Autofocus woes

[Davey-T]

Trying to get autofocus to work using Lakeside motor focuser and Maxim Sharpstar, tried FocusMax but it won't connect to the focuser.

I set the focus position around the centre of focus start autofocus and it winds the focuser right in until it's at 0 and complains that it can't reach focus not sure how it reckons focus is getting better and it needs to go in more when it's already wound it in miles, if wants to go in further I'll have to fit an extension tube which seems a bit silly.

I can achieve nearly perfect focus manually but would like autofocus to work.

Dave

Starting position 1290 is nearly in focus, when it gets to position 114 winding in and the focus getting worse how can it the think focus is better at position 30 and keep winding in until it runs out of travel ?

 

 

[Ibbo!]

I think its something to do with step size.

I would be able to help more if my obs PC was playing nice, it was fine this afters now it is about to be launched into orbit or I go to bed.

[Starflyer]

Have you set up the focuser in Maxim? This tutorial may help;

http://www.cyanogen.com/help/maximdl/Autofocus_Tutorial.htm

I don't use Maxim to auto focus myself (I use SGP) but I know with both my Lakeside focusers it's essential to have backlash compensation on in order to get good V curves.  It's not important to know the backlash amount exactly, only the direction you approach focus from. I used a figure of 100 steps and tried both inward and outward and just settled on the direction that gave me the best curve.

I don't think backlash compensation is your main problem though, but if the focusing routine is anything like SGP's then you will need it enabled to get the best curves.

I'd suggest going through the tutorial and reporting back here. It can be a pain to setup but once you have it working you'll wonder how you ever managed without it.

 

Cheers,

Ian

 

 

[steppenwolf]

It is critical that you enter the correct step size in the set up but don't interrogate the focuser's ASCOM driver to do this automatically as it has NO knowledge of the focuser's physical attributes. A manual method for determining this figure can be found here near the bottom of the page under the heading 'SharpStar Autofocus'.

Well worth persevering with as it works very well indeed and I swapped to it from FocusMax.

[Davey-T]

Hi all,

Spent a few more nights messing about trying to get autofocus to work, had some partial success but not complete success.

Peter at Lakeside is sending me some notes on sorting  backlash which may help to get Sharpstar working properly.

Got some V curves out of F'Max4, does this look right. should the bottom of the V be in the centre ?

Dave

[steppenwolf]

Don't worry too much about the bottom of the 'V' not being in the centre. I had this with FocusMax and it still achieved good focus.

A backlash setting should be entered in either your focuser's driver or MaxIm but NOT both. You can easily determine the backlash for you focuser in the following manner:-

1. move the focuser outwards 100 steps under motor control 

2. place your finger lightly on the motor's drive shaft

3. move the focuser inwards in single steps counting them as you go until you 'feel' the shaft move. Don't rely on sight!

4. the number of counts is your backlash figure

[Davey-T]

3 hours ago, steppenwolf said:

Don't worry too much about the bottom of the 'V' not being in the centre. I had this with FocusMax and it still achieved good focus.

A backlash setting should be entered in either your focuser's driver or MaxIm but NOT both. You can easily determine the backlash for you focuser in the following manner:-

1. move the focuser outwards 100 steps under motor control 

2. place your finger lightly on the motor's drive shaft

3. move the focuser inwards in single steps counting them as you go until you 'feel' the shaft move. Don't rely on sight!

4. the number of counts is your backlash figure

Tried that method with my finger on the shaft but the Lakeside seems to twitch a bit on first movement so tricky to tell when it's actually moving.

Is this normal for the Lakeside focus motor ?

Dave

[ChrisLX200]

You could try using a cable-tie wrapped around the focus knob, leave a long tail so movement is magnified. SGPro has a simple backlash compensation feature which forces all final focus movements to move in one direction only - ie., it over-runs then reverses direction. It works very well and, provided you set the amount of overrun to be greater than the amount of backlash, pretty much foolproof.

ChrisH

[steppenwolf]

Quote

Is this normal for the Lakeside focus motor ?

Sorry, I can't comment on this as I don't have a Lakeside focuser but I am sure an owner will be along soon!

[Barry-Wilson]

Dave

I use SGP as you know.  I do use Maxim's focus routine when building a model and I can get a good focus curve but I don't really understand its settings as I have used SGP for much longer.

Like Chris, I set a backlash value in SGP and it is vital for an accurate focus as my FT and Tak focuser both have a small amount of backlash being R&P.  What type of focuser are you using?

41 minutes ago, Davey-T said:

Tried that method with my finger on the shaft but the Lakeside seems to twitch a bit on first movement so tricky to tell when it's actually moving.

Is this normal for the Lakeside focus motor ?

Dave

Is there a tension set on the focuser drawtube so that there is a tiny bit of 'stiction' to over come when the stepper motor starts up?  Is the bracket holding the stepper motor flexing slightly giving the impression its the motor twitching?  I have had to periodically reset the set screws securing the motor coupler as they can work a little loose over time, thermal cycling across the seasons etc.  All your fixings set tight?

When the Airy disk of a very out of focus star is large and resembles a donought, the algorithms used to detect a star shape from background noise can struggle to measure HFD/HFR/FWHM (whichever metric is being used by the AF programme) hence the sudden drop at position 30 in your Maxim gtaph.  This was documented extensively on the SGP forum when the developers were adding a 'smart focus' utlity for long focal length RCs, SCTs etc.  The focus routine now works well with all types of scope.  Perhaps you are aware of all of this, so apologies if its doesn't help.

HTH

[Davey-T]

Thanks for the helpful replies folks.

Browsed several tutorials on F'Max4 and there are lot's of do this then do that but no in depth explanation of what and why.

If I put my finger on the focus motor and start moving it I can feel a knocking before detecting any movement, I think this knocking is present even if the motor isn't connected to the shaft,  job to tell if it's moving or ramping up and taking up backlash before it moves, I've got two Lakeside focusers and they both seem to do this so I presume it's normal.

I've adjusted the tension just enough to stop it slipping when aimed straight up, it's a WO110FLT with the stock focuser, I'm considering getting a better focuser, is there no end to astro spending ? now you guys want me to get SGP 

Barry I see you changed the focuser on your WO132FLT was that because the stock one was rubbish ? presumably I could fit the Lakeside MF to a FT focuser.

Dave

[Barry-Wilson]

Hi dave

I bought the WO FLT132 with the 3.5" FT focuser so I have never used a WO R&P focuser so cannot comment from first hand experience.  A few years ago I had a TS 115mm scope - good optics but a terrible crayford focuser.  I swapped this for a 3" FT focuser and performance improved dramatically by giving a flat field and so improving focus.

I'm not sure if the stock focuser of the WO110 is secured onto the OTA by 3 set screws each 120deg apart rather than a screw threaded fit?  If so, I am inherently suspicious of this attachment mechanism for its ability to deliver a secure and stable flat field.  However, I am slightly off topic here.

With both my WO and Tak, I do not have any tension on the focuser tube because the stepper motor itself holds the ccd/filterwheel/reducer without any slippage, and I think is a non too often mentioned advantage of autofocus.  I would take off the tensioning and see what effect it has on your v-curves.

SGP isn't compulsory and Maxim I'm sure will get you were you want to be.  Having used both though (I bought Maxim to build a model with Per's MM for the GM1000HPS) I personally would not use Maxim DL for image capture and co-ordinating my observatory.  SGP is $99, the framing and mosaic wizard is $39 (and is worth ten times that price on its own) all updates are given at no cost and the developers are very active and listen to their customers.  I cannot recommend it enough, especially being an early UK user.  The AF routine in SGP is straightforward and gives repeatible and accurate focusing.

Regarding FT focusers, I have just taken the plunge to buy the Tak FSQ85 FT 3" focuser and adapter.  The Tak focuser is good, but not quite good enough .  You're absolutely right and AP expenditure just seems to go on and on and on .

[Davey-T]

One other thought does anyone / everyone here use focuser temperature compensation ? I've read conflicting opinions on this.

Dave

[steppenwolf]

FeatherTouch focusers really are very good indeed, I have one on my FLT 98 and that works beautifully with my SharpSky autofocuser so I see no reason why a Lakeside unit shouldn't work just as well with one.

SGPro has a great following -and rightly - but if you already have MaxIm DL (as I do) and you get on with it (as I do!) then there is no reason why it shouldn't continue to be the hub of your imaging operations but it isn't compatible with SGPro'. I didn't wish to shift from MaxIm so I orchestrate my automated sessions using CCD Commander which is compatible with MaxIm.

I would thoroughly recommend that you take this step by step and don't throw money at the issue straight away as this can be very unrewarding! My fully automated observatory uses MaxIm as its hub, it is NOT the only solution but it works very well and you have already made that investment so I'd be tempted to use it as a fault-finding vehicle if nothing else and once you know that your system can achieve focus automatically, think very carefully about your next move.

[ChrisLX200]

1 minute ago, Davey-T said:

One other thought does anyone / everyone here use focuser temperature compensation ? I've read conflicting opinions on this.

Dave

Not me. My opinion is that a temperature probe which doesn't measure the temperature of the actual optics and tube interior is not much use, any changes will not track those of the optics. Through experience I know that approximately a one degree change [in ambient] temperature will significantly affect focus - but this is only useful to know for making compensation _after_ the scope has reached equilibrium. So for the first 1-1/2 - 2 hrs (with the 'frac) and 3 hrs (with the ODK12) I will autofocus every 1 or 2 subs regardless, and after that rely on the 1deg change in ambient.

ChrisH

[Barry-Wilson]

Like Chris, I don't either Dave.  I set SGP to re-focus after a specified drop in temperature, 0.5deg C for the large WO FLT132 and 1.5degC for the Baby Q.  These values come from observation of how the focus position changes with temperature.  I know many imagers have plotted a simple graph to show the relationship between the two for their optical train - it can even be used for predictive purposes when the temp is detrmined enabling one to move to the desired position.

[Barry-Wilson]

7 minutes ago, steppenwolf said:

SGPro has a great following -and rightly - but if you already have MaxIm DL (as I do) and you get on with it (as I do!) then there is no reason why it shouldn't continue to be the hub of your imaging operations but it isn't compatible with SGPro'. I didn't wish to shift from MaxIm so I orchestrate my automated sessions using CCD Commander which is compatible with MaxIm.

I would thoroughly recommend that you take this step by step and don't throw money at the issue straight away as this can be very unrewarding! My fully automated observatory uses MaxIm as its hub, it is NOT the only solution but it works very well and you have already made that investment so I'd be tempted to use it as a fault-finding vehicle if nothing else and once you know that your system can achieve focus automatically, think very carefully about your next move.

Steve's right Dave.

We all know how frustrating it can be when something/system doesn't work first time, especially when we've invested so much time and energy.  Be methodical and I'm sure you'll get on top of the integration of systems and devices.  The Lakeside is widely used as is Maxim DL, so there are plenty of folk to help . . . . just keep firing out updates and questions.

[Davey-T]

Thanks all.,

I originally tried Maxim Sharpstar as I've had Maxim for several years so thought I should use it, after Steve pointed out the Sharpstar option, didn't even know it existed lol.

Failing to get that to work I tried F'Max3 which refused to work and F'Max4 which is proving not to be as simple to use as advertised 

Presumably they all use the same basic principle so just need to get one of them working.

Will try again when the sky clears ( the simulator works everytime  )

I'll be carting the whole kit and caboodle to SGLX11 and I see you're a couple of pitches up from me Steve so if you've got a spare minute perhaps you could pop along and hit it with something

Dave

 

[steppenwolf]

Quote

I'll be carting the whole kit and caboodle to SGLX11 and I see you're a couple of pitches up from me Steve so if you've got a spare minute perhaps you could pop along and hit it with something

I always bring a rubber mallet with me on any kind of camping expedition, so we can do this literally or metaphorically 

Hopefully you will be on top of this by SGLX11 but if not, I'd be very happy to have a look with you.

Quote

Presumably they all use the same basic principle so just need to get one of them working.

A 'V' curve is a 'V' curve I guess although FocusMax does have the advantage of storing a profile comprising numerous test 'V' curves to compare against and make the best 'average' focus calculation.

[Uplooker]

Hope you sort your troubles Dave. 

I am following this thread as I have just recently purchased a Lakeside and am not sure what way to jump with regards to software. I currently use APT but am leaning toward SGPro

[Davey-T]

14 minutes ago, Uplooker said:

Hope you sort your troubles Dave. 

I am following this thread as I have just recently purchased a Lakeside and am not sure what way to jump with regards to software. I currently use APT but am leaning toward SGPro

Good luck Ian, SGP would appear to be a good choice, FocusMax3 is available for free but I couldn't make it work.

Dave

[Uplooker]

24 minutes ago, Davey-T said:

Good luck Ian, SGP would appear to be a good choice, FocusMax3 is available for free but I couldn't make it work.

Dave

Hi Dave,

i will stop after this post so as not to distract your thread.

I had seen FocusMax 3 for free but the sites offering it seemed a bit iffy. Where did you download it from? I would just like a bit of a play with it.

Thanks in advance

[Davey-T]

35 minutes ago, Uplooker said:

Hi Dave,

i will stop after this post so as not to distract your thread.

I had seen FocusMax 3 for free but the sites offering it seemed a bit iffy. Where did you download it from? I would just like a bit of a play with it.

Thanks in advance

There you go.

When I tried it I got an error message in the log "error 13 type mismatch" couldn't find out what it meant.

Dave

http://www.astronomylog.co.uk/focusmax-downloads/

[ChrisLX200]

I tried the older free version of FocusMax and didn't get on with it, although it was certainly fast - faster than SGPro. On the other hand SGPro just works and it nails focus with the refractor. It used to struggle a bit with the ODK12 and 10" SCT because it didn't like donut stars, it's much better now though. I prefer the ability of SGPro to work with a large sample of stars scattered across the whole frame, and in the example below (the old version of the focusing algorithm) the scope was pointing at the target - ngc6888 - and doesn't need to be moved to some other FOV to find a suitable star.

ChrisH

[Davey-T]

Looks impressive Chris, just about to order an Astrodon 3nm HA filter so can't spend any more on astro stuff just yet or the astro spending police will be on my case 

Dave