6 stars.

Chen