External focuser on an SCT; What about temperature change induced focal shift and back-focus?

David Jones:
if needed. I’m thinking 7.5mm in each direction should cover a large temperature swing

John Stone · Mar 25, 2026 at 08:45 AM

I'm sorry but I think you're missing the point I'm raising.    Did I misunderstand you?

It’s not you, it’s me. I haven’t sweated that tolerance but I don’t have the data nor have I compared the results across the images of the night if that is a wise choice or not. I need to do that and reassess. I’ve been looking into the Optec and the IF ONAG as much better solution - just a budget constraint for now. I think I can tolerate the spikes but would attempt the curved spider veins On that and to back up Andrea’s statement Gaston’s write up https://www.innovationsforesight.com/support/celestron-edgehd-back-focus-tolerance/ shows the mirror spacing tolerance (large) but getting that requires either the Optec or, if memory serves, someone (maybe John Hayes) had worked up a solution for moving the primary on a C14 with support so that mirror flop could be avoided it worked well - I think the assemblies vary but principles could apply.

I ran a non-Edge C8 for a while and managed with a motor focuser and was either fortunate or oblivious to any flop significance.

An electrical mechanism for mirror locks might prove interesting if successful but would shift the need a focus motor for the primary and eliminate the Esatto.

Dave

John,

You raise an important point regarding external focusers. basically what you have to do is make a guesstimate of the average temperature of the night and the temperature spread, and start on the short end getting longer over the night (as the tube contracts). usually from previous focussing runs you should have a pretty good idea how much the focuser moves per degree over the night. for my EDGE 11 and the Esatto2LP I get some 6000 steps per 1C, which corresponds to a bit less than 0.2mm (and is consistent of what you would expect for the thermal contraction of aluminum times the SCT amplification of 25). and I get typically a spread of 5C over a typical night. I have the back focus set with the focuser in the middle, so at ~210 000 steps.

So shortly before I start my run I put the focuser to 195 000 and do a focus by hand with the focus knob on the main mirror - does not need to be super exact, basically try to get HFR for a typical night. Then lock the mirror. and run the focussing routine. After that I run autofocus once per hour, and over the night the focuser moves from 195000ish to somewhere around 225000 or so.

the ESatto moves very smoothly, so you can even do a first order correction over the night, is shift after every exposure the focuser by 6000*Delta T (take a fraction of a second).

Hope that makes sense.

So nothing for a remote setup and problematic in desert climates, where Delta T over the night can be considerably larger than 5 C.

Summary: it works, but not the ideal solution. In hindsight, starting anew I probably would go for the shifting secondary (pay once, cry once)

Matthias

I wonder if this is all chasing a red herring. Yes, there is a perfect back focus position and Celestron tells you what it is but do they tell you what the temperature of the measurement is?

As the entire optical responds to temperature changes I would expect is that the focus position, even when it changes is still the best position for that temp. Have you not found that to be true in practice? I can’t see changing changing the back focus (somehow) during a run. If that needs to happen then I would consider the design unsuitable for astrophotography.

The “perfect” BWD isn’t given by Celestron for your telescope. What they provide is the design spec. Small manufacturing variations mean that the best spacing varies a little between telescopes. The correct position is the spacing that minimizes field aberrations and the tolerance on the spacing is determined mainly by the size of the sensor. The small variation due to thermal variations won’t make any noticeable difference-even with a fairly large sensors. I gotta run but I can spend more time in another day or two going over this if you have more questions..

John

Spacey,

Thanks for you post but you and I have differing opinions on this.

Celestron’s optical designers specified a back focus and tolerance that meets their published performance specs. Whether or not that back focus distance is exactly 133.35mm ±0.5mm for my particular telescope is immaterial for this discussion.

I’d like to always run my equipment within these published guidelines and the main point of this topic is how are people doing this with the ever popular external focuser on the back of the SCT in the face of temperature induced focal plane movement?

It seems like people have been adding external focusers on these telescopes for decades so I’m surprised this isn’t a front-and-center issue that’s been solved somehow.

So far it seems that most people are ignoring or minimizing this issue when using external focusers except for GalacticRAVE who works around it by starting the night with his back focus on the -0.5mm inward side and lets the dropping temperature move the focus plane through Celestron’s 1mm of tolerance to end the night at +0.5mm on the outward side which gives him an 8°C temperature window (measured on his scope) in which the telescope is operating within spec. This does require him to reset the mirror on the nights when the cumulative temperature change exceeds this range.

I’d like to move this telescope to a dark sky site like StarFront or SoFar2.com but this temperature induced shift is throwing a wrench into the works for me since I’m not there to reset the mirror.

I think the right solution is to either;

A.) Create a motorized mirror lock/unlock system and then add code to your image acquisition software that monitors the external focuser’s drawtube extension and performs a “reset” with the movable mirror when the back focus goes out of tolerance (you decide how much that is for yourself).

Or

B.) Go buy an Optec SMSF focuser and fix your back focus with extensions tubes. The problem with the Optec for me is twofold;

1.) You have give up your diffraction spike free telescope (I hate diffraction spikes; religion I know)

2.) They aren’t made for the 8” or 9.25” model so that’s not even an option for me

Option A is what I’ve been doing but it’s a manual process. I have to go outside each night when starting up (and sometimes in the middle of the night if the temperature changes a lot), point the scope to the zenith, unlock the mirror, reset the drawtube to the proper back focus, focus with the movable mirror, re-lock the mirror, and continue my imaging run with the precise external focuser.

I’ve noticed a significant improvement in the repeatability HFRs across focus runs and generally more consistently and smaller HFRs throughout the night using the Esatto so it really does make a difference to me.

I appreciate the effort you are putting into this and I’m interested in the results you ultimately get to see whether I would go down the same road.

The back focus is specified but I don’t see it as absolute. What I mean is we have to focus the instrument and in this case it means the main mirror must be moved and that alone means you cannot maintain a set length back focus.

If you focus the instrument and then reset the backfocus distance to the published spec, you’ve then unfocused the instrument. You cannot maintain backfocus length and focus at the same time.

It practical terms if the camera sensor is within a few mm away from the optimal point the only distortions you will see will be near the edges of the frame and you’ll be cropping those out of necessity on this scope imaging across multiple nights. You may try to experiment with spacers and save yourself time. You could purposely alter the backfocus by 1-2 mm and compar your stacked results with a setting closer to ideal. I think you will not see any difference.

I’m interested in seeing an HFR trend with what you are doing if you can post in the future.

I see stable HFR across a meridian flip on my setup provided conditions are good. On good nights it stays between 2.5 and 3 all night.

The temperture induced changes affect this instrument in a non linear fashion. I don’t see a gradual rise in HFR like I do with my refractors. It stays stable for long periods and then changes dramatically all of a sudden as the tension in the optical tube gets to its breaking point and causes a focus change.

The mirror spacing can be adjusted to bring the system into focus at a wide range of positions behind the telescope. However, the BWD is a basic design parameter that sets the mirror spacing to minimize field aberrations. (Let’s not use the term “distortions” to describe field aberrations since optical distortion is a specific 3rd order aberration.) When you have the sensor positioned correctly, the system will come into focus when the mirrors are properly spaced. If you put a focuser on the back of the telescope (which isn’t the best way to focus the telescope), you have to lock the mirrors into a fixed position and that should be done at the spacing that minimizes field aberrations. When you use the focuser, it should only change focus by a small amount to track thermal changes, but keep in mind that the focus position changes as the square of the optical magnification of the secondary mirror times the change in space that is induced by thermal changes. Celestron SCTs have an optical magnification of 5x so the focal plane position changes by 25x the amount that the mirror spacing changes. That makes these telescopes very sensitive to thermal variations but in spite of that, the variation in field aberrations with temperature change is minor and it’s not something that you will likely be able to easily see.

As John Stone mentioned, the best way to focus these scopes is to use the Optec SMFS with the mirrors and the camera sensor locked in position. @Niall MacNeill wrote about setting the proper BWD on his C14 Edge in great detail under one of his images (I can’t recall which one). He demonstrated that for a C14 Edge with a 16803 sensor (36 mm x 36 mm), the tolerance on the BWD was around ± 50 microns (as I recall). The tolerance goes down rapidly as the sensor size gets smaller. I’m not sure what it would be for a C8 Edge so you’ll have to experiment to see what it is on your system. (I’ve got the results from a ray tracing calculation somewhere for a C14 Edge but it won’t apply to a C8 so I’m not going to look for it. I posted a lot about this subject years ago on CN but it’s probably been archived.)

John

After spending some time using an external focuser on my C8 Edge HD, I switched back to primary mirror focusing and I’m quite happy with the results. As John Hayes pointed out, the best option would be the Optec system, but it can be expensive or difficult to implement for users who are not technically inclined.

When you lock the mirror and use an external focuser attached to the busines end, each frame after a focus adjustment ends up with a slightly different focal length. For example, with a C8 Edge, one frame might be at 2125 mm and the next at 2132 mm, and so on. The plate scale changes with every focus adjustment, and aberrations begin to creep in as you move away from the designed spacing between the two mirrors. When registering these images, they also need to be interpolated.

With sufficient backlash compensation and a Starlight microfocuser, I’m achieving very consistent focus and a stable plate scale, so I’m quite happy with the results.

Below is a photo of my current setup in my remote observatory where you can see the focuser setup.

📷 C8 Edge Focuser.JPG

John, nowhere in this thread have I seen any anything showing the effect of the issue on actual images or even spot diagrams. I would want to be really sure I wase’t tilting at windmills before I would be too concerned. I think this line in your OP sums it up:

“I've never heard this discussed before”

The reason might be that it’s an insignificant effect. I don’t know that it is but I would want to be sure it was before trying to fix it.