galactique I bought an adapter following his advice and it works reliably, like a dream
By the way, I can confirm that the problem is how ZWO had implemented WiFi.
I have been writing some code to use INDIGO Sky; it runs in a Raspberry Pi 4 enclosed in one of these ubiquitous Amazon metal boxes. The WiFi has been solid for weeks on end (never had to reboot the INDIGO Sky in a month or two). That being said, when I switched to Ethernet, I do get better speeds even with the INDIGO Sky -- getting upwards of 80 MB/s on it.
There is also some bad bug or poor implementation with the guide camera in the ASIAIR. If you recall, I had to use Bin2 on an ASI290 (USB-3 version) to get better than 1 FPS guide rate (needed because d/dt of a the periodic error of my RST-135 is extremely large). With Bin2, I could get 2 FPS when the exposure of the ASI290 is set to 0.5 seconds. But Bin2 has its associated problems too.
With INDIGO Sky, I am getting a sustained frame rate from the same ASI290 of about 9 FPS, at bin1 and going through an Ethernet connection on a Raspberry Pi 4, which the ASIAIR does not have to go through since autoguiding is done locally in the ASIAIR box. There is a factor of 8 in there, using the same ARM processor, and using Ethernet as the system bus.
I cannot guide at 9 FPS of course, the exposure duration would be ridiculously short to capture enough guide stars. But it allows me to start playing with guide algorithms at even 3 or 4 FPS. (I am contemplating using a feedback loop that synthesizes a rough approximation of the periodic error curve on the fly, and be able to do some predictive feedback, so even 1 FPS could be enough -- I hope I don't have to use something as sophisticated as a Kalman filter, but will do so if I have to -- make good use of the class on Kalman filters that I took some 50 years ago :-).
The Ethernet speed should allow me to implement the guide algorithms inside a Mac indoors (my primary computer is a Mac Studio Ultra with 20 cores), with the INDIGO Sky outdoors.
Chen
