Increase cpu load causing mount disconnect?

[Beepa]

Hi,I've struggled with Celestron mount disconnects for years. Have added dedicated router, minimized interference, circuit board heater, etc but unrecoverable disconnects still occur typically within a few hours of running my Sharpcap sequence (v4.1.12802 64bit) on Windows 11. In trying to understand, I used a Dell & HP laptop and ran the very simple program shown using no mount connection but using Camera V3 simulator. Sharpcap starts with 2 to 4% of cpu usage on both laptops. But after 13 hours cpu usage was at 25 to 35% with this very simple program (factor of 10 increase). Is it possible that this trend is the cause of my disconnects after a few hours? Is there any possible mitigation?

[Beepa]

Program I was using: (Don't know if downloaded properly on prior message.)
SEQUENCE
ARRANGE FILES IN ONE FOLDER
LOOP 2000 TIMES
DELAY 5
LOOP 3 TIMES
PLAY SOUND Alert
LIVE STACK FOR 50 FRAMES SAVING AS AllFormats COUNTING Stacked FRAMES
DELAY 55
END LOOP
DELAY 5
LOG "So far so good"
DELAY 30
END LOOP
END SEQUENCE

[Beepa]

Also Camera V3 simulator was running with 4-second exposure and each stacked fit file showed 200sec total exposure.

[admin]

Hi,

well, it's hard to be sure about the CPU usage and the disconnects, but it's certainly not *supposed* to be like that with the CPU usage rising over time. I will run through the same sequence and see if I can catch the same problem and work out what is going on - I have a few suspicions about what typically can cause problems like this, but might as well be sure.

Thanks for the clear report!

Robin

[kaymann]

Two things come to mind with Celestron mounts:

1) Every time I have an issue with my mount I would truck on down to Starizona with my mount in hand. First thing they would do every single time is update/reload the firmware on RA/DEC/HC and every single time the problem was resolved.

2) I purchased a new power supple from them ($30) and no more need to reload the firmware as flaky power was causing disconnects and firmware errors...

YMMV and just something to consider...

[Beepa]

Hi Kaymann,
Thank you for your observations. I have run my system with 3 different chargers and I seem to get disconnects with each one. I keep them plugged in to the mount and it normally reports it is fully charged. I don't use the hand controller but use the wifi direct.

And Robin, I'm delighted you're taking a look at the increase in the cpu load because that does seem to correlate with my delayed disconnections. Thank you for your efforts. Also, anecdotally, the mount appears to stay connected somewhat longer with my Dell which is consistent with it's cpu usage increasing a bit more slowly. If this work doesn't resolve the issue, I'll take a closer look at the power supply experience.

FYI, I'm running my scope with an ASI 294MC pro camera using a Sharpcap sequence for all night automated photography. I use live stacking with plate solve reorientation and auto refocus every 40min or so on alt-az scope.

Best Regards,
Dave

[admin]

Hi Dave,

I trimmed your sequence down to remove the delays (since I thought it unlikely that they would be part of the problem) and left it running for several hours yesterday evening. By the end, CPU usage had indeed increased. Fortunately I'd planned ahead and started the run with a code profiling application pre-set-up ready to find out what was using more CPU.

The result was unexpected... Some code that was designed to make the movement of the frame progress bar at the bottom right smooth (rather than it progressing in steps a few times per second) was adding an animation to make the smooth motion happen, and over time those animations were building up and using processing time to calculate, rather than the ones that were finished with being cleared. I have now made a small change that should mean that the old animations are fully cleared when a new one is applied, and I think it will fix the issue. Look out for an update containing this next Monday.

Of course, the disconnects may be coincidental, but it's definitely a bug fixed on the SharpCap side

cheers,

Robin

[KenHarpster]

I am so glad that I just found this thread tonight, I will sleep better now that I can stop trying to solve the issue. I noticed an increase in CPU usage with SharpCap as my imaging session progressed tonight. Initially everything ran as it should, but, as the hours ticked by things started disconnecting (RDP, mount, focuser). I am very glad that Robin is so incredibly responsive to the users of SharpCap. I will be waiting for the update, but by then my testing will have to be indoors as the clouds once again roll in...

Wanted to add that after several hours of imaging, not only did devices start to disconnect, but SharpCap froze up (not the PC) a few times (standard Windows message ...not responding, with wait or close options), I had to close SharpCap and restart it.

[admin]

Hi Ken,

if we're lucky then the issues you hit will be the same as this one and all will be fine. However, there are other possible reasons that SharpCap could lock up, so if you can find the log files from the crashed session(s) and share them here, I will look to see if there is anything inside them that points to a cause. Normally you can spot the right log file - it will be quite large from a long session, the file name will reflect the date/time that SharpCap was launched for that session and the file timestamp will reflect the time that SharpCap exited (or crashed).

cheers,

Robin

[admin]

Hi,

the version with the potential fix for the CPU usage increase is now available to download.

cheers,

Robin