TPoint Flexure War – Part 2

Went to Paul Lind and designed/built a new focuser bracket for the Sky90. This bracket attaches to the Sky90 finder scope position, so now the telescope can be rotated in the bracket. This in turn allows the camera to be positioned so it doesn’t need counterweights.

Took everything apart and rebuilt. Ended up moving the Paramount Versa plate up as far as it would go, then reversing the Homeyer cradle and moving it farther up the Versa plate. After a couple of iterations, managed to get the Edge11 mounted so it is balanced. I have the 4 bolts holding the cradle to the Versa plate all in the front slots of the cradle; the Versa plate doesn’t give access to the back slot. Seems like it holds fine. Connected the E11 wiring and balanced.

Worked on getting the Edge11 rotated so the top dovetail is on the axis. Don’t want the Sky90 leaning to one side:) I can line up bolts centered on the E11 with bolts centered on the cradle to get it correctly aligned. Put a level on the dovetail and it looks good.

Mounted the Sky90 on top, with the wiring. Turns out the weight of the lens caps affects the balance, so I balanced everything without lens caps. This means I need to remove the E11 lens cap even though I am not imaging with it.

Nice – everything balances neutrally without the extra 25 lb weight room counterweight I had been using. I suppose that could also have been an issue since it does not clamp to the bar. I can balance everything without any counterweights on the OTA – nice. I left the maple dovetails on the cradle, but nothing is mounted on them.

First shot at TPoint model shows a better raw scatterplot. Many of the outliers are the first 12 samples which I do manually before letting AAG_TPointMapper do its thing. The overall distribution seems better.

RawScatter

The orthographic shows a scattering of error.

364ModelOrthoErr

The supermodel, however, seems to again get squeezed North/South. Not sure why that is.364ModelPolar

 

364ModelMultiple

And, I still see the drifting of a star in RA/Dec.

DriftingWest

Next step, get the polar alignment better so I can eliminate that from drifting concerns, then redo TPoint.

Advertisements

TPoint Flexure War

The initial telescope setup has strange flexure? issues – the TPoint scatterplot has a very distinct N/S distribution. I suspect the various counterweights, which are necessary because the focuser is attached to the rings, which means I cannot rotate the scope. The camera is all threaded connections, resulting in the camera being sideways (assymetric to the scope stack). This in turn requires significant counterweights (about 12 pounds) to balance the scope. I’m not convinced it is well balanced in the end anyway, and probably leads to flexure issues.

Photo Jun 23, 10 23 26 PM

Here you see the large SBig2000XM and the heavy filter wheel. The camera is all threaded mounting, resulting in the filterwheel/camera being sideways. This in turn requires counterweights.

Note 10 lb counterweight to offset SBig sideways camera

Note 10 lb counterweight to offset SBig sideways camera

Here you see the 10 pound counterweight attached via my maple dovetail. The weight is likely an issue, and it might flex on the bolt, and the wood structure maybe flexes as well.

Photo Jun 23, 10 22 16 PM

The 10 pound counterweight needed another counterweight on the opposite side. These two 2 lb weights bolt onto the aluminum dovetail, but might not be as solid as one would like.

Another reason for the 10 pound weight is to counterbalance the Edge11. This scope is crazy back end heavy, necessitating the heavy counterweight. I may be able to reposition the Homeyer cradle to balance the E11 by itself, removing the need for some of the counter weight.

To summarize:

  1. Need to shift E11/cradle
  2. Need to rotate Sky90 so SBig is symmetrically positioned instead of sideways. This requires a new implementation of the focuser on the Sky90.
  3. Eliminate use of maple dovetails.
  4. Minimize counter weights.

All of this yielded the following 354 point Tpoint raw scatterplot

rawdata

RawOrthoA lot of the poorer data points are in the North.

Here is the resulting supermodel. A lot of the northern points were thrown out

supermodel

In practice, this would probably be fine for imaging. Maybe some of the northern slews would be off, but I suspect ACP could realign fine. However, I think this is bad enough to prevent ProTrack from guiding.

Various Issues

I drove up to New York and picked up all of the electronics, cabinets, etc for completing the installation of telescopes into the observatory.

Got everything installed, and began to discover various issues. I will deal with each issue in detail in separate posts.

First, using a dome brings in a number of problems. First, the telescope stack is big – everything just fits inside the dome. The top Tak is within a foot of the dome wall. This makes it tricky to see out of the slit, especially to the South and North. The scope looks out the slit at a very oblique angle.

My existing equipment will not work in the dome. Specifically, I had previously had the Tak and William Optics Guide scope mounted on the sides of the Edge11. This does not work in the dome – first, the geometry requires the scope to be centered over the RA/Dec axes. It cannot be offset to the side, the calculations do not take that into account. Thus, a side positioned scope will often not be able to look out the slit.

This implies the William Optics cannot be used, I can’t mount three scope vertically, because a) I do not have the hardware to do it, b) if I managed it, flexure would be terrible, c) the stack will not fit inside the dome, and d) I can’t get two scopes to look out the slit. Well, I might be able to get the Tak and WO, but not the WO and E11.

I got an ST-i guide scope and lens kit, which is supposed to provide a very small, lightweight guiding option. Had problems getting it to work; somehow the Windows system was not properly handling the USB drivers. Eventually I ran Windows Update, did a bunch of updates, and the ST-i started working. Now I find that in fact, the ST-i is not usable as a guide scope. It has an image scale of 15.2 arcseconds per pixel, which is way too big for guiding the Sky90 accurately, much less the E11. It sounds like the camera itself is nice, but the lens piece is not appropriate at all. I don’t know what SBig is intending with this setup.

So, the Tak will need to guide using the internal ST2K chip. Alternatively, if I can get the Paramount set up accurately enough, maybe the ProTrack system will remove any need for guiding.

The Edge11? Once I decided to remove the WO guide scope, I had intended to guide using the sky90. However, this doesn’t look feasible – I don’t think I can get both scopes to look out the dome at the same time. I think I will need an Off Axis Guider (OAG) to guide the E11. I will use the ST-i on the OAG, once I figure out which one to get.

OK, other issues.

Dome Operation

It has been a real pain getting the Dome shutters to work. See separate post. They still do not work correctly; I have to be there to close the shutters because they physically do not close right. The software is wacky, has a poorly designed interface and various bugs. I am currently working with a beta version of 2.1.3. It is impossible to get hold of Stan, their forum has very little traffic.

Paramount Problems

See separate posts.

1. The initial setup has strange flexure? issues – the TPoint scatterplot has a very distinct N/S distribution. I suspect the various counterweights, which are necessary because the focuser is attached to the rings, which means I cannot rotate the scope. The camera is all threaded connections, resulting in the camera being sideways (assymetric to the scope stack). This in turn requires significant counterweights (about 12 pounds) to balance the scope. I’m not convinced it is well balanced in the end anyway, and probably leads to flexure issues.

2) Odd thing, if the scope is just tracking without any guiding, then both the RA and Dec positions steadily drift over time. Guiding will remove this, but clearly ProTrack does not handle it. I tested both with and without ProTrack running and the results were identical. Bisque thinks this is also due to flexure? They clearly don’t like my setup and will blame it for all problems.

3) I still do not have a perfect polar alignment. I doubt the slight offset from the pole is enough to cause the drift I am seeing, but I want to get it more accurate. It seems like every time I run a TPoint it changes the offset instructions. I don’t know if a 25 point run gives the same results as a 50 point run as a 300 point run. I guess I need to test some of these options to get a better feel for the results.