Drift Problem Continuing

After more than a year of trying, I am still having the same problem with my Paramount MX mount “drifting”.

The Goal: I would like to be able to do 10 minute unguided exposures using a wide field setup. My system has two telescopes and cameras:

Widefield: Sky90 refractor with SBig ST2000XM camera, 10 position filter wheel. Image scale is 3.7 arcseconds per pixel unbinned. The camera has a second guiding chip so sometimes I can guide. However, the various filters are pretty dim; even the regular RGB filters dim out guide stars. HAlpha, Oiii, and Sii need a very bright guide star. This is why I am looking to do unguided exposures. Various people have assured me that this mount should be capable in this configuration. As seen in the pictures, the Sky90 is mounted on the C11 with a wide Losmandy rail.

Narrow Field: Celestron 11″ mounted in a Homeyer cradle for stability. Uses an Optec focuser with a SBig STF8300M camera, 8 position filter wheel. Has the STi off axis guiding option. Image scale is 0.40 arcseconds per pixel. Guiding works fine with this setup, so unguided operation is not particularly needed. Might be nice, since I sometimes have to play around to get a guide star on the chip.

The Problem:

Without guiding, using the widefield setup, my stars drift (mostly in the RA direction). I end up with distinctly trailed stars in a 5 or 10 minute unguided exposure. This occurs with or without ProTrack running; I have a 350 point Tpoint model in place, with pretty good model results (I think:) as shown below. It appears that the drift is bigger than what ProTrack can handle.


To demonstrate/measure the problem, I run tests on the C11 (just to get more precise data at 0.4 as/px). I set up to autoguide using the STF8300 main camera (0.4 as/px). In Maxim I disable the guider relay outputs so the actual guiding corrections are not sent to the mount. PEC is enabled. I can then start “guiding” and watch the guide star’s movement over time.


The screen shot above is a typical example. In this 5 minute exposure I am looking at a star in the East. The RA guide error on the Y axis (my camera is turned 90 degrees to help balancing) steadily drifts about 20 pixels (8 arcseconds) over 300 seconds, for a net drift of 0.027 arcseconds per second. This same drift rate on the wide field setup results in a displacement of 4-5 pixels on the widefield system for a 10 minute exposure. I generally see 5-10 pixel star trails on the widefield system.

When looking at a star in the West, the RA drift may or may not be in the opposite direction. This example shows the drift in the same direction.


Note that there is also a slight drift in the Dec. In different tests, the Dec drift is either slight or negligible.

I know, the general buzzword is always Flexure. A convenient error, since we have no tools to measure flexure in the various hardware components. I have been blindly replacing components for awhile now.

Things I have done:

– Regreased the gears.

– Rerun PEC (after regreasing). The screenshot below shows the result of the new PEC measurement (Red line) versus the original PEC curve (green) from a year ago. THE PEC is essentially identical. PEC is obtained from 15 worm cycles using PemPro.



After installing the PEC curve, the PEC error is essentially zero as seen above. This also tells me I have not gotten the PEC curve installed inverted.

– Replaced the flimsy stock aluminum dovetail for the C11 with a Homeyer cradle. This makes the C11 incredibly solid on the Versa plate.

– Replaced the Sky90 rings with a Takahashi clam shell.

– Removed Telrad.

– Replaced Sky90 compression fitting holding the camera with a completely threaded optical train.

– Double checked balancing. The scope is as close to evenly balanced as I can make it, both in RA and Dec.

– Worked on polar alignment. I ran a number of 50 point TPoint models to obtain no error (Excellent Alignment, No Further Adjustment) in the Polar Alignment results screen. However, the 350 point model indicates I am still off by 2.1 arcminutes in Azimuth and 2.9 arcminutes in Altitude.

– Replaced previous external guide camera with STF8300 and  STi Off axis guider. This doesn’t affect the Sky90 system directly, but perhaps could have contributed some sort of flexure.

– Ran some cables through the mount to reduce the number of wires in the pigtail. Rerouted the wires to correctly avoid “dragging” as the mount turns. The dragging had shown up in TPoint models. The current model does not show that pattern.

– Double/Triple checked tightness of various bolts.

– Tied down wiring on mount using Velcro straps. I assume wiring position changes would not lead to this type of drift.


Regreasing MX Adventure

It has come to my attention on various Bisque and ACP posts that I should grease the gears on the MX yearly. Since I have never done this, I figure it’s about time.

Of course, I immediately screwed up. First, I removed the knob on the Dec worm gear box. Wrong, the knob is supposed to come off with the cover. Second, I left the mount in Gear Engaged mode, thinking I would need to run the gears to access the various segments of the gear. Wrong, one is supposed to work with it in Balance mode. Third, I saw the gear for the belt drive and thought that was the target gear, so I greased that. Nope, wrong gear.

After emails with Bisque, I was led to a nice video describing how to disassemble and reassemble the worm block.

I was able to correctly reassemble the knob on the shaft.

I then tediously managed to clean the grease off the belt and gear.

Subsequent removal of the worm block (with motor intact) was successful. Cleaned the worm with a rag, then with toothbrush. Used a rag to clean the main gear, got everything reassembled. There could be an issue that I did not get the lower worm block screws tight enough; I don’t have T Allen wrenches.

Finally, did the cam adjustment. Seemed easy, hopefully it is correct.

Repeated the procedure on the Ra gear. Ran the TCS mount exerciser 6 times, everything seems to be working.