|
|
|
|
|
There is no error message. CCDciel just gone without a trace of any error messages. Once 0.9.92 is running fine, I will run 0.9.93 again in the debugger. |
|
|
I make many test but I cannot reproduce the crash. One of the change between this two version is the plot of the star mark is now done in pu_main DrawGuideImage(). This is to improve the performance by drawing the guide image only after the guiding is complete and the next exposure is started. For that in cu_autoguider_internal it store the position and size to plot in an array. This was not checking the array range explicitly, and I fix that now, even there is very little chance to reach the maxstars limit. https://github.com/pchev/ccdciel/commit/42bef02772e793fb514740fd894ac48ca9d28631 |
|
|
Thanks for looking into it. Here the guider typically finds only a few stars. Next time I will run CCDCiel in the debugger. The only other thing observed here, is that the guider USB link has become unreliable probably due to a poor USB connector contact. Maybe this is causing the crash. |
|
|
Tested this night the following: CCDCiel 2025/08/19, same problem crashes random when guider is active CCDciel latest code in debugger, Crashes inside debugger, Not clear where in code. See attached screenshot. At this moment, I'm pretty confident the guider driver is to blame. Setting USB speed slower and USB traffic seems to help but not fully. I'm surpriced the driver can cause this but it looks like. A better shorted USB connection may help. So not a problem for CCDCiel and this issue can be closed. Untitled1.png (3,629 bytes) |
|
|
The error code is "status heap corruption" . This can be from the driver outside of a try except block? I will look what change in the Ascom camera interface between 92 and 93. |
|
|
This morning I run the program in the debugger again. I only connected the guider camera and started looping of the guider camera. Simular exception. So the new Win10 QHY5II driver combined with a poor USB connection results in these crashes. It would be nice if CCDCiel could handle this with just an exception. Attached a screenshot and the last three lines of the several logs from last night. My previous Win7 QHY5II camera driver could handle this better. Unfortunally I can't use it in Win10. last lines of logs.txt (1,929 bytes) Untitled.png (32,409 bytes) |
|
|
Can you try to connect this camera as the main camera, loop and change binning. Do it crash in this case? This is to be sure this is from the driver and not from the guider code. |
|
|
The QHY5 as main camera has now been looping for more then two hours without an error. This hints that the main camera option could work more reliable, but it is too early to tell. I will continue tomorrow. Unplugging the USB and replugging helps with making the connection but likely only temperary. More tomorrow. |
|
|
During the day I had three times an error while looping the QHY5 as main camera. Looping stopped and could be restarted. No crash. During the nightwith the GHY5II as guider, I had a about 6 crashes. After that I had enough and replaced the cable by a spare one. System is now running smooth for one hour now. So I have the impression that connecting the QHY5II as main camera works more reliable then as guider. Is the software very different and if so can the guider part be made simular as main camera with some tweaks? |
|
|
All the ASCOM code in cu_ascomcamera.pas is exactly the same for the main camera and the guider. This is only after the reception of the image data that the code in onNewImage property is different. Do it work better with the new cable? |
|
|
Yes it works reliable with the new cable. If the software is the same, then there is nothing to improve and this issue can be closed. |
|
|
Ok, I close it now. I also search more about the "status heap corruption" exit code, it look like this not raise an exception, so this cannot be trapped in a try...except. As the error is most probably in the driver after the USB error there is nothing that can be done in CCDciel. |
- Anonymous
