SCT Telescope Focus Variation vs Temperature

Another astrophotography video on focusing! We spend some time looking at the Celestron C9.25 SCT telescope focus variation as the outside temperature drops during the night. Focus points are made periodically over a four-night period using a Bahtinov mask and the Celestron Focus Motor. Unfortunately, at present, neither Astrophotography Tool (v3.84) nor NINA will allow for temperature-based focus compensation for focusers, like the Celestron Focus Motor, that do not have a dedicated temperature sensor. Oh well.

[@salomon1639]

I switched to NINA and recognizes Pegasus ultimate power box. I used to use APT but NINA is awesome especially for such an analytical person as you. I always enjoy your videos. Very scientific and detailed. Thanks

[@Aero19612]

Thanks Salomon! I'm sitting here right now getting NINA set up for my next night out (a few days away). I'm ready to give it a try!

[@ColeRees]

This is fascinating! You’ve got a great setup

[@andygaras]

Brilliant work, thank you

[@Aero19612]

Glad you enjoyed it, Andy! Thanks for watching.

[@yosmith1]

Great analysis, and I'll be interested to see if this patterns holds consistent. I will say this, after following your issues with that focuser backing plate. I have the same SCT/focuser setup, and keep a closer eye on mine for any issues. I also recently saw someone mention this problem with their SCT on a FB post, so it's obvious this problem isn't exclusive to your setup. Thanks again for the analysis.

[@Aero19612]

Thanks for watching, Neil! Absolutely. Keep and eye on your focuser for "odd" behavior. Good luck!

[@salomon1639]

Your correct in that NINA does not have temperature compensation yet but working on it. Just use autofocus routine and specify to perform an autofocus routine with any % chamber change in HFR. I

[@Aero19612]

Yes, I could do that. I'm being very picky, by trying to avoid the V-curve back-and-forth focus algorithm. But that may be my only option.

[@bjdiss]

James- I just l learned something interesting about NINA that I'll share here. In the Sequence panel there is an option to trigger an autofocus run based on temperature change. The temperature on that screen must come from an focuser. The Pegasus probe temp is not considered. In NINA, temperature is not unified across the application. This fact, and your findings here make me want to look at other focuser options. Something temperature compensated that could move the focuser independently based on observed temp.

[@Aero19612]

Exactly! APT doesn't recognize the UPBv2 but NINA does. Seems likes it would be a REAL simple thing to allow the user to pick any temperature source and assign it to the focuser for those of us with a focuser that has no temperature sensor. I was about to buy the Pegasus Astro Focus Cube for my SCT because I liked it so much on my refractor. I paused and happened across the Celestron focuser. I like that the Celestron Focuser has a direct mechanical link rather than the belt drive used by the Focus Cube. In retrospect, knowing what I know now, I might go back and consider the Focus Cube again.

[@bjdiss]

@@Aero19612 I was looking at that one and the ZWO. I like the way ZWO attaches better.

[@Aero19612]

I wasn't aware ZWO's focuser had an SCT attachment option. Will have to look into that.

[@CuivTheLazyGeek]

@@Aero19612 It will apparently get released in September! I am avidly waiting for it...!

[@Aero19612]

I will certainly take a look. Very interested in how it connects to the OTA. The only question is: Current Frustration = How Many $$. Thanks for the heads up guys.

[@comeraczy2483]

Brilliant! This is a lot of extremely useful information. There is still one important question: is the frequency of focus adjustment consistent with focusing error mean and SD? I am asking because your curve at 10-13mn shows 16 measurements, 2 are more than 20 steps away from the curve, and 4 are more than 10 steps away. This might be an indication that in up to 40% of the cases the focus might be out of the expected 7-steps wide critical focus zone.

[@Aero19612]

Good question. First, I suspect that the frequency of focus adjustment based on critical focus zone is overestimated compared to what the eye can perceive. Second, the randomness (SD) comes from several sources including Bahtinov focus error (seeing variation) and the difference between the effective OTA temp and the outside air temp. So, I think the trend lines are a good measure of the behavior that compensates for the error sources. Thanks for watching!

[@DiegoGomez-px6lh]

Very useful information James! From the picture of your rig, it seems that you're using a dew heater strip and the end of the OTA. Was it running while collecting the focuser data? It would be interesting to see if there are variations in the readings with and without the dew heater strip running. I've read a several post in CN indicating that temperature compensation doesn't work very well with SCT because the relation is not completely linear. Those fellows recommend to track ambient temperature and adjust focuser using the V curve or refocus every 20 to 30 mins.compensation

[@Aero19612]

Hey Diego, I don't think the dew strap on the SCT has any effect. SCT focus changes because the tube changes length with temperature of the tube. Of course the tube temperature will lag behind the air temperature. But the effect is effectively linear (as my data show). For refractors, the focus/temperature relationship is more complicated because the index of refraction of glass is highly temperature dependent and heating of the objective lenses with a dew strap could introduce one effect while temperature drop of the tube would create another. I do agree that V-curve (I use the hyperbolic fit) focus routines are more accurate. The only reason I'm "fighting" the V-curve focus method for my SCT is because my stupid Celestron focus motor works itself loose from the SCT because of the back and forth motion of the focuser. The V-curve focus method works great for my refractors! Thanks for watching, Diego!

[@JohnMcGFrance]

Interesting as always. Could you perhaps use the hfd percentage change feature in NINA to reduce the need to refocus. As the temp drops and th hfd changes by calculating what percentage change a 1 degree drop induces it could be that NINA then moves the focuser a set number of steps. The guys developing NINA seem very open to suggestions so it might be something they could implement so that rather than run a new autofocus it just makes the step adjustment then resumes the sequence.

[@Aero19612]

I could do that. I'm trying to avoid too many V-curve-type autofocus runs. Your idea is a good one: maybe I should try and relate HFD growth VS temp. Thanks for watching, John!

[@scottrk4930]

Hi James . Question , do you have actual Star Images taken on either side of the Critical Focus Zone , or beyond , that connects actual Images to the measured Focus Position Data ? I am curious as to when our critical eye will actually trigger our brain into believing the Image is "out of Focus" . Obviously we have the Digital and Mechanical Tools to reassure ourselves that our System is as perfect as we can get but I'm wondering what the allowable error might be to get acceptable results ? /SRK

[@Aero19612]

That’s a good question, Scott. I did a video showing how to use APT and the hyperbolic curve fit where you would manually do what NINA does: Move the focuser way out, take picture, move the focuser in a bit, take a picture...and so on. So there should be pictures in and outside of focus that could be used to answer that question. I wonder if I still have those images. Would be easy enough to repeat though. Might put that on my list!

[@andrebergeronastro]

What would happened if you were using a full heating blanket (witch maybe does not exist) to cover the whole tube assembly to keep constant temperature ?

[@Aero19612]

HaHa. I've been think along those lines too. I did buy some insulation to wrap around the tube to see if it at least delays the temperature loss as the outside air temperature drops-haven't done the experiment yet. But in the extreme case, why not have a temperature control system like is used for the camera sensor that maintains a constant tube wall temperature using maybe two dew straps. If you could do that, you could maintain focus all night long! Thanks for watching, Andre!

[@briandenley]

Excellent engineering analysis. I am not crazy about my Celestron focuser thus far. It juts doesn't feel stable. I was more interested in your results showing that maybe the choices of aluminum is probably the wrong way to go. My scope is the Evolution 8 with the same material. While it's rugged, as you point out, Al's coefficient of thermal expansion is rather high. Thanks for sharing!

[@Aero19612]

I never would have thought of that at the time I bought my SCT (my first scope). I don't think Celestron even offers SCTs in carbon fiber. If they do, I bet it costs an arm and a leg. Thanks for watching, Brian!

[@briandenley]

James Lamb interesting site: carbonscopetubes.com/

[@Aero19612]

Interesting indeed! Thanks for passing this along!

[@poruatokin]

Great video. One point, my refocusing algorithm only takes maybe 2 minutes to complete. At around 16:29 you suggested o\yours is taking 10 minutes. Isn't that rather excessive? (I'm using EKOS)

[@Aero19612]

Haha. I think calling a 10-min procedure in astrophotography "excessive" is excessive. Was there ever a bigger time sink than astrophotography? No, I didn't say it "was" taking 10 min. If said "if" it took 10 min -- I hadn't performed a hyperbolic focus run at that time. Back then, I was having all sorts of issues keeping the Celestron focuser attached, so I didn't trust it to get through one autofocus run, let alone 6 autofocus runs through the night. I have since solved the focuser issue (fingers crossed) and have run many autofocus runs in NINA. As you say, a typical autofocus run takes 2 min or less.

[@anata5127]

I started imaging with 9.25 SCT (edge variant). No reducer. I have been shocked about inconsistency of focusing by NINA. It is all over place compared to refractor. I am getting impression that it is inconsistent because of changes in seeing during night. Some info. I use inline Moonlite focuser. It doesn’t have backlash. What do you think?

[@Aero19612]

Probably need to tweak your NINA autofocus parameters. Once I adjusted those, I got very consistent results even with my moving mirror focuser. With the focal reducer I'm using now, I get R^2 values of 0.98 to 1.00 - very good and very consistent. See kzbin.info/www/bejne/g2iqqXecZ9KWrZI

[@anata5127]

@@Aero19612 I did it but it is inconsistent during night. It usually fails 2-3 times per night. Maybe exposure needs to be increased. Some FOV contains very few stars.

[@anata5127]

@@Aero19612 I see very useful video! I will try it.

[@Nickle314]

Carbon fibre can have negatve coefficients of expansion. The magnitude as you say, is very small

[@Aero19612]

Yeah, carbon fiber has some weird traits. Too expensive for me though!

[@Nickle314]

@@Aero19612 I was looking into it for a pendulum clock. The idea being having some carbon fibre for the pendulum rod combined with some other metal. It should be temperature invariant.