Automated focusing and filter offsets

[alan4908]

I've been using filter offsets in conjunction with my new automated set up for about 6 months, whilst everything appears to be working, I was wondering if I had achieved optimum focus results.

I've attempted to analyse the situation come to a conclusion but I'd appreciate any comments !

Some background: I use FocusMax in conjunction with an ACP script to generate the filter offsets. I currently choose the Lum filter as the reference filter, ACP then finds a focus star near the zenith within a specified magnitude range. Focusmax then adjusts the exposure to ensure that the resulting flux will yield a suitable non-saturated star on the linear portion of my CCD.  This process is repeated for each filter and then is repeated five more times, so that I end up with five focus positions per filter which can then be averaged.  Back in May 2017, I ran this and obtained the following results for my 7 filters:

Filter	Offset
L	0
R	-42.8
G	-15.2
B	16.6
H	-85
O	-52.6
S	-115.6

Recently, I took the opportunity to run these again. The result in January 2018 was:

L	0
R	-47.2
G	-13.8
B	18
H	-67.8
O	-25.2
S	-93.8

I then decided to try to work out which result was better. 

After a bit of searching, I discovered that you first need to work out the Critical Focus Zone (CFZ) for your system, which is how much you can move the focuser and still maintain focus.  I came across this useful presentation by Steve Brady http://www.focusmax.org/Documents_V4/Precision Focusing Using FocusMax.pdf  which details three formulas for calculating the CFZ, unfortunately they give three very different answers. 

CFZ = 4.88 x FR x FR x wavelength -  Equation 1

NCFZI = 1.6 x FR x FR x wavelength - Equation 2

NCFZII = 0.00225 x total seeing x SQRT (focus tolerance) x Aperture x FR x FR - Equation 3

The first two formulas do not take into account seeing or the telescope aperture. The last formula takes into account both of these items including an additional item for focus tolerance.  More details of on the third equation can be found here: http://www.goldastro.com/goldfocus/ncfz.php

So, if I plug in the numbers for my system and I choose to use a wavelength for red light of 0.65 micrometers, a total seeing of 2.5 arc seconds and a willingness to tolerate a 10% focus error I get:

CFZ = 161 microns

NCFZI = 53 microns

NCFZII = 135 microns

I decided to take the result of 135 microns as my theoretical critical focus zone since equation 3 would appear to be the most accurate predictor. 

Now, Focusmax records the absolute position of my Feathertouch focuser which is driven by Starlight Instruments stepper motor.  After looking up the motor specifications it turns out that 1 focuser step equates to just over 1 micron, which allows me to convert the critical focus zone into focuser steps, if I do this, my theoretical CFZ for my system is 129 focuser steps.

So, to me, if my L filter focus position moves by more than 129 steps during the focus offset run, it means that the calculated filter offsets are invalid. Here, I'm assuming that the focus position could move between measurements because of either a temperature or seeing change. 

If I then go back to my logs of the filter offsets generated in May 2017, it turns out that my L filter position changed by 115 steps during the offset measurement, the temperature also changed by 1.7 C.  In January 2018, my L filter changed by 67 steps with the temperature changing by 0.7 C.

So, whilst both filter offsets results are within the CFZ it would appear that the January filter offset result is likely to give more accurate offsets.

Do you agree ?

Alan  

 

 

 

[steppenwolf]

On 07/02/2018 at 16:33, alan4908 said:

So, whilst both filter offsets results are within the CFZ it would appear that the January filter offset result is likely to give more accurate offsets.

Do you agree ?

Yes, I do agree with your conclusion. I too have found some disparities between different test sessions but in all cases, I have been satisfied with the focus achieved for each filter. I think it is all too easy to read too much into the amount of movement going on here but based on the two sets of data that you have, it would make good sense to go with the January, 2018 offsets.

As an aside, although I have the same focuser as you, I have a different focus motor drive connection to yours and I have a drive spindle connected to the the slow motion shaft and I have noted that over a period of about six months, my stepper position changes by several thousand steps but the reality of this is that the focus slip equates to about 1/4 of a turn of the slow motion shaft. Every six months or so, I reposition everything and set the 'L' position to 20,000 and use this as my start point when the session begins. Naturally, this doesn't affect the offsets in any meaningful way.

[alan4908]

1 hour ago, steppenwolf said:

Yes, I do agree with your conclusion. I too have found some disparities between different test sessions but in all cases, I have been satisfied with the focus achieved for each filter. I think it is all too easy to read too much into the amount of movement going on here but based on the two sets of data that you have, it would make good sense to go with the January, 2018 offsets.

As an aside, although I have the same focuser as you, I have a different focus motor drive connection to yours and I have a drive spindle connected to the the slow motion shaft and I have noted that over a period of about six months, my stepper position changes by several thousand steps but the reality of this is that the focus slip equates to about 1/4 of a turn of the slow motion shaft. Every six months or so, I reposition everything and set the 'L' position to 20,000 and use this as my start point when the session begins. Naturally, this doesn't affect the offsets in any meaningful way.

Thanks for taking the time to look at this Steve. OK I shall use the January 2018 offset data. I shall see if I can spot any difference !

Yes, I also see a stepper motor position change - over a similar time period - in my case between the May 17 and Jan 18 it moved by about 1700 steps but as you say, it shouldn't impact the offset results.  

Alan

[steppenwolf]

29 minutes ago, alan4908 said:

Yes, I also see a stepper motor position change - over a similar time period - in my case between the May 17 and Jan 18 it moved by about 1700 steps but as you say, it shouldn't impact the offset results.  

That's interesting and makes me feel a little bit better as yours is the official motorised focuser. I don't see it as a problem, just a function of the ball-bearing drive reduction.

I look forward to hearing how you get on with the new offsets.

[alan4908]

On 2/11/2018 at 17:12, steppenwolf said:

That's interesting and makes me feel a little bit better as yours is the official motorised focuser. I don't see it as a problem, just a function of the ball-bearing drive reduction.

I look forward to hearing how you get on with the new offsets.

Well, after a few months imaging with my new filter offsets....my conclusion is that my focusing has improved with the new offsets. 

Alan

[steppenwolf]

28 minutes ago, alan4908 said:

Well, after a few months imaging with my new filter offsets....my conclusion is that my focusing has improved with the new offsets. 

That's great news, Alan. I have to say that I am pleased with the way mine behaves despite that long-term 'drift'.