No matter what step size, or how many data points I set in the autofocus options, the focus curves are always right shifted. That means there are more focus points to the right, than to the left of the minimum.
Quite often the curve runs out off left points without finding the botton of the curve, therefore the autofocus starts all over again (automatically shifting the first point some steps to the left) waisting precious time…
Is there a setting to set the first point more to the left?
It would be great if SGP notices that there are a few points to the left amissing to finish the curve and automatically adds the needed data points to continue the curve to the left without restarting the whole autofocus from the beginning…
BTW, why does SGP block all functionality while focusing? It’s impossible to do sometihing else on SGP while focusing…
Thanks for the hints.
I have tried smaller and larger step size with the same result.
I think this maybe to little backlash compensation or some loose coupling.
Well check and report back my findings.
Thanks
Martin
So are you saying that if say SGP finds a focus point at 20200 (regardless of how many points there are left or right) then you refocus again the focus point keeps shifting to the left by couple hundred focus points? If that’s the case, the issue lies with the focuser/autofocuser.
I put this down to mechanical contraction as it cools. Annoying, but better to lose 3 mins every sub than lose the subs due being to out of focus. My backlash is higher than it needs to be, I tried to set it exactly using a dti but this didn’t stop the focus point moving as the night progressed, so I increased it to double what it really needs to be. Focus is usually fine.
Not sure this is what your seeing, but I thought I’d mention it.
Apparently you used SGP AF LogViewer. When you load the same data as in the shown AF graph, what is outputted when you click on ‘Position-Temp Graph’ (I’m interested in the line: “Linear regression = …”)?
The temperature sensor is built into the focus motor assembly, so I wouldn’t expect that temperature to be the same as the scope itself. As the scope is a lot bigger, it will take more time to cool down.
You can see it got down to -7deg C (pretty cold for South UK!) but the pattern is the same in the summer too.
I can connect an external sensor to the scope / focuser so I may do that at some point.
I’ve noticed that even after many hours, the focus point is still moving to the left, hence why I refocus before each 30 minute sub. It works for me, it may not work for others.
All I do at the beginning of the next session is reset the focuser position in the software back to 1500 so I don’t eventually get to 0, which it will if I don’t do that.
Without further data (either the order of the points in the graph or a temperature vs. time graph and the information whether and when a meridian flip was performed) one can only have a guess.
Assuming that a meridian flip was performed and ambient temperature decreased continuously, the break might be at meridian flip: before m.f. temperature decreased slowly, after m.f. more rapidly. The orientation of the focus motor assembly toward wind or heat sources may play a role. However, not knowing the conditions, this is only speculation.
It is also possible that the location of the temperature sensor is causing this behavior as you supposed. An external sensor which can be mounted directly on the scope might improve the correlation.
I don’t know for sure, but I do know that there were no meridian flips, wind or heat sources (it’s in a ROR obsy, west of meridian targets only).
I will try the external temperature sensor touching the scope or focuser at some point and see if that makes a difference.
I don’t want to detract from the OP’s issue, but I find sometimes more information can spawn more questions, and eventually an answer.
Just to give you some feedback on my issue that may help others.
I bought a dial micrometer and did several back and forth manual movements of the focuser with the focus motor.
I set the micrometer to 0, and made X steps forward and the same amount backward (backlash compensation was active and large enough). In theory the micrometer should return to 0. This was the case when scope (and focuser) were horizontal, but when inclined the focuser would not return to exactly the same position, instead it was 0.02-0.03mm off. After repeating the procedure, it accumulated another 0.02-0.03mm offset, and so on. Or put another way, the original 0 position was getting further and further away. I still have no new motor to see if this behaviour dissappears, but I am 99% confident that my motor is jumping some steps for whatever the reason (I think due to the weight of my camera, FW, etc).
This is a simple way to check the focus motor behviour.
Martin
BTW my motor is directly connected to the focuser shaft, not to the typical ball reducer, those ball reducers are not geared and may slip, this makes the problem even worse (which I also tested and verified)
