Hi,

i'm running into quite some frustration with this.
I was starting up Stellarmate (1.7.4) connected via VNC, connected devices, configured a job, started, slewed, started to take shots, activated guiding...
Suddenly the exposure after starting phd2 didn't download, no more exposures to be taken. Quit the running job, tried again, failed again.

Then i discovered that the USB camera drivers (ZWO ASI294MC and ASI120mini) were showing strange behaviour on profile startup.
Updated to 1.7.5. no errors. Reboot - same story

Now, the 120mini didn't get recognized in INDI panel, so i tried some options, found one that worked: ASI294 connected to USB2, oddly enough, with 120mini connected to USB2 hub on the 294MC. So far so good, but the optical train data is reset, so i set it according to my setup, suddenly i can't take pictures with the 294MC selected as main camera (EDIT: Everything in EKOS camera tab is greyed out). So either i can't shoot, or only one of both cameras are recognized and connected.
Seems like 1.7.5 is messed up with the USB3 connectivity somehow? It worked before flawlessly. The external PSU of the 294MC is working. Whatever i was trying, in the freezing cold, didn't work.
Errors shown were "could not connect", "no camera found" and "could not connect while already connecting".

Any known issues that fit this behaviour?

Thanks,
M

Hi,

Which Zwo driver are you using? Zwo CCD ? Zwo Camera 1 ? A mix of both ?

I think I recall that when using 2 Zwo cameras you need to stick to Zwo CCD only.

Just a guess, trying to help. Might not be that.

s.

Thanks for your input, Steve.
I tried both and mixed, but when mixing ZWO-CCD with ZWO-Camera*, the device assigned to ZWO-Camera was showing up twice in INDI control panel. while one of them wasn't able to function. Originally i had ZWO-Camera drivers set to both cameras, and it worked a charm until this incident.
If i get both cameras to connect, by using USB2 port for the 294MC and the on-camera hub to connect the 120mini, it worked until i started guding with PHD2. Images don't get downloaded from the camera.
When using ZWO-CCD for both devices, PHD2 tried to guide usiing the 294MC, but i can only select the 120mini in optical trains setup (which is cleared and empty) as main camera.
Seems like an USB enumeration issue (Kernel?) but i could not figure out how to reset USB enumeration in the OS.
In Windows, for example, it would be done by removing devices from Device Manager and reboot to recognize and map these USB devices.

I'm stuck. Didn't change anything before doing the 1.7.5 upgrade. I might try with a backup sd-card (1.7.4) tonight, to rule out hardware problems.
Giving ASCOM/Alapaca and NINA a try if this won't work, but i love Stellarmate for various reasons over other solutions and got used to it, so i would like to continue with it.

Have you looked the devices mappings in the Web manager? Maybe you have old mappings that screw up things.

I think i got it. That's the good part. The bad part is: It's not fixable :/

It's all about poor Linux compatibility of the ASI 120MM mini. That's why all the problems start when guiding is started.
I found loads of threads about the problem, which can be (unrealiably) fixed with a firmware update, but there's no matching firmware for the mini version of the ASI120 series. It's working on Raspi3 hardware, but i am running a Pi4 with stellarmate, so i will have to replace this one to achieve stability under Linux. There's a S version, which is USB3 capable, and most of the Linux users are happy, but still some are reporting problems.

I will go the safe route and replace it by a compatible model/brand, i hope the dealer will take back the 120MM mini for a refund.
The strange thing with the mini is, it was working over two sessions, but PHD2 was throwing some strange warnings and it was hard to get a lock on stars in some cases.

So, sorry ZWO, but i have to warn every reader of this thread of using the ASI120 models with Linux.
The root problem seems to be variable packet sizes sent over the USB, which is rated as security violation by the Linux kernel. The effects seem to vary over hardware implementations (USB chipsets) and Kernel versions.

I'm currently running a 90x30s exposure session without problems after removing the guiding cam from the bus and the EKOS/INDI profile.
Thanks for trying to help.

As far as I know the issue you are describing is historical from around three years ago……
I have never had any issues with my two ASI120MM minis bought in 2020 and 2022.

Interesting.
I have no idea where to look further. As long as the guiding cam is connected, there are problems. Either i can't take exposures with the main camera, my mount device gets thrown out of EKOS/INDI, or it shows up doubled, one time with limited settings and unable to connect...
What is your configuration for the cameras in your EKOS profile? ZWO-camera-X for each, or ZWO-CCD for both?
Did you connect your 120mm mini directly to USB2 or via the main camera hub (with the former connected to USB2 instead of USB3)?

Full disclosure: I use Astroberry which is still 32bit and uses KStars 3.6.0 and Indilib 1.9.7. Both rigs have the ASI120MM mini plugged through the main cameras, ASI533MC and ASI2600MC, each of which is connected with good quality USB3.0 cables into USB3.0 ports on the two RPi4s 4GB. In Ekos profiles I use ZWO cameras 1 & 2. Both rigs have ZWO EAF focusers also plugged through the main camera hubs. One thought - I never use the 2.4 GHz WiFi band to communicate since there is evidence of interference between USB3.0 ports and the 2.4GHz band on the RPi4.

OK, based on your info i tried again, but i plugged the main cam into the USB2 port, connected the 120mm mini to that.
I set ZWO-CCD in EKOS for only the primary camera. I had set it to camera1 and camera2 respectively before. As i am inside i defined the mount as "simulator". After starting the profile both cameras popped up in INDI panel as connected.
PHD2 started up and refused to connect to INDI 120mm mini (timed out), so i changed that to ZWO-ASI in the settings and PHD2 successfully connected to the guiding cam.
Then i set up the scope to capture a few dozens of flats, which worked without problems.
SO far so good, tomorrow i will do a live test with a real mount, real stars and so on.

I'm desperate to try the same setup using an USB3 port if the above works well.

Thanks for all the suggestions and info, i will keep this thread updated.

EDIT: At this point, i can only pin the source of the problem down to indilib 2.0.0
Maybe that's why it works flawlessly (as it did) with your configuration.

Well, it's working again.
No idea why it failed after dist-upgrading and incorporating indilib 2.0.0, but all i did was removing the guide cam and after everything worked using this setup with only ZWO-CCD set for camera1 in EKOS/INDI profile. After connecting the guide cam again, it was autodetected by INDI, PHD2 behaved well. I just did a 90x30seconds star-guided exposure on C-2022-E3(ZTF), everything works OK.

Thanks again for your input,
M

Good news!
Clear skies…..

Happy for you