×

INDI Library v2.0.6 is Released (02 Feb 2024)

Bi-monthly release with minor bug fixes and improvements

AZ-GTI in AZ-mode tracking problems

I did some calculations in this spreadsheet and they differ from your values above. Btw, microstep rate is inversely proportional to the offset. It's actually the T1 hardware timer value, so large values mean the timer will trigger less and the stepper speed would be slower.
11 months 3 weeks ago #91135

Please Log in or Create an account to join the conversation.

  • Posts: 80
  • Thank you received: 11
I can see the same tendency in your spreadsheet (perhaps i made a mistake while copy/pasting the lines). But what i see now is, that my theory cant be right, because the speed would be a bit higher with every step with aggr 0,6 and that doesnt happen. it just stabilized like without the factor (or with factor 1) and that i cant understand :unsure:
Last edit: 11 months 3 weeks ago by Mat. Reason: typos
11 months 3 weeks ago #91137

Please Log in or Create an account to join the conversation.

  • Posts: 80
  • Thank you received: 11
I made another test, i think the implementation of the track factor or aggressiveness was a very good idea (could solve the problem if th mount goes a little bit to fast), but the implementation will not work, by simply changing the trackrate:

if you look at this records the only time the speed changes is with every change of the factor (while running) -> see green values and how the offset changes after setting a new factor. It changes ONE time, and than speed is back und the offset is on a new level:

AXIS1 offset (28) microsteps rate (571428) tracking rate (514285) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (15) microsteps rate (1066666) tracking rate (1011428) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (17) microsteps rate (941176) tracking rate (923344) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (18) microsteps rate (888888) tracking rate (881634) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (17) microsteps rate (941176) tracking rate (930326) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (17) microsteps rate (941176) tracking rate (932279) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (18) microsteps rate (888888) tracking rate (886554) direction (0) aggressiveness (1.00) hysteresis (0.10)
...
AXIS1 offset (17) microsteps rate (941176) tracking rate (937989) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (18) microsteps rate (888888) tracking rate (891169) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (17) microsteps rate (941176) tracking rate (938146) direction (0) aggressiveness (1.00) hysteresis (0.10)

CHANGED AGGR TO 0.80

AXIS1 offset (17) microsteps rate (941176) tracking rate (768840) direction (0) aggressiveness (0.80) hysteresis (0.10)
AXIS1 offset (13) microsteps rate (1230769) tracking rate (976817) direction (0) aggressiveness (0.80) hysteresis (0.10)
AXIS1 offset (15) microsteps rate (1066666) tracking rate (858913) direction (0) aggressiveness (0.80) hysteresis (0.10)
AXIS1 offset (13) microsteps rate (1230769) tracking rate (976894) direction (0) aggressiveness (0.80) hysteresis (0.10)
AXIS1 offset (14) microsteps rate (1142857) tracking rate (913829) direction (0) aggressiveness (0.80) hysteresis (0.10)
AXIS1 offset (15) microsteps rate (1066666) tracking rate (858985) direction (0) aggressiveness (0.80) hysteresis (0.10)
AXIS1 offset (14) microsteps rate (1142857) tracking rate (913683) direction (0) aggressiveness (0.80) hysteresis (0.10)
AXIS1 offset (14) microsteps rate (1142857) tracking rate (913700) direction (0) aggressiveness (0.80) hysteresis (0.10)
...
AXIS1 offset (12) microsteps rate (1333333) tracking rate (1050540) direction (0) aggressiveness (0.80) hysteresis (0.10)
AXIS1 offset (15) microsteps rate (1066666) tracking rate (858903) direction (0) aggressiveness (0.80) hysteresis (0.10)

CHANGED AGGR TO 0.60

AXIS1 offset (13) microsteps rate (1230769) tracking rate (755396) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (954418) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (954535) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (12) microsteps rate (1333333) tracking rate (810647) direction (0) aggressiveness (0.60) hysteresis (0.10)
...
AXIS1 offset (10) microsteps rate (1600000) tracking rate (954623) direction (0) aggressiveness (0.60) hysteresis (0.10)

CHANGED AGGR TO 0.50

AXIS1 offset (11) microsteps rate (1454545) tracking rate (745261) direction (0) aggressiveness (0.50) hysteresis (0.10)
AXIS1 offset (7) microsteps rate (2285714) tracking rate (1118981) direction (0) aggressiveness (0.50) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (810808) direction (0) aggressiveness (0.50) hysteresis (0.10)
AXIS1 offset (7) microsteps rate (2285714) tracking rate (1119202) direction (0) aggressiveness (0.50) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (811011) direction (0) aggressiveness (0.50) hysteresis (0.10)

i tried it the whole day and even wrote a "+400000" in the code for the trackingrate -> but it happens exactly the same thing i showed in the new logs... the offset value should change continuously (become higher or lower depending on the factor)

I'm a bit out of ideas now...

Regards
Last edit: 11 months 3 weeks ago by Mat.
11 months 3 weeks ago #91144

Please Log in or Create an account to join the conversation.

Ok that makes due the inverse relationship between offset and clock rate. Let me test locally today and I'll try to figure out an optimal solution to minimize the offset at least indoors.
11 months 3 weeks ago #91149

Please Log in or Create an account to join the conversation.

  • Posts: 80
  • Thank you received: 11
I am available for testing!
11 months 3 weeks ago #91155

Please Log in or Create an account to join the conversation.

Ok, I'll try using PID controller imported from Celestron AUX driver to see if it helps improves the situation.
11 months 3 weeks ago #91164

Please Log in or Create an account to join the conversation.

Ok PID is added now, please recompile and test.
The following user(s) said Thank You: Mat, Michael
11 months 3 weeks ago #91173

Please Log in or Create an account to join the conversation.

  • Posts: 80
  • Thank you received: 11
yay thank you!

i'll test it as soon as any star shows up and report back! (at the moment... clouds...)
11 months 3 weeks ago #91176

Please Log in or Create an account to join the conversation.

  • Posts: 80
  • Thank you received: 11
Hello Jasem,

again: it is really great that you have implemented the PID changes.

Unfortunately the weather is bad, but I checked the sky again and again throughout the evening and dragged the mount onto the balcony when there was a small gap in the clouds.

I only did one platesolve and had the impression that the tracking drift was much less. So I started guiding and it just worked! The guiding was able now to correct the drift without going crazy:



It was so cool - after a slew to another spot, i restarted and the guiding worked again for a few minutes with low RMS until the clouds came.

I think to really draw a conclusion I need a whole evening (or a few more), but I just wanted to report back that the first impression was very very good! B)

Regards
11 months 3 weeks ago #91182
Attachments:

Please Log in or Create an account to join the conversation.

  • Posts: 39
  • Thank you received: 15
Thank you for implementing the PID loop as part of the tracking controls, Jasem. I have re-built the drivers and installed on the stellarmate. If I get clear skies (fingers crossed), I will test tonight. From Mat's recent post, it looks like you hit a homerun on the first pitch with this version. To confirm, no guide scope is required for the PID loop (i.e. optional as before), correct?

Mat, did you change any of the default values of the PID control or the T1 Clock Rate when you tested? if so, what values did you initially use if not the default? Also, thank you for putting in the testing this week on the other versions.
11 months 3 weeks ago #91196

Please Log in or Create an account to join the conversation.

  • Posts: 80
  • Thank you received: 11
Hi Michael,

No guiding is required!

for me it was a good test to see if Tracking drift is so low now that guiding can correct. I only was able to test 15 minutes with values
1
0
0
And
1.2
0
0

Regards
11 months 3 weeks ago #91197

Please Log in or Create an account to join the conversation.

so this was tested under real sky condition and using default PID, the target does not drift?
11 months 3 weeks ago #91209

Please Log in or Create an account to join the conversation.

Time to create page: 0.838 seconds