Does your polar alignment technique address SCT mirror movement?

[g__day]

This is a question for anyone who has really tried to nail polar alignment on a permanent mount using an SCT - which is exposed to some degree of mirror movement that can confuse the accuracy of your alignment corrections.

I tracked Rigel Kentaurus for about three hours last Saturday - the star was about 1.5 arc minutes off target. Now normally one could do the maths and say well I'm an arc minute off in DEC, but with an SCT life isn't so easy. I went slightly squirrelly earlier this year trying to use PEMPro's polar alignment wizard to nail polar alignment - until I discovered mirror movement was confounding ultra precise alignment.

I switched to a side mounted MAK OTA and got polar alignment much more precise. So now whenever I see drift I wonder how much of it is real and how much is gear or mirror movement?

I have in the past tried MaxPoint to model my pointing parameters - out to 100 stars. My assumption is any mirror shift is consistent all across the sky (a dubious simplifying call).

How do other folk account for this challenge (other than locking your primary mirror in place)? Does Tpoint handle this challenge alot better than MaxPoint anyone?

Many thanks,

Matt

[Gama]

Hi Matt,

Problem with using any pointing precision programs (Tpoint etc) rely on repeatable goto accuracies. It can take gear errors, angles and eccentricities into account, but they are all absolute, and never really change.
Mirror flop is different, as the mirror may not always slip/slop/flop along the baffle tube or its mirror support the same amount every time, so its relative to the last position it was "Sync'd". For this reason your accuracy will never be excellent. An easy test is to setup your CCD camera and image a bright star low on the horizon, then slew the scope to the Zenith (Up) and slew back to the same star using goto, numerous times and see if it centres you star within the ccd sensor by stacking the images together using Overlay or Sum.
What you should see will be similar looking image like a bullseye in a shooting range, where the star images are tightly grouped. This may give you a best fit value to use in your calculation for high precision.

Theo

[GrampianStars]

Mirror shift has nothing to do with polar alignment
Once aligned that's it it its "permanent" and doesn't change when you move to different stars arond the sky It has nothing to do with the mirror
Its the latitude and asmuth angles of the scope
If your getting movement its either flexture of the mount or
IMO you are describing "PE" the slight "squirrelly" drive changes from the worm wheel
BTW LOCK the mirror in place to prevent shift ALWAYS

[g__day]

Robert,

If the mirror is shifting - well that could throw off a polar alignment by my reckoning. So when I switched back and forward to a side mounted MAK and got different polar alignment correction data (from PEMPro's polar alignment wizard) on the same target at the same elevation to me that says the mirror is moving. The PE PEMPro describes as under 2 arc seconds corrected - which I find hard to believe. The C9.25 unfortunately has no mirror lock (unlike the Meade) but the Feathertouch focuser helps reduce mirror shift.

So if I see DEC drift on one OTA during polar align but not another side mounted one (Losmandy side saddle bars) - something unusual is happening!

[GrampianStars]

G'day Matt
check out http://www.sciencecenter.net/hutech/misc/focuslk/index.htm
for the Focus locks for Celestron :thumbsup:

[Tandum]

Why would unreliable focus effect pointing? Wouldn't you be pointing to right place but just be out of focus?

[g__day]

Robert,

You can't fit both a Feathertouch microfocuser and a Hutech focus lock - the microfocuser does most of the job of the focus lock.

The mirror shift is tiny - so it isn't effecting focus - focus isn't my problem - its affecting where the star appears - if its wandering a few arc seconds left or right, up or down - this is throwing out correct drift alignment adjustments. So if the mount is a tad off polar alignment you'd expect point and therefore tracking (which is just real time pointing to be off).

Matt

[JohnG]

Hi Matt

Is everything setup square to each other .......

Errors can be introduced where the axes of the mount and your telescopes are not perfectly orthogonal.

Most setups have slight imperfections in this regard.

Cheers

JohnG

[g__day]

John,

I would say as perfect as Losmandy and Vixen can machine their gear. MaxPoint reports on a 100 star alignment that the non-orthognal issues are generally in the low arc seconds - say 5 - 15 arc seconds and accounts for this.

This is very different from tracking a star that is due East about 30 degrees above the horizon. After 10 minutes PEMPro saying the mount is your dead on for one scope whilst the other side-saddle mounted scope given exactly the same test reports the mount is 2 arc minutes off SCP. This reads like mirror shift on one of teh OTAs - the larger one - to me. Non orthogonal issues shouldn't be able to affect a test where the star is individually tracked - RA and DEC movement on the OTA is determined and arc seconds per pixel worked out by examining 10 second star trails then a star is tracked for 10 - 30 minutes at known bearing to compute drift correction required.

[gary]

What is often an issue with unlocked mirror shift in SCT's is the shift can occur suddenly
in a discontinuous manner. As a precursor to this, one will often see a small continuous
shift occur followed by a discontinuous larger shift as stiction is overcome.
Unfortunately this sudden 'klunk' is, for all practical purposes, impossible
to correct using a software approach.

During the systematic precursor phase, the amount of defection is a function
of the mirror's orientation with respect local gravity and is not constant. In other
words, it is a function of both your mount's RA, Dec and the local observing latitude.
If this systematic part was all that you had to deal with, you might be in a position
to able to correct it in software.

If you have it, the discontinuous, non-systematic and often seemingly random
shift component is only possible to correct by taking some hardware intervention,
i.e. by locking the mirror in some way.

Based on our own analysis work, we often see sudden mirror shift occur in larger SCT's,
in particular the C14, and less often in smaller aperture SCT's, such as in the
8" aperture class.

If you are serious about trying to determine the source of errors using a software
approach, you should indeed consider switching to TPoint. Though TPoint will not
be able to correct any sudden mirror shift, if such a shift is indeed occurring, careful
use of its graphing and data point masking capabilities combined with
some knowledge of statistics can often help you 'catch it with its pants down'.

In any case, TPoint's richer set of terms covering a broad gamut of systematic
errors means it is the best choice. I know through experience you will
most probably have several other significant systematic errors on your mount that
your current analysis software does not have the capability to uncover.

With this last point in mind, though mirror shift might be a candidate, I wouldn't
point my finger at it until you have done some further TPoint analysis.

Finally, as I have often posted on this and other forums, never lose sight of
the fact that a 'perfect polar alignment' does not exist. There is only an optimum
alignment for the part of the sky you happen to be observing in at any one
instant.

Best Regards

Gary Kopff
Managing Director
Wildcard Innovations Pty. Ltd.
20 Kilmory Place, Mount Kuring-Gai
NSW. 2080. Australia
Phone +61-2-9457-9049
Fax +61-2-9457-9593
sales@wildcard-innovations.com.au
http://www.wildcard-innovations.com.au

[gary]

Matt,

Keep in mind that for the change in zenith distance of Rigel Kent from an observing
location near Melbourne over a three hour run, a back of the envelope calculation
shows that you might get as much as a 2 arc minute apparent shift in altitude
from refraction alone, for an observing run from about midnight to 3am at this time
of the year.

As I mentioned, there is no such thing as "an exact polar alignment". Your
best compromise for an extended session is to align the mount to the refracted
pole.

Best Regards

Gary

[g__day]

Gary,

Huge thanks for your replies. I know the SkySensor2000-PC is meant to calculate and handle refraction - its old fashioned but smart.

I ponder is there any polar misalignment or mirror shift that can manifest solely as an error in RA rate rather than DEC? This stumps me a bit - the shift is always along the RA axis - which to me reads like a tiny tracking rate error (like its one part in a thousand faster than it should be).

I track a star for 30 minutes and see maybe a 1-2 pixel drift in DEC (so say an 1-3 arc seconds or 25% of a star diameter out of position) but in RA it is likely to be 4 star diameters out of position in RA.

I am trying now to see if I shoot a series of 30 * 10 second shots - spaced a minute apart over half an hour how much drift I see in RA and DEC and determining:

1. If I change the elevation of the mount slightly - does it improve or worsen
2. Is the error growing, shrinking or consistent with elevation.

Data empowers one to diagnose the source of the problem I guess!

Matt

[gary]

In a nutshell, no. Polar misalignment and mirror shift will bring about apparent
shifts in both RA and Dec.

There are other error terms that will be predominantly in the RA axis,
notably Collimation in Hour Angle (CH) and Non Perpendicular Axis (NP)
error terms. However, since your mount is polar aligned and since
you are imaging a star with a constant Dec, these will not be involved
in this instance.



Do you see the same magnitude of error in both RA and Dec for all zenith distances?
I see further below the answer to this may be that you are unsure at this stage
since you have insufficient data.

The reason I ask is that even if your pointing computer is performing refraction
calculations, it won't be adjusting the actual elevation of the mount to account for
refraction. All it can do is change the tracking rates in both RA and Dec.
Are you auto guiding?

Best Regards

Gary

[g__day]

Gary,

No I am not auto-guiding at all as I want to see the drift and measure it. I have done PEC - its down to around +/- 3 to 4 arc seconds. I am on my second shooting run now.

The first shooting run tracked Achernar from 9:30am - 11:30am in Sydney. So it moved for two hours when it was about elevated about 30 degrees above the horizon until it got to about 60 degrees. In the first hour it drifted maybe half a star diameter due East and down (RA was too fast) about 2 star diameters. In the second hour at higher elevation it drifted the same in DEC but I would say 1.5 star diameters in RA (again mount too fast)

Now I am tracking a star (Zarauk) which is almost due East about 30 degrees elevation - so any drift will be much more apparent. This star is 1 mm across in my shots. After 30 minutes its drifted 0.5mm in DEC and 5mm in RA (mount is still too fast). At 45 minutes the drift is around 7mm in RA and still under 1 mm in DEC. By an hour it had moved around 10 mm.

To give you my precise scale - I shot the Moon yesterday - on this monitor it is 293 mm in diameter equating to about 31 arc minutes (by comparison Achernar was 3mm and Zarauk a bit less than 2mm on my shots).

Day 2 - Imaging Atria on Celestial Equator near the Meridian

After a very slight adjustment East tracking now shows DEC in drift about 1 star diameter per half hour and minimal RA drift - so stars look slightly stretched on a 10 minute unguided shot.


Matthew

[AlexN]

do you have a scope that isnt affected by mirror shift? I drift align using my refractor, then pull the refractor off put up a different scope - the mount is still aligned..

[g__day]

I'm thinking of only using the 5" MAK for this - it seems alot more stable!

[g__day]

I re-ran a 80 star alignment with MaxpPoint on my main SCT - it says 14 arc seconds too West and 3 arc minutes too low.

Now that 3 arc minutes too low surprises me after the hours and hours I've spent drift aligning - so that could be mirror shift in the SCT.

I plan to repeat the alignment run on the side-saddle 5" MAK once the skies clear.

Another interesting factor - pointing before MaxPoint run was around 3 arc minutes - with Maxpoint 1 arc minute - but the error is all along one axis - like a sausage - maybe indication of slight slop in the DEC gears still?

[Geoff45]

No, it just affects the goto but not the tracking. Think about using a guidescope. It's common practice to move it around a bit to find a good guidestar, which is equivalent to mirror flop, but you still track accurately. With mirror flop, you will be off when you use the goto, but once you have centered the object, it should track accurately.

[g__day]

Not if the mirror is changing position as the mount changes position - then you have two factors to account for by my thinking.

[AlexN]

the mirror flop has nothing to do with the tracking accuracy of the mount.. The mounts tracking may cause the mirror to flop, but this does not throw out the mounts alignment or tracking, its just something inherent in the design of the SCT...

You cant set the mount to compensate for the mirror flop/shift.. The mirror however shouldnt shift or flop unless its flipped from one side to the other, Ie, during a cross medrian slew.. or perhaps slightly as the tube rotates while tracking an object the mirror might change its position slightly.. Unfortunately, appart from focusing the telescope, then permanantly fixing the mirror in that spot, I think its just something you're going to have to put up with...

[g__day]

Problem pretty much solved I think - manually setting the SkySensor2000-PC's alignment parameters (which I guess are a starting RA, Starting DEC, RA tracking adjust, DEC tracking adjust) third setting from 000 to -007 has resulted in near perfect tracking in RA :) So on a 20 minute unguided test shot of NGC 2070 at 1,500 mm focal length drift is almost totally missing - a big success!

My earlier question was really could mirror shift upset polar drift alignment enough to lead to a bad set-up that couldn't be corrected!

Thanks all

Matthew

So here's a 10 minute shot - some coma from the focal reducer (that's my next adjustment - I don't think its seated square in the OAG tube) - but very little drift in RA or DEC I believe.

[AlexN]

Geez... 10min unguided exposure like that... Looks fantastic.

Glad to hear you got the problem sorted Matthew...

[Matty P]

I agree. :)

A 10 minute unguided exposure with no apparent drift is quite an achievement. I wish my EQ6 could perform like that.

Great work Matthew. :thumbsup: