Re: GOTO Accuracy


Christopher Erickson
 

One of the scopes I am responsible for is a 20" RCOS on a 3600GTO and it suffers from a problem that is similar to what you are describing.  I have carefully studied every truss junction and structure fastener and everything is torqued down to appropriate levels.  I have narrowed the problem down to the support systems of the primary and secondary mirrors and right now I am mostly suspicious of the primary mirror baffle tube support design.  I believe it is flexing/sagging as the weight of the primary mirror is shifting as the scope points around the sky.
 
I hope this helps.
 
 
Christopher Erickson
Consulting Engineer
Summit Kinetics
Waikoloa, HI 96738
www.summitkinetics.com
 


From: ap-gto@... [mailto:ap-gto@...] On Behalf Of Joseph Zeglinski
Sent: Thursday, April 17, 2014 3:36 PM
To: ap-gto@...
Subject: Re: [ap-gto] GOTO Accuracy

Thanks Chris,
 
    One suspicion I have,  about the cause of “truss flexure”, such as on RCOS scopes – I wonder if the (pinned)  truss poles move slightly out of their truss pole sockets. I realize that the truss design should be rigid, based on mechanical engineering of truss plates, and their triangular arrangement – top & bottom,  push & pull strain forces are supposed to balance out.
 
    Don’t know for sure, but I thought the RCOS carbon fiber truss members were “supposed to be” epoxied in place, but I have my doubts, since I see each pole is also “pinned” through the pole base, with a tiny (~ 2-56) Allen screw at each end socket. That would make sense for easy pole replacement if necessary.  If not epoxied, the hollow carbon poles might “stretch” out of their pin holes. Even that tiny movement might get magnified to a 28 arc-minute “lean” on either side of the meridian flip, given the approximately one metre of secondary cage moment arm. Don’t know if Jeff’s solid carbon fiber tube might have similar cause to shift out of its (drilled/pinned) rings. Such movement might account for the GOTO accuracy we have been discussing ... among other more common causes of course.
 
    Just a thought,
Joe
 

Sent: Thursday, April 17, 2014 3:50 PM
Subject: RE: [ap-gto] GOTO Accuracy
 


FWIW, I take care of a lot of university scopes and shimming the dovetail bars works well for tweaking optical axis orthogonality.
 
I have also found that Meade, Celestron and even some of the "premium" RC scopes available have a lot of orthogonal variability (slight sag, flexure or flop) when moving around the sky and that plate solving usually ends up being the easiest solution to robotic-operation pointing errors.
 
I hope this helps.
 
 
Christopher Erickson
Consulting Engineer
Summit Kinetics
Waikoloa, HI 96738
www.summitkinetics.com
 
 

From: ap-gto@... [mailto:ap-gto@...] On Behalf Of Joseph Zeglinski
Sent: Thursday, April 17, 2014 4:34 AM
To: ap-gto@...
Subject: Re: [ap-gto] GOTO Accuracy

Thanks for the explanation Rolando,
 
    And also Chris for the suggested solution. I have done many collimations on the truss, in an effort to at least lessen the error, and the meridian flip error remains “precisely” the same. So, in my experience, re-collimation had no effect on increasing or decreasing the error offset. The collimations were done at zenith, using CCDInspector. I have also used a wrench to check the tightness of all the integral dovetail/OTA attachment screws, as well as all others on the truss poles, for good measure.
 
    However, Rolando, even if Chris’s suggestion would (hopefully) work, it should make no difference to your observation that “ The tube begins to become a very slight banana shape curving away from the mount”.
Unfortunately, even with a shim under the “integral dovetail”, the OTA would still go bananas, from its now compensated offset :-)
 
Joe
 
Sent: Wednesday, April 16, 2014 11:08 PM
Subject: Re: [ap-gto] GOTO Accuracy
 


"As for OTA's with an integral dovetail bar, I loosen the screws on one end of the bar a bit and insert different thicknesses of shims between the bar and the OTA structure"
 
Well that sounds like a really great idea.
 
Rolando
-----Original Message-----
From: Christopher Erickson
To: ap-gto
Sent: Wed, Apr 16, 2014 9:06 pm
Subject: RE: [ap-gto] GOTO Accuracy



Just a general FWIW, I have found that adjusting collimation also changes orthogonality and I check it after every recollimation.
 
As for OTA's with an integral dovetail bar, I loosen the screws on one end of the bar a bit and insert different thicknesses of shims between the bar and the OTA structure and then re-tighten the screws.
 
 
Christopher Erickson
Consulting Engineer
Summit Kinetics
Waikoloa, HI 96738
www.summitkinetics.com
 
 

From: ap-gto@... [mailto:ap-gto@...?] On Behalf Of Joseph Zeglinski
Sent: Wednesday, April 16, 2014 1:52 PM
To: ap-gto@...
Subject: Re: [ap-gto] GOTO Accuracy

Rolando,
 
    Something I don’t quite understand in your “gravity” flex comment.
 
    On a meridian flip, the before and after position is just about “vertically”  identical, so the effect of gravity shouldn’t account for a repeatably precise image shift. I have seen this on my meridian flips at all “OTA elevations”, even for a target right at zenith, where the mirror and spider end up perfectly level, anti & post meridian. So, I think,  any shift should end up zero, even if it rattled all over the place during the entire mount meridian reversal contortions. Even if I wacked the OTA in that zenith position, there isn’t enough saddle looseness to shift it back the requisite 28 arc-minutes.  Only thing left is Orthogonality, which is a  likely culprit.
 
    Sorry for introducing some  confusion, but unfortunately (unlike Ross’s AT8RC “closed” carbon tube),  my RCOS RC “open truss” has no rings to be shimmed – it’s fixed D-Plate is solidly bolted to the truss frame. Might be worth trying slipping a shim under that D-plate, but I don’t relish loosening the saddle lock knobs and lifting the 85 lb. OTA to do it. I will probably see if the mirror optical axis can be better aligned, though I have tried drastically shifting the secondary spider with its “OTA ring attachment” screws, with no visible difference.
 
    Finally, your suggestion to “Add an RA offset” in ASCOM – will that work both ways? Does ASCOM subtract the offset if the meridian flip is switched back to the other side? It would be great if it does – I wouldn’t need to use APJOG to do it manually every time. I need to think about that.
 
    Pointing Models – haven’t tried one yet. If a model is created using the east, then the west halves of the sky during training, then if a Meridian DELAY is in effect, and the OTA is now pointing somewhat past the meridian, wouldn’t the model become ineffective to fix this meridian flip problem by software? The mount would be pointing on the post-flip side of the sky, but it has not yet mechanically flipped, so the “trained” model and Meridian Delay position no longer agree – (hope I am explaining this right).
 
Hope the weather gets warmer soon, so I can go back to retesting (my) meridian flip problem.
 
Thanks Rolando,
Joe
 
Sent: Wednesday, April 16, 2014 1:09 PM
Subject: Re: [ap-gto] GOTO Accuracy
 


You suspect wrong. Even a carbon fiber tube does not eliminate flex. The secondary sits in mid-air and is held in place by a set of thin vanes. That secondary can move ever so slightly with gravity, and can do all kinds of wonderful things. In other words, the exact optical axis of your telescope is not reliably orthogonal to the exact mechanical mounting point of your dovetail plate in all orientations. This is not the only source of flex. The main mirror itself can flext slightly, and it does not take much in an RC. The focuser end can flex, as well as the camera connection. Lots of things add up.
 
Rolando
 


Join main@ap-gto.groups.io to automatically receive all group messages.