×

INDI Library v2.0.7 is Released (01 Apr 2024)

Bi-monthly release with minor bug fixes and improvements

Read More...

AZ-GTI in AZ-mode tracking problems

  • Posts: 80
  • Thank you received: 11

Replied by Mat on topic AZ-GTI in AZ-mode tracking problems

thank you for your very quick reply - too bad I thought I was on to something. but then we can rule out this movement as the cause…
1 year 1 month ago #91083

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

Replied by Jasem Mutlaq on topic AZ-GTI in AZ-mode tracking problems

I pushed an experimental change to replace Track Factor with Tracking Aggressiveness & Hysteresis. Please try to see if this helps by tuning the values. By default aggressiveness is set to 1 (100%) and hysteresis is 10%
1 year 1 month ago #91084

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

  • Posts: 80
  • Thank you received: 11

Replied by Mat on topic AZ-GTI in AZ-mode tracking problems

Hey Jasem,

thank you very much! i will try!

Sorry, but do you have short instruction how i can update this single driver. I only installed from the astroberry bullseye repo until now.

Regards
1 year 1 month ago #91085

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

Replied by Jasem Mutlaq on topic AZ-GTI in AZ-mode tracking problems

You need to build it from GIT
1 year 1 month ago #91086

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

  • Posts: 80
  • Thank you received: 11

Replied by Mat on topic AZ-GTI in AZ-mode tracking problems

OK found the instructions. I will install, test and report as soon as the clouds are gone!
1 year 1 month ago #91087

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

  • Posts: 80
  • Thank you received: 11

Replied by Mat on topic AZ-GTI in AZ-mode tracking problems

hey Jasem,

i installed the new version from git and took a small LED as an artificial star to test.

First of all, the "ghost movement" after stopping the tracking via padlock is gone. If i uncheck the padlock the mount really stops. Very cool!

Than i tried to test the new aggresiveness setting. So i slewed to my arificial star an gave the mount an random alignment point in KSTARS to have a tracking.

With Aggr 1.0 and Hys 10% (all values allways for both AZ/AL) the star needed 1 minute and 24 seconds to drift through the square.



Than i changed to Aggr 0.8 and hysteresis 10% and the value was exact the same (about 1:24). But i think it should take 20% longer? With 0.7 and 10% same time. If i increase hysteresis to 20% its also 1:24, if i go to hysteresis 30% the star jums like the green line, an than the mount slews away:



It seems to be a bit like the old track factor - if it gets to low the mount jumps. Or did i somethin wrong? It should take longer to move, becaus its only x percent of the old speed, isnt it?

Regards
Last edit: 1 year 1 month ago by Mat. Reason: typos
1 year 1 month ago #91100
Attachments:

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

  • Posts: 80
  • Thank you received: 11

Replied by Mat on topic AZ-GTI in AZ-mode tracking problems

Hey Jasem,

sorry, I had a new thought: Could it be that the track factor or the new aggr doesn't work (see post above) because the calculation of the offset simply increases the rate again (with every step). So maybe the problem is that the rate slowdown is being nullified because the mount thinks "oh I was too slow" and then just increases the rate just enough so that it hits that point it thinks it should be.

i dont know how to explain better:

<code>Mount runs -> calculates offset -> calculates matching speedrate -> applies trackfacktor 0.7 -> NEXT POINT -> claculates offset "noticed: oh i was 0.3 to slow" -> calculates matching speedrate (but compensates and adds the missing 0.3) -> applies trackfacktor 0.7 -> NEXT POINT ... etc</code>

So in the end the speed with 0.9 is always nearly same... that also could explain jumping or the crazy slewing with low values, because if i make the factor very low (>0,5), it loads up the speedrate over a few iterations.
just a thought.

cya
Last edit: 1 year 1 month ago by Mat.
1 year 1 month ago #91110

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

Replied by Jasem Mutlaq on topic AZ-GTI in AZ-mode tracking problems

I'm not sure if my current implementation is correct, but it works as following assuming an Alt-Az mount:

1. Every pulse width (1 second), we convert target equatorial coordinates to horizontal coordinates.
2. Calculate AZ/AL offsets from current encoder position.
3. propotionalRate = offset * aggressiveness (1 by default).
4. finalRate = (1-hysteresis) * propotionalRate + hysteresis * average(rate_history)
5. Send finalRate to mount.
6. Add offset to rate_history

Now that I think about it, perhaps the proportional rate is the one that should be added to history, and not just offset. For 10% hysteresis, the above final rate takes into consideration 90% of the proportional rate and 10% of the average rate history.
1 year 1 month ago #91120

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

  • Posts: 80
  • Thank you received: 11

Replied by Mat on topic AZ-GTI in AZ-mode tracking problems

You might be right, I'm always ready and motivated to test a new change!
1 year 1 month ago #91124

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

  • Posts: 80
  • Thank you received: 11

Replied by Mat on topic AZ-GTI in AZ-mode tracking problems

Hey Jasem,

i took a look at the code and in Line 1099 for AXIS1:
SetClockTicksPerMicrostep(AXIS1, AzimuthRate < 1 ? 1 : AzimuthRate);

and in line 1152 for AXIS2:
SetClockTicksPerMicrostep(AXIS2, trackingRate < 1 ? 1 : trackingRate);

So one time you make SetClockTicksperMicrostep with the variable trackingRate and the first time with AzimuthRate - but I guess it shouldn't be like that, should it?

Regards
The following user(s) said Thank You: Jasem Mutlaq
1 year 1 month ago #91128

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

Replied by Jasem Mutlaq on topic AZ-GTI in AZ-mode tracking problems

Good catch! I pushed a fix.
1 year 1 month ago #91131

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

  • Posts: 80
  • Thank you received: 11

Replied by Mat on topic AZ-GTI in AZ-mode tracking problems

Hello Jasem,

thank you, i downloaded and testet your fix. (i already tested a local fix with same result.) the factor again had no influence on the result.

What i tried to explain earlier is, that i think the mount falls behind by 40% !once! in the first step (because it's slower) and corrects that because it already thinks in the second step, "oh I have to get faster to reach the desired point, im 40% behind" perhaps that could be the reason why neither track factor nor aggr worked to slow down.

the interesting thing is the ratio of the microsteps rate and tracking rate. With aggressiveness (0.60) it just boosts the microsteps rate after the first step to a higher level.

(only AXIS1)
AXIS1 offset (26) microsteps rate (615384) tracking rate (332307) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (11) microsteps rate (1454545) tracking rate (785454) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (8) microsteps rate (2000000) tracking rate (1158545) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (961200) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (961200) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (960651) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (960534) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (960454) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (960395) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (960351) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (960316) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (960287) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (11) microsteps rate (1454545) tracking rate (881681) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (9) microsteps rate (1777777) tracking rate (1055651) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (10) microsteps rate (1600000) tracking rate (960312) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (9) microsteps rate (1777777) tracking rate (1056294) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (11) microsteps rate (1454545) tracking rate (882298) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (9) microsteps rate (1777777) tracking rate (1056365) direction (0) aggressiveness (0.60) hysteresis (0.10)
AXIS1 offset (11) microsteps rate (1454545) tracking rate (882308) direction (0) aggressiveness (0.60) hysteresis (0.10)


if you look at this points (AXIS1) with NO slow down (aggr 1):
AXIS1 offset (27) microsteps rate (592592) tracking rate (533333) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (15) microsteps rate (1066666) tracking rate (1013333) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (14) microsteps rate (1142857) tracking rate (1105905) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (14) microsteps rate (1142857) tracking rate (1105905) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (17) microsteps rate (941176) tracking rate (936760) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (16) microsteps rate (1000000) tracking rate (990496) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (16) microsteps rate (1000000) tracking rate (991922) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (17) microsteps rate (941176) tracking rate (940019) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (16) microsteps rate (1000000) tracking rate (993091) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (15) microsteps rate (1066666) tracking rate (1053781) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (18) microsteps rate (888888) tracking rate (894940) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (16) microsteps rate (1000000) tracking rate (994446) direction (0) aggressiveness (1.00) hysteresis (0.10)
AXIS1 offset (16) microsteps rate (1000000) tracking rate (994863) direction (0) aggressiveness (1.00) hysteresis (0.10)



Or am I totally wrong here?

Regards
Last edit: 1 year 1 month ago by Mat.
1 year 1 month ago #91134

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

Time to create page: 1.170 seconds