Telescope tube flexure

[SkyHiker]

I have been aware for quite a while that my 12" F/4 GSO Newt has flexure issues. This is easy to see by inserting the Howie Glatter into the focuser and following the location of the laser dot on the primary mirror when slewing the scope around manually. When I rotate it 180 degrees with the DEC axis horizontal, it moves from the outside edge of one side of the donut to the other. This has effect on collimation, coma patterns, plate solving based polar alignment and autoguiding.

About the magnitude of the effect: If the shift of the laser dot is about 10 mm, that corresponds to 3600*((12/1200)*180/%pi)/(12*60)=2.8" per minute (worst case). For comparison, a 1' polar alignment error results in 60*((2*1)/(12*60))=0.17" per minute (worst case). If a 2' polar alignment error is borderline acceptable (0.34" per minute), then the flexure effect is about 8 times worse. The worst case is with DEC horizontal, which is unrealistic, so in practice maybe it is 2 to 4 times worse. Also, the shift seems to go faster in an upright position, in other positions it may just be 1 to 2 times worse. This is still enough to worry about, for astrophotography.

I tried to track down the cause of it:

1) Flexure of the spider vanes in combination by the weight of the offset secondary mirror. There is indeed flexibility when pressing against the vanes around the OTA axis direction. I actually made a counterweight to compensate for the offset mirror weight. It had no effect. Also, moving the vanes took more force than what I suspect the off-axis secondary weight causes. In addition, I rolled the OTA around over the floor on its rings and did not see the laser dot shift much. So, this is not the issue.

2) Bending of the OTA. I found that the laser dot only moves in one direction, namely perpendicular to the DEC axis. This could indicate bending in the plane perpendicular to the DEC axis. I measured any possible flexure using a laser collimator aimed at the mirror that rotated on the inside of the OTA opening, then flipped the OTA doing the same, and checking if the chord on the mirror described by the laser was different. It was not. In addition, I built a tensioner rig using metal wire and turnbuckles to stiffen the OTA. It had no effect. So, there is no bending of the OTA.

3) Cross-section flexure (in the plane perpendicular to the OTA axis). I took the secondary out and projected the Howie Glatter on the other side of the OTA. I found a 2 to 2.5 mm displacement when flipping the OTA 180 degrees with the DEC axis horizontal. That looks significant. This type of sag will cause the spider vanes on opposite sides to be offset in opposite directions and rotate the secondary. If the effect is multiplicative, it would be 2*(1200/(14*25.4))=6.7 mm to 2.5*(1200/(14*25.4))=8.4 mm at the donut, which is about what I see. I do not see the laser dot shift much when rolling the OTA over the floor on its rings, so it happens only when the OTA weight is hanging off the side.

So, it looks like I have a sagging cross-section flexure. The OTA is made of sheet metal that has very little stiffness in that direction. The 14" tube rings are pretty sturdy but have some flexure. So, I think I need stronger rings. Has anybody experienced this type of flexure, and if so, what did you do about it? Does anyone know a good source for stronger tube rings?

Maybe a sandwiched carbon fiber or carbon fiber truss type OTA fixes it, but those cost thousands of dollars, and that is beside the point.

[JayTee]

First off, awesome troubleshooting. My brain is having a hard time visioning tube rings that flex. Are the tube rings attached to a saddle? If so, could it be the saddle that is flexing?

Maybe I just haven't had enough coffee yet to fully understand...

Cheers,

[SkyHiker]

Are the tube rings attached to a saddle? If so, could it be the saddle that is flexing?

The tube rings are attached to a Losmandy dovetail. The tube rings are 14" in diameter and only so thick, so when you press down on them there is some flexure. The OTA is 48 lbs. plus gear, quite a torque when you hold it sideways.

[Lady Fraktor]

https://agenaastro.com/gso-12in-f4-newt ... r-ota.html

Is this the telescope you are referring to?
If so I would put a bar across the top of the rings and check if it is still flexing.
That Losmandy plate that is included does not really look HD enough for a 48 lb tube.

[Petrol]

That's quite a comprehensive and technical analysis of your flexture problem and my brain is also having a hard time of it. What I would say though is, don't rule out cumulative flexture. It's possible the primary mirror is flexing in it's mountings, as is the tube and the secondary held by the vanes. No.1 culprit in my book though would be the focuser flexing on the tube especially if you have extension tubes to accomadate a camera. That's a lot of leverage / weight for the focuser to handle when mounted on a thin, large diameter tube. Iv'e looked at your scope and whilst it's a steel tube, everything has limitations engineering wise.
I may be wrong but I doubt it's the rings that's the problem. Every setup has differential flexture but from my experience of Newts, I'd look at the rigidity of the focuser mounting to the tube.
I hope that helps a little and look forward to how you get on sorting it

Clear skies
Pete

[SkyHiker]

https://agenaastro.com/gso-12in-f4-newt ... r-ota.html
Is this the telescope you are referring to?

Yes.

If so I would put a bar across the top of the rings and check if it is still flexing.

That will do nothing for the type of flexure that I identified. Thanks for the suggestion though.

That Losmandy plate that is included does not really look HD enough for a 48 lb tube.

The Losmandy shoe clamps the whole plate including where the rings are located. Because of the full contact, together they are rock solid.

[SkyHiker]

It's possible the primary mirror is flexing in it's mountings, as is the tube and the secondary held by the vanes.

I checked that. The mirror is held in place by a metal ring that has at most a 1 mm gap. The three retainer caps are rubberized metal clamps with screws that go fairly deep down. It's not flopping around, movement is at most 1 mm.

No.1 culprit in my book though would be the focuser flexing on the tube especially if you have extension tubes to accomadate a camera. That's a lot of leverage / weight for the focuser to handle when mounted on a thin, large diameter tube.

I ruled that out based on my tests where I tried to flex it in any direction manually with a fair amount of force.

Iv'e looked at your scope and whilst it's a steel tube, everything has limitations engineering wise.

Steel, heh I can bend it with my thumb and index finger. It is as strong as sheet metal, just a bit stronger than cardboard, flexing wise. Don't get me wrong, the GSO build is quite respectable because they compensate for the weak sheet metal by providing solid rims at both ends, and the tube rings that are fairly solid. Altogether it's quite good. Just not up to snuff for quality astrophotography IMHO.

I may be wrong but I doubt it's the rings that's the problem. Every setup has differential flexture but from my experience of Newts, I'd look at the rigidity of the focuser mounting to the tube.
I hope that helps a little and look forward to how you get on sorting it

Thanks Pete, I appreciate your thoughts!

[SkyHiker]

That's quite a comprehensive and technical analysis of your flexture problem and my brain is also having a hard time of it.

Let me add some background to the formulas.

First, the effect of polar misalignment on the error rate. Think of the situation as two spheres that are rotating at the same rate with the axes off. Near the poles the error will be near circular, at the equator it will be nearly a vertical line, both with an amplitude the size of the polar misalignment. Thus, near the poles, in 12 hours that misalignment will go from one side to the other. If the polar misalignment is 1' then it will move 2' in 12 hours, or, in arc seconds per minute, 60*(2*1/(12*60))=0.17" per minute.

Secondly, if the laser dot is moving 10 mm over the focal distance of 1200 mm, the angle is 10/1200 radians or (10/1200)*(180/%pi) degrees. If this happens when rotating the OTA over 180 degrees, timewise this represents a time span of 12 hours tracking the sky. Thus, the rate is 3600*(10/1200)*(180/%pi)/(12*60)=2.39" per minute. There was a typo where I replaced 10 with 12, but that does not change it much.

The test was done with the DEC axis horizontal, which is impossible for AP, and the effect is less at "normal" angles. If this is a factor 2, say, then the rate error is 1.2" per minute. And that is approximately tracking error that I see when running the autoguider with the controls turned off. So, the effect of the OTA sagging in terms of equivalent polar misalignment is 1.2/0.17=7.1' misalignment. 1' polar misalignment is quite good, 2' misalignment is borderline acceptable but 7.1' is definitely bad.

So, the math was right except for one typo, and this needs to get fixed. The best thought so far is to add stronger rings to prevent the warping of the OTA.

[SkyHiker]

A glimmer of hope. I moved the rings right next to the focuser on both sides and noticed a significant reduction of the laser dot excursions.

The final test is of course when the scope is mounted, so I ordered a pair of 14" rings from Agena. They are out of stock, so I put myself on the wish list.

I also tightened the spider vanes more by adding Belleville washers under the spider vane screws to the point where the metal flattens out. Below you can see one screw.

I have been searching on CN for more info and it turns out that I'm not alone! Removing the flexure is a big thing in getting a cheap Newt working for AP. It is now clear why quality Newtonian astrographs are so expensive. I am looking for simple fixes, not a complete re-engineering.