Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Correct way to setup electronic focuser on SCT

  • Please log in to reply
18 replies to this topic

#1 Owk

Owk

    Vostok 1

  • -----
  • topic starter
  • Posts: 135
  • Joined: 18 Sep 2016

Posted 29 March 2020 - 09:41 AM

G’Day members

 

im fairly new to electronic focusers. I purchased the ZWO EAF, and am struggling to set it up.

 

ive installed the hardware and ascom driver. All seems to work fine. 
 

the problem is I use my C925 for visual and EAA, both with and without a reducer. There doesn’t seem to be enough focus motor travel for the entire range focus. Is that how it should be?

 

if so, what’s the best way to setup so the focuser covers the focal lengths I need? There is very little instruction on how to do this, what to set zero position to, max step etc. all quite bewildering to me!

 

cheers

 

hemi



#2 carolinaskies

carolinaskies

    Mercury-Atlas

  • *****
  • Posts: 2,974
  • Joined: 12 Dec 2014
  • Loc: Greenville SC

Posted 29 March 2020 - 10:07 AM

What focuser are you using? Main focus knob or an external attached to the back?

If you can post a picture of the rear of your setup showing the EAF installed it would help.  

If you're using an electric attached to the main focus knob you shouldn't have any issues, if the EAF is somehow stopping then see if there is a limit being set in direction of travel. 

If you are using an external focuser remember that an SCT external has depending on which focuser only .5 to .9 travel.  In this case typically you set the external to 1/2 that distance (either .25 or .45) and do rough focus with the main primary mirror focuser first then move to the external.     

 


  • Jon Isaacs and Old Speckled Hen like this

#3 Bob_Stan

Bob_Stan

    Vostok 1

  • -----
  • Posts: 173
  • Joined: 08 Jan 2006
  • Loc: Northeast

Posted 29 March 2020 - 10:37 AM

As already mentioned, an external focuser will not have enough travel.  I use my 11in SCT the same way you do.  My procedure is (each time I add or remove the reducer) - set the electronic focuser to the mid range of its travel. Use the SCT focusing knob to get focused as close as possible.  From then on use the electronic focuser to refine the focus.  There is no way the electronic focuser alone will adjust for the addition or removal of a reducer.



#4 pkrallis

pkrallis

    Mariner 2

  • *****
  • Posts: 277
  • Joined: 06 Oct 2007
  • Loc: Colorado Springs, CO

Posted 29 March 2020 - 10:50 AM

Bob Stan is right.  If you want to be able to run the full focal range  of the OTA you would need install a unit that drives the OTA's focuser directly such as the 'Celestron Motor Focuser.'


Edited by pkrallis, 29 March 2020 - 10:51 AM.


#5 Owk

Owk

    Vostok 1

  • -----
  • topic starter
  • Posts: 135
  • Joined: 18 Sep 2016

Posted 29 March 2020 - 04:30 PM

Thanks Guys,

 

im using an EAF attached to a feather touch micro focuser for sct. It drives the mirror directly.
at some point the eaf stops turning and beeps.

Bob Stan, are you saying that the eaf cannot drive the full focal range of the telescope set up in this way, ie if I add or remove the reducer? That would be a bit of a pain, as to manually focus to get in the ball park, I need to take the EAF off, which requires undoing some grub screws etc.

 

 

cheers


Edited by Owk, 29 March 2020 - 04:37 PM.


#6 carolinaskies

carolinaskies

    Mercury-Atlas

  • *****
  • Posts: 2,974
  • Joined: 12 Dec 2014
  • Loc: Greenville SC

Posted 29 March 2020 - 09:07 PM

From the install manual for the EAF

For the standard bracket installation:

The 0 position is recommended to be set to the position when the
focuser is fully retracted.
The maximum number of steps is recommended to be set to a limit
position less than or equal to the full focuser stroke.

For the curved saddle installation:

The 0 position is recommended to be set near the focus of the
focus of the focus holder. The maximum number of steps is recommended to be set to a limit
position less than or equal to the focus seat stroke.


It is likely you have not set the EAF correctly.  I'd suggest going back and looking at the manual because for the main mirror focuser you should be able to rack it between full CW and full CCW.  
 



#7 Owk

Owk

    Vostok 1

  • -----
  • topic starter
  • Posts: 135
  • Joined: 18 Sep 2016

Posted 30 March 2020 - 11:24 AM

Thanks PaulT,

 

well I’m not using the standard bracket or curved saddle. I’m using the sct adapter for feathertouch micro focuser for SCT, as I said.

 

unfortunately the manual doesn’t elaborate much about setting it up software wise, and I’ve read it over several times...it’s pretty light weight.

 

in any case, I’ve worked out that indeed the motor Focus has considerably longer throw then I was encountering. In ascom setup, if I reset the focus position to zero it will then allow further winding in the direction it stopped and beeped at. It stops at the value you set the maximum step count to be. Unfortunately it doesn’t seem you can put the max step count beyond 60000. Which is way to short, the Amount of times I reset the zero position suggests I need a range of 0 to 300000, to encompass the range between hard stops on the sct.

 

cheers

 

hemi



#8 carolinaskies

carolinaskies

    Mercury-Atlas

  • *****
  • Posts: 2,974
  • Joined: 12 Dec 2014
  • Loc: Greenville SC

Posted 30 March 2020 - 12:21 PM

Thanks PaulT,

 

well I’m not using the standard bracket or curved saddle. I’m using the sct adapter for feathertouch micro focuser for SCT, as I said.

 

unfortunately the manual doesn’t elaborate much about setting it up software wise, and I’ve read it over several times...it’s pretty light weight.

 

in any case, I’ve worked out that indeed the motor Focus has considerably longer throw then I was encountering. In ascom setup, if I reset the focus position to zero it will then allow further winding in the direction it stopped and beeped at. It stops at the value you set the maximum step count to be. Unfortunately it doesn’t seem you can put the max step count beyond 60000. Which is way to short, the Amount of times I reset the zero position suggests I need a range of 0 to 300000, to encompass the range between hard stops on the sct.

 

cheers

 

hemi


Hmm,  It sounds like you are attached to the fine-micro focus position and you need to be attched to the coarse position. 
Recall that the fine focus is 10x sensitivity. 

If you still have the stock focuser I'd attach it to that and see if it will cycle for you. 

Of course the reality is you don't typically need to cycle fully only between use of the reducer and use in native configuration.
If you can figure out if that difference is within the steps you have available then you'd center the focuser midpoint in the range and there you go.   

The installs I could find on SCTs typically aren't on the primary focuser but on the external focusers.  



#9 Owk

Owk

    Vostok 1

  • -----
  • topic starter
  • Posts: 135
  • Joined: 18 Sep 2016

Posted 31 March 2020 - 03:09 AM

Thanks again PaulT,

 

I setup the EAF, following the instructions and an online video precisely. Its pretty simple standard stuff, in any case...but I did check, and is definitely attached to the coarse focus. The coupler is gripping the same thing im attaching the coarse focus knob, while I jump to and fro whilst sorting this out :)

 

Cheers

 

Hemi



#10 carolinaskies

carolinaskies

    Mercury-Atlas

  • *****
  • Posts: 2,974
  • Joined: 12 Dec 2014
  • Loc: Greenville SC

Posted 31 March 2020 - 09:17 AM

Thanks again PaulT,

 

I setup the EAF, following the instructions and an online video precisely. Its pretty simple standard stuff, in any case...but I did check, and is definitely attached to the coarse focus. The coupler is gripping the same thing im attaching the coarse focus knob, while I jump to and fro whilst sorting this out smile.gif

 

Cheers

 

Hemi

OK, then it's pretty clear the EAF isn't up to the job with the stepper count limit for full travel.  Have you checked to verify whether it will fall in the step limit with and without the reducer in place?    

 



#11 Rac19

Rac19

    Vanguard

  • *****
  • Posts: 2,440
  • Joined: 05 Aug 2016

Posted 22 July 2021 - 10:04 PM

I haven't read all the replies but I can add my 2 cents worth, based on experience with a newly purchased SCT this week.

 

The EAF has a numeric range of 0 to 60,000 because it uses a 16 bit integer. There are 5,760 counts per turn (revolution) which is 10.4 turns in total. Full mirror travel on my Evolution is about 45 turns.

 

The advice from ZWO is that when you reach the either limit, reset the counter and continue in the same direction. If you reach 60,000, reset the counter to 0 and continue in then same direction. If the limit of 0 is reached reset 60,000 and continue in the same direction.

 

If a boundary limit is close to focus, I think that there is little choice but to reset to 30,000, which I think creates a potential problem.

 

In my case, I picture the mirror travel being divided into four zones, 5 to 15.4 turns, 15.4 to 25.8 turns, 25.8 to 36.2 turns and 36. 2 to 46.6 turns. They can be called zones 1, 2, 3, 4, or whatever you like. I have always kept a spreadsheet of focus positions and now have an extra column for the "Zone". The problem is, of course, that you need to mentally keep track of the "Zone".

 

It can be a problem if the point of focus happens to be right on a Zone boundary. That is why I start at 5 turns, not 0.

 

I have proposed some ideas to Zwo to extend the numeric range, but there has not been much interest. One suggestion was to simply make the 16 bit integer to a 32 bit integer. Another is to introduce an automated "Zone" counter.



#12 glancey

glancey

    Vostok 1

  • *****
  • Posts: 164
  • Joined: 05 Jun 2019
  • Loc: Santa Ana, CA, USA

Posted 22 July 2021 - 10:25 PM

I haven't read all the replies but I can add my 2 cents worth, based on experience with a newly purchased SCT this week.

 

The EAF has a numeric range of 0 to 60,000 because it uses a 16 bit integer. There are 5,760 counts per turn (revolution) which is 10.4 turns in total. Full mirror travel on my Evolution is about 45 turns.

 

The advice from ZWO is that when you reach the either limit, reset the counter and continue in the same direction. If you reach 60,000, reset the counter to 0 and continue in then same direction. If the limit of 0 is reached reset 60,000 and continue in the same direction.

 

If a boundary limit is close to focus, I think that there is little choice but to reset to 30,000, which I think creates a potential problem.

 

In my case, I picture the mirror travel being divided into four zones, 5 to 15.4 turns, 15.4 to 25.8 turns, 25.8 to 36.2 turns and 36. 2 to 46.6 turns. They can be called zones 1, 2, 3, 4, or whatever you like. I have always kept a spreadsheet of focus positions and now have an extra column for the "Zone". The problem is, of course, that you need to mentally keep track of the "Zone".

 

It can be a problem if the point of focus happens to be right on a Zone boundary. That is why I start at 5 turns, not 0.

 

I have proposed some ideas to Zwo to extend the numeric range, but there has not been much interest. One suggestion was to simply make the 16 bit integer to a 32 bit integer. Another is to introduce an automated "Zone" counter.

 

I use the EAF on a C9.25 with the ZWO SCT bracket. My experience is similar to the above, but the problem remains from one setup to the next, you never really know what zone your focuser in. You have no markings to go by. Even if you write it down, can you guarantee you haven't moved it anytime since.  Say you have a focuser setting of 18,000. Is that in zone 1, or zone 2, which would put it at 78,000, or zone 3, at 138,000?  You get it.

 

Once you lose track, you have to recalibrate and that means backing the focuser all the way until it stops, at 0, and then restart your focus routine. At the speed of the EAF, that could take all night.


Edited by glancey, 22 July 2021 - 10:27 PM.


#13 carolinaskies

carolinaskies

    Mercury-Atlas

  • *****
  • Posts: 2,974
  • Joined: 12 Dec 2014
  • Loc: Greenville SC

Posted 22 July 2021 - 10:39 PM

I think the better method for use with an SCT is to change the ratio of the EAF to focuser coupling. Since the inherent problem is a lack of steps then changing the gear ratio would provide meaningful steps to achieve enough focus without having to reset the count.

#14 Rac19

Rac19

    Vanguard

  • *****
  • Posts: 2,440
  • Joined: 05 Aug 2016

Posted 22 July 2021 - 10:40 PM

I use the EAF on a C9.25 with the ZWO SCT bracket. My experience is similar to the above, but the problem remains from one setup to the next, you never really know what zone your focuser in. You have no markings to go by. Even if you write it down, can you guarantee you haven't moved it anytime since.  Say you have a focuser setting of 18,000. Is that in zone 1, or zone 2, which would put it at 78,000, or zone 3, at 138,000?  You get it.

 

Once you lose track, you have to recalibrate and that means backing the focuser all the way until it stops, at 0, and then restart your focus routine. At the speed of the EAF, that could take all night.

I get it, that's for sure. Life becomes difficult if you lose track of the absolute mirror position.

 

Something that would help would be a simple mechanical (or electronic) counter that fits over the EAF flexible coupling, such as the one shown below. The challenge is that it would need to have a 25 mm bore and be no more than 20 mm thick, so as not to render the flexible coupling rigid.

 

Also, I think that it is a serious omission in the basic SCT design that there is no built in mirror position indication, such as a small window in the OTA or a graduated pin projecting from the rear cell.

Attached Thumbnails

  • RevCounter.jpeg

Edited by Rac19, 22 July 2021 - 10:41 PM.


#15 glancey

glancey

    Vostok 1

  • *****
  • Posts: 164
  • Joined: 05 Jun 2019
  • Loc: Santa Ana, CA, USA

Posted 23 July 2021 - 11:42 AM

I get it, that's for sure. Life becomes difficult if you lose track of the absolute mirror position.

 

Something that would help would be a simple mechanical (or electronic) counter that fits over the EAF flexible coupling, such as the one shown below. The challenge is that it would need to have a 25 mm bore and be no more than 20 mm thick, so as not to render the flexible coupling rigid.

 

Also, I think that it is a serious omission in the basic SCT design that there is no built in mirror position indication, such as a small window in the OTA or a graduated pin projecting from the rear cell.

It doesn't have to be a mechanical counter. The change could be done in the ASCOM and indilib drivers for the EAF and possibly a firmware fix.


  • Rac19 likes this

#16 Rac19

Rac19

    Vanguard

  • *****
  • Posts: 2,440
  • Joined: 05 Aug 2016

Posted 23 July 2021 - 05:03 PM

It doesn't have to be a mechanical counter. The change could be done in the ASCOM and indilib drivers for the EAF and possibly a firmware fix.

I have made three suggestions but none of them has received an enthusiastic response.

 

1. Use a 32 bit integer, probably best done in EAF firmware.

 

2. Implement an automatic "Zone" counter, at driver level.

 

3. Offer a timing belt SCT adapter kit with a 4:1 speed/travel multiple.

 

Item 2 would increment a "Zone" counter on reset from 60,000 to 0 and would decrement it on reset from 60,000 to 0, Another option could be attempt to simulate a 32 bit integer at driver level, but that could be a bit dodgy.

 

Item 3 could also be used improve back-end clearance (see image), which is quite tight on my Evolution 8. It would multiply the resistance torque 4x, so that would have to be evaluated. It would be more expensive than the current flexible coupling, but I would buy it.

 

My guess is that both the ASCOM and indlib drivers are based on the native driver. The parameters for ASCOM driver are same as ASICAP for example.

 

Meanwhile, I have used the EAF on 2 nights so far and it works well for focusing at any particular mirror position. I think that a range of 15 to 25 turns will cover 95% of my requirements. It will get awkward if one my optical configurations happens to focus near 0 or 60,000.

 

EDIT: Further to Item 3. There is a question as the minimum number of teeth possible on the smaller pulley (larger shaft), it's probably more than 15 T. This, in turn, affects the size of the larger pulley. There is also the question of whether or not a 4:1 ratio is feasible for a timing belt. If not, 3:1 or even 2:1 would still be helpful.

Attached Thumbnails

  • DD65925B-FDBE-4D74-8415-306A2AAA0649.jpeg

Edited by Rac19, 23 July 2021 - 06:28 PM.


#17 glancey

glancey

    Vostok 1

  • *****
  • Posts: 164
  • Joined: 05 Jun 2019
  • Loc: Santa Ana, CA, USA

Posted 23 July 2021 - 06:31 PM

I have made three suggestions but none of them has received an enthusiastic response.

 

1. Convert to a 32 bit integer, probably best done in EAF firmware.

 

2. Implement an automatic "Zone" counter, at driver level.

 

3. Offer a timing belt SCT adapter kit with a 4:1 speed/travel multiple.

 

Item 2 would increment a "Zone" counter on reset from 60,000 to 0 and would decrement it on reset from 60,000 to 0, Another option could be attempt to simulate a 32 bit integer at driver level, but that could be a bit dodgy.

 

Item 3 could also be used improve back-end clearance (see image), which is quite tight on my Evolution 8. It would multiply the resistance torque 4x, so that would have to be evaluated. It would be more expensive than the current flexible coupling, but I would buy it.

 

My guess is that both the ASCOM and indlib drivers are based on the native driver. The parameters for ASCOM driver are same as ASICAP for example.

 

Meanwhile, I have used the EAF on 2 nights so far and it works well for focusing at any particular mirror position. I think that a range of 15 to 25 turns will cover 95% of my requirements. It will get awkward if one my optical configurations happens to focus near 0 or 60,000.

 

EDIT: Further to Item 3. There is a question as the minimum number of teeth possible on the smaller pulley (larger shaft), it's probably more than 15 T. This, in turn, affects the size of the larger pulley. There is also the question of whether or not a 4:1 ratio is feasible for a timing belt. If not, 3:1 or even 2:1 would still be helpful.

I like this discussion. The problem with #3, as I see it, is that you might never achieve perfect focus regardless of the number of teeth involved. Celestron's SCT focusers are extremely sensitive, especially in poor seeing conditions. 

 

As for adding a counter, well, that assumes SCT mirror travel remains constant over time and temperature. I'm not so sure that's correct. 

 

Your first choice, re-typing the step size from short to long integers, or even to a double, so you can have fractional steps, in the firmware and/or drivers is the simplest and most foolproof fix. The only issue I can see with that is determining the value of each step. You don't want to make it so small as to be unnoticeable nor too large as to miss focus.

 

Yeah, then again, maybe ZWO isn't so far off, requiring the user to do some calibration on their own SCT. After all, even Celestron's motorized focuser requires calibration on startup.



#18 Old Speckled Hen

Old Speckled Hen

    Viking 1

  • -----
  • Posts: 597
  • Joined: 18 May 2020
  • Loc: 56.4451° N, 3.1670° W

Posted 23 July 2021 - 07:17 PM

I think the better method for use with an SCT is to change the ratio of the EAF to focuser coupling. Since the inherent problem is a lack of steps then changing the gear ratio would provide meaningful steps to achieve enough focus without having to reset the count.

I think the better method is to ditch one featherlight and slap a replacement one on the back...at least then you will have a scale on the drawtube to help.



#19 Rac19

Rac19

    Vanguard

  • *****
  • Posts: 2,440
  • Joined: 05 Aug 2016

Posted 23 July 2021 - 07:55 PM

I like this discussion. The problem with #3, as I see it, is that you might never achieve perfect focus regardless of the number of teeth involved. Celestron's SCT focusers are extremely sensitive, especially in poor seeing conditions. 

 

As for adding a counter, well, that assumes SCT mirror travel remains constant over time and temperature. I'm not so sure that's correct. 

 

Your first choice, re-typing the step size from short to long integers, or even to a double, so you can have fractional steps, in the firmware and/or drivers is the simplest and most foolproof fix. The only issue I can see with that is determining the value of each step. You don't want to make it so small as to be unnoticeable nor too large as to miss focus.

 

Yeah, then again, maybe ZWO isn't so far off, requiring the user to do some calibration on their own SCT. After all, even Celestron's motorized focuser requires calibration on startup.

The following are comments/clarifications are not intended to be criticalsmile.gif.

 

I agree that 4x the speed could potentially affect positioning accuracy, if it is not a stepper motor, but I think that it probably is a stepper motor. As for resolution, 5,760/4 = 1,440 counts per turn. The Celestron focus motor has 1,000 counts per turn. The default coarse and fine steps appear to 1,000 and 100 steps respectively, so reducing them to 250 and 25 should be equivalent. The main areas requiring evaluation, I think, are torque and physical dimensions & layout.

 

The mechanical counter would count the turns of the focus shaft (0 to 45) and provide an indication of the rough, absolute mirror position. I can't find one with a suitable bore and length (along the shaft) unfortunately.

 

Changing from 16 bit (0 to 65,535) to 32 bit (0 to 4,294,967,295) would have no effect on resolution or accuracy, it would simply increase the range of values available. In many ways it would be (by far) the best solution but would require a firmware update for the EAF. Even if the EAF doesn't natively support 32 bit integers, I am pretty confident that something could be done in the EAF's logic using 2 x 16 bit registers.

 

I don't think that ZWO is giving any of this serious consideration by the way.


Edited by Rac19, 23 July 2021 - 08:09 PM.



CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.







Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics