Focusmax and Robofocus fun!!!! NOT

[cventer]

Been messing around for last 3 hours trying to get decent focusmax curves

Started off well with a series of 4 curves with a PID of around 2 - 3.2

Tried to get things better and thatswhen things went south

Curves looked good but PID was all over the place.

I also noticed the absolute focus position seemed to be changing every run

At first I thought the focuser must be slipping. Tightened the coupling to robofocus.

Same result. Changing absolute focus point.

Then I thought I may have overdone backlash compensation so I decided to use some digital vernier calibers on focuser to check what was going on

So next i tried following sequence
1. Measure distance
2. Tell focuser to move out 200 steps
3. Tell focuser to move in 200 steps
4. Measure distance

The distance was .5mm different :eyepop:

This got worse further out I moved focuser.

Played with backlash comp but no difference.

Rebooted everything. reset robo. Same result.

Decided to recalibrate the travel points and noticed the motor was buzzing and not moving scope in. I then remembered I read something a while back and changed duty cycle from 50 down to 5.

Eventualy I changed the duty cycle on the motor to 50%

Bingo. Travel recalibrated straight away.

Started re running curves and am now getting 1 - 2.5 PID's as I type this...

!@$%#&* Some nights things just dont go your way.

At least I sorted it in the end I guess and this may be of use to someone one day who sees same behaviour

[Hagar]

Hi Chris. I have found focusmax and robofocus to be very good but it does require a very good smooth and stable focuser. The focus must be repeatable to achieve somewhere near good focus. I have takahashi rack and pinion focusers with almost zero backlash and still the results night to night can vary so much I have gone back to using robofocus with a Bahtinov mask and manually controlling the focus from the PC screen with bahtinov grabber.

I find using this method much more acurate overall and doesn't take much extra time in the long run.
I set up on a brighter star near the target, Insert the mask, do a focus, move to the target, synch the mount on the target, start imaging. After a few frames when the temp is changing or when I notice the FWHM changing I move back to the star and re-do my focus, move back to the image and start again.
This method definitely gives me more consistant results.

Temprature compensation is another mine field set to make life a missery.

Just my thoughts on auto focusing.

[TheDecepticon]

Hi Doug.

I have had a lot of trouble with focus max. I also have robo focus which works well and I have very high repeatable focus. I have a good focuser with almost zero backlash, but my problem is getting focus max to actually work and connect to the focuser. I am on the FM group, but really it is no help. I now focus the same way Hagar does and find the results to be excellent.

[cventer]

I have no problems with focusmax and robofocus working together. And now the duty cycle is sorted its all working again.

I have found focus max to do a very good job of focusing, especillay if you use multiple images per move to average the seeing.

I use 3 images per step and this works great.

Have you got latest focusmax ?, the latest robofocus ascom drivers ?

If you can connect to robofocus via the robofocus server which is what you should choose when installing robofocus drivers, then ascom should have no issues seeing the robofocus.

[Hagar]

Hi Chris, I have been able to get the 2 to connect OK but still find the final focus to be a little bit flakey. Results look good on initial inspection but if I then drop in the mask and use Bahtinov ghrabber I find that the focus can be improved, and often by quite a bit.
I am yet to decide which one is the most accurate due to issues with seeing etc. I just find I feel more confident using the mask and getting it all done manually.