Eric,

I'm sorry I didn't explain what I did. I just took a 50s exposure of a random star field with the telescope tracking. While doing it, I sent 5 1000ms guiding pulses each to the north, east, south and west. The double lines have no relevance, they have no impact on the result.That's true, I still have no idea what happaned the other night. At least I now know the mount works properly.
Kevin, excellent idea! I've been thinking about it for a while and believe it might have played a role. I'll misbalance the mount slightly next time and see what happens.

When I noticed that RA+ guiding pulses were moving the star twice as much as did RA- pulses, I thought this could have been caused by a massively missbalance mount. So I checked and found it was well balanced. I do balance it by hand, move each axis to one direction an then to the other and when I have the feeling it moves similarly easily in both directions, I'm done. So, well balanced isn't necessarily perfectly balanced. But then, maybe it was! My G-11's mesh is rather tight though, thus very little backlash and I'd be surprised if this alone could have caused such horrendous calibration plots.

I'm wondering will these crazy plots ever come back so I can find out? Do I really want that to happen? I'm afraid they will, most probably during a moonless production night.

Many thanks for your suggestions, Gentlemen, I'll report back what I find out!
CS!

What's wrong with this editor? Quotes and pictures disappear as soon as I submit the post!

What's wrong with this editor? Quotes and pictures disappear as soon as I submit the post!

Yeah I know. I just manually copy/paste what I'm quoting now, like I just did.
 

Ah, good to know it's not some exotic Firefox plugin's fault. The original picture that I took is attached (at least I hope so).

@Alfred re asymmetric RA motion:

No expert here, but here's a theory.
The mount is tracking in RA, so, RA has no backlash (it's constantly pushing against the gears).
When you give it RA+ pulses, you get the full push.
When you give it RA- pulses, you will go against the normal RA motion and so reverse motion and realize some backlash.
This backlash will cause the system to lose some "push" as compared to the other direction.

The DEC calibration (assuming you have the DEC backlash box checked) removes backlash on the way out, but not on the way back in (where no measurements are made). Dec isn't constantly pushing the way RA is, though, so DEC will eventually overcome the backlash on the way in, but because of the constant tracking RA motion you may always fight backlash in the RA- direction.

Haven't thought deeply about this, does that make sense?
Hy

Hy,

thanks a lot for chiming in and sorry for being late, I've been dealing with a nasty back injury. Regarding RA movement I'm not sure... I certainly am no expert either. My current understanding goes like this: As my guiding rate is 0.5 (I assume that's 0.5x siderial tracking speed), I believe normal tracking (no correction pulse) means 1.0x siderial speed movement. While executing an RA+ pulse it means 1.5x siderial speed and while executing an RA- pulse it means 0.5x siderial tracking speed. So in RA there should never be any backlash involved as the gears "push" the mount in the same direction all the time. Of course I could be wrong here. In any event I would be shocked to find backlash account for the difference in RA+ and RA- movement as it would equal roughly half of 5 800ms (or even 1000ms) pulses = a 2s correction pulse. 2 seconds!  I'd deem that very bad even for a cheap mount Made in China.

Alfred, Sorry about your back. Hope you get better soon.  Let me try my theory one more time. I agree with you that for RA going outward/westward, there would be no backlash. I was talking about RA going the opposite direction as normal sidereal tracking. With RA+ you'd get the full push, but with RA- you'd always have backlash since you are going opposite to tracking. That seems to be what you experienced. Does that make sense?

Hy,

you wrote: "<em>but with RA- you'd always have backlash since you are going opposite to tracking.</em>"

That's the only point where I am of a different opinion. I believe with RA- the mount does not actually move in the opposite direction, it just tracks at half speed but still in the same direction.

So in my theory during RA- you still have the mount moving westward but at 0.5x siderial speed. During normal tracking it moves at 1x siderial speed and with RA+ it moves at 1.5x siderial speed. With a guiding rate of 0.5 there is no (absolute) change in direction at any time (just relative to 1.0x siderial). During all these states the mount continues moving westward, just at different speeds.

In order to experience backlash in RA the guiding rate would have to be greater than 1.0x siderial. With a guiding rate of exactly 1.0x the mount would simply stop tracking during RA- pulses. This is my understanding of "guiding rate". I am not 100% sure about it but to me it does make sense.

I agree, but not 100%. I guess it depends on the reverse RA pulses, and how they compare to the tracking ones. I believe you can get that info directly from the debug logs if you have the indi mount logs enabled, and perhaps that depends on the particular mount driver.

Also note that you can configure the strength of the calibration pulses in the calibrations options menu.

My guess, but not confirmed, is that you can overwhelm the tracking with e.g. 1000ms calibration pulses. I don't know how you set yours.

If you can get the logs to show, let us know what you find out.

Hy,

IMO the "strength of the pulse" is just the pulse duration which defines the amount of time the mount moves at a reduced or increased speed in RA. A 1000ms RA+ calibration pulse means the mount moves at 1.5x siderial speed for 1s. No calibration pulse can ever overwhelm tracking, provided the guiding rate is <= 1x siderial.

I had a look at the manual of my G-11 mount. It explains:

"You can select a guiding rate between 0.2x sidereal and 0.8x sidereal. While the guiding occurs at a constant speed north or south in Dec., it is added to or subtracted from the tracking speed (1x sidereal) in RA. The RA drive always tracks the telescope to the west; it just speeds up or slows down when it receives a guiding correction. The standard Guiding Speed is 0.5x, giving 0.5x and 1.5x sidereal speed in RA. The highest guiding speed is 0.8x, giving approximately 0.8x tracking speed in Dec., and 0.2x and 1.8x in RA."

This seems to validate my view provided calibration pulses and guiding pulses are of the same "nature".

Hello,
Tonight I am getting the same weird behavior as you described.
The east calibration is using the wrong direction, using both st4 or direct pulses, using phd guiding or ekos internal guider.
Sending manual orders works fine.
I changed the gemini2 battery today because it was depleted.
Still have no clue on your side?

Florian,

I didn't have a chance to observe again. I will swap RA and DE motors just to check whether a faulty motor is a part of the problem. In my case equal pulses in Ra+ and RA-  result in unequal correction moves. I never figured out why calibration sometimes goes haywire.