I fixed the SGP API 'NO_ERROR' issue! And share my approach ^_^

Normally, PHD2 connects to the SGP API local TCP service via port 59590 and maintains only one connection (‘netstat /an’ command):

PHD2 error message occurs when an unexpected condition occurs:

PHD2 repeatedly attempts to connect, will establish multiple connections, generally can have 3 of connections. But at this point, it is clear that the SGP API is in some state and cannot respond to PHD2 requests.

After PHD2 attempts to communicate repeatedly, stop trying, disconnect the camera, and system will automatically release the invalid TCP connection:

When the system releases the failed TCP connection, the SGP error message box appears: (The problem especially repeats when waiting meridian flip. )

In this, I suspect that due to the limited processing capacity of the computer, cause the coordination problem between SGP API and PHD2. So I managed to mitigate the performance load of the SGP API. Miracles happened, SGP and PHD2 resumed the tacit agreement, and continued to work stably for several evenings, without appearing an error message box.

Below, I share my approach: enable PHD2 's “Use Subframes”, only download the search area of the range of guided star images, let the SGP API lighten the burden. and increase “Time Lapse”, 100MS, prevent SGP API unexpected error:

I tested it for a couple of nights, and both options have a relationship. The system works fine as long as they are selected.

My English is not good, with translation software. I hope the above information will help friends who have the same problem. Good luck!

this is strange, i have very underpowered hardware and almost never see this problem. i have my timeout set to 300s though. it happens to me maybe once every 2-3 months.

rob

Hi Rob,

My computer: Pentium Dual-Core E5200@2.5G.
According to my PHD2 settings, guiding never raise error, but when taking Bias photos SGP will report error.

Yang
2017-10-29

Last night, I pause a sequence to waiting the moon down, PHD2 was still guiding, ‘NO_ERROR’ error happened.
I guess SGP API has some issue or not robust enough when waiting meridian flip or time constraints.

what firmware are you running? this is an STT-8300M + FW8G right? as i mentioned i occasionally see a guider camera timeout but it is very, very infrequent. it sounds like yours is failing all the time.

i have not upgraded my firmware in years, so it’s possible one of the newer camera firmwares causes this problem?

rob

Yes, Rob. I’m using STT-8300M + FW8G, and I used to upgrade the camera to the newest firmware. Just like you said, ‘it’s possible one of the newer camera firmwares causes this problem’.
Addition, I’m curious that have you ever taken a try to set the ‘Time Laps’ to 0ms and don’t check ‘Use Subframes’, what will happen?

well i have never used subframes… so i’m always downloading a full frame. i also have time lapse set to 0 already.

disconnect on non-responsive camera is set to 300 seconds (!). that seems to have fixed the problem for me.

rob

Well done!