Byrdsfan1948 I guess at this point in my life I'd rather spend a bit more dollars as in the RST 135 , but know what I get is what I expected to get
From the PE curve that ZWO has publish for the AM5, I already see problems that they will potentially face. I have no idea if they are aware of it themselves.
As you said, PE can always be autoguided down. Very few mounts don't need to be autoguided, and once you need to autoguide anyway, the question is can you really get the total error down low enough through the feedback loop
One of the reasons the RST-135 can be corrected is that the PE curve is very close to sinusoidal. There is only a small third harmonic term -- what the hobbyists call a "smooth PE curve;" hobbyists don't talk Calculus :-). When autoguiding, you are chasing two things -- the PE from the gears and the measurement error caused by atmospheric turbulence. The first term requires you to be able to keep up with the first time derivative of the PE curve. So you need as fast a feedback loop as possible so that the mount has not erred by more than 0.25 (or whatever) arc seconds in delta-T.
In the past, the second term can only be solved by using very long guide exposures, often 2 to 4 seconds. So these two are conflicting specs.
Fortunately, if you assume ergodicity, exposing for 4 seconds is exactly equivalent to averaging the exposure of 4 stars with the same SNR, each at 1 second. Thus we can get the measurement error down by using centroids of multiple stars.
Unfortunately, in the past (until ASIAIR v1.8), exposing for shorter time did not increase the sampling rate. You could not get autoguiding on ASIAIR to be faster than 1 FPS, even if you cut the exposure time to 0.5 second. The guys and gals running raw PHD2 were ruthlessless beating out the people running ASIAIR. I had pleaded directly with ZWO for faster updates for 2 years now, especially after their chief bottle washer whispered in my ear that they were planning a harmonic drive mount about a year ago. I had told him that the only way to tame a harmonic drive mount is to use very small MinMo (they did that in v1.7) and very fast frame rates. They allowed 2x binning in v1.8, and we finally could get 2 FPS guiding.
The RST-135 has a worst-case specs of about 70 arc seconds peak to peak error. Since is is pretty much sinusoidal, that is equivalent to 70 * 0.5 * 0.707 arc second RMS. I.e., 25 arc second RMS error in the RA axis.
With multi-star guiding, and finally 2 FPS autoguiding rate, I am finally getting a total (RA plus declination) RMS error of around 0.35" RMS. So, reducing error from 25" to 0.35" is possible now on ASIAIR.
Now, all this is possible because the RST-135's PE curve is smooth (i.e., the time derivative is also sinusoidal).
Go take a look at the curve that ZWO published for the AM5. Although their PE is smaller than the RST-135 (the RST-135 number I gave is the worst case; I don't know if ZWO's curve are also also worst case, or one of their selected prototypes), if you try to take the time derivative of their PE curve, you get a lot of large peaks. Those peaks are going to be tough for autoguiding to keep up.
Anyway, lets hope that this time they do some mathematical analysis before pushing the results on their "officers." Won't be me, I can assure you of that.
Speaking of loose specifications, Cloud Break Optics has already changed the documentation on their web site -- the ASIAIR "Plus" is not longer spec'ed to work to 15 volts on their web site. They didn't mind that I created an open-box situation, since they are already using the returned ASIAIR as a training tool for a new employee.
BTW, the harmonic drive mount market is getting hot. Used to be only Hobym and RainbowAstro produced them (both Korean companies, and their CEOs used to be partners in a joint venture). Followed by a hand made mount from a Japanese boutique company called SS-One (caters to Takahashi owners).
https://www.iceinspace.com.au/forum/attachment.php?attachmentid=240738&d=1550715062
SharpStar (the company that produces the SharpStar telescopes, and rebranded into names like Radian Raptor) has released something called the Mark-III. And now we have the AM5 into the fray. (Don't be surprised if Radian also rebrands the SharpStar as a Radian Mount "designed by" a YouTube shill :-).
Notice that the SS-One is a single armed fork. So, like the Avalon M-zero, does not need to Meridian flip for most declination angles and instrumentation. Only harmonic drive fork mount that I am aware of.
Chen