Return to home Index
Return to LX200 Astro Programs

This material provided by Frank Loch. I thank him for allowing me to publish this fine article.

-- or The Best Change I Ever Made To The 12" LX200
To Improve Its Performance!

Also My Experiences with TPoint Software

Early on after the purchase and initial use of the 12" LX200 Classic, I began to realize that there was substantial friction in the DEC axis bearings or a problem with axis alignment. Several times I measured that about 3-3.5 lbs. of force applied to the corrector end of the OTA at a right angle was required to initiate motion even though the OTA was balanced (?). I question the "balance" here because the friction was so high that it was very difficult to determine that the OTA was really balanced. Also, initially, even with the DEC backlash set at the max (99), there was still very significant backlash . I never tried to measure it, but it was a lot -- sometimes 5-8 seconds would elapse before motion was detected in DEC axis.

As a result of the high friction, I was pretty well convinced that the GoTo performance was severely compromised. I struggled with this lack of pointing accuracy for several years. At one time I loosened one of the forks in an attempt to find a better position for it to improve the alignment and thereby reduce the DEC axis friction. This action was not very helpful.

In November of 1998 I purchased TPoint from Software Bisque. I was hoping that TPoint would improve the pointing accuracy (i.e. make a silk purse out of a sow's ear!) without having to make any other changes. No such luck!

TPoint is a fine program designed to improve pointing. After working with TPoint for several weeks, the data TPoint produced very clearly showed me that the most significant pointing errors were DEC related and a very high scattering of data points could clearly be seen in the DEC axis direction compared with the RA axis scattering. Studying the data as it was accumulating, it was obvious that "inconsistent" and "unrepeatable " errors were killing the pointing accuracy in the DEC direction. TPoint needs "repeatable" and rather consistent errors to perform its magic correctly (and magic it has, believe me). I estimated that pointing accuracy even with TPoint was un-predictably erratic and at least +/- 15 arcminutes. Thus, the high DEC "Stiction" was killing both the LX200 pointing accuracy and "GOTO" performance.


At that point I made the decision to perform Michael Hart's DEC bearing replacement as nicely described on Doc G's website Decfixmajor. I was initially very "scared" of undertaking this procedure (as it turns out a needless fear). Even though I consider that I have above average mechanical skills and a Professional Mechanical Engineering background, the very thought of tearing apart a $5000 scope and getting it back together again with satisfactory alignment was very daunting.

I followed Michael's and Doc G's instructions very closely. I will only add now a bit here and there to report my own additions, findings, or possible suggestions to an already fine procedure (many thanks Michael & Doc).

First, I obtained the the two needle bearings (as Torrington BH1616 - $5.49 each) originally specified SCH1616INA -1" shaft ID x 1.313 OD x 1" long from the Brown Bearing Co, (a well known bearing supply house here in S.E. PA). I also obtained 6 (not knowing exactly how many I would need - actually I only used one) Torrington TRA1625 Thrust washers ($0.64 each), and a 3/4 oz supply (smallest available --$12.00) of Locktite #242 (I actually only used a few drops).

Secondly, I prepared a sturdy table in our family room to receive the scope and then with the assistance of my son Kermit, loosened (at the pier) the front mounting bolt (scope to super wedge) and removed the two rear (lower) mounting bolts, and finally sliding the scope off the super wedge, we carried it into the family room and placed it on the previously prepared table in the Alt/AZ orientation.

Next, using an Exacto knife, I scored the fork to base locations on both forks as in Michael Heart's Figure 4. Then I removed the DEC drive cover and released the DEC drive electrical cable from its lower fork socket. Finally I removed the DEC drive motor and control assembly, checked it carefully for any unusual bearing slop (found none) and placed it in a plastic baggie to keep it clean.

I marked the worm gear for future re-orientation and then removed it. Using an old tooth brush and a little alcohol, I cleaned off all the old "black moly" grease and any dirt particles from the worm gear and placed it in a plastic baggie to keep it clean.

Now the "fun" part (read really scary part here, because if this fails, there is no way to proceed further) begins. This is the action of removing the two allen screws from the inner clutch plate. Check out your Propane torch first and make sure you can get a "hot" 2" flame with a 1" super blue "very hot" inner cone showing. If you can't get a really hot flame, get another torch that can produce a really "hot" sky blue inner cone. Get the proper size allen wrench ready (a long one with at least a 3 1/2 " torque arm available).

Heat and remove the allen screws one at a time. Place the point of the hot blue torch cone right at the head of the socket screw (maybe a 1/8 to 1/4 inch space from the tip of the hot blue torch cone to the screw head). Heat for 20 seconds (Michael Hart says 5-10 seconds). Try to remove the screw. If it feels too tight yet (this is a judgment call - you don't want to strip the socket in the screw head), apply heat for 30 more seconds and try to remove the screw again. Thirty seconds did not work for me -- neither did 40 seconds -- I'm getting very worried now! Try 60 seconds! That worked for me for both screws -- whew! Sit down and pour a sherry! The rest is much easier and less stressful! Now remove the Clutch plate (after it cools a bit so you can touch it).

Gently lay the scope over on its side so the bulk of the OTA is laying down on the table. Remove the four long allen screws from the left fork and remove the fork carefully from the OTA. Remove its old plastic sleeve bearing and set it aside. Now, very carefully "jiggle" the OTA off the other (Right) fork and base assembly. Remove the plastic bearing from the right fork. Now using Duct tape or masking tape, tape up all the cracks and openings in the LX200 computer base assembly to keep it clean and "chip free" during the "boring out" procedure.

Researching in the phone book, I located a machine shop only a few miles from my location and after a short phone call I was pretty sure that they could do the "boring" out successfully. I took the separate fork and taped up fork assembly to them and within a day the job was finished. They even "pressed" the new bearings in place for me saving me from having to do that. Total charge $60.00.

I re-assembled the scope following Michael Hart's instructions. This was pretty straight forward and not very difficult. I lubricated the new needle bearings with red Bosch bearing grease. I used a few drops of "locktite" to lock the DEC clutch plate Allen screws again. No "primer" was used.


WoW! Zero DEC bearing friction! I even think that's an under statement! Better hold on to the OTA as now it really "books" if you let it go "unbalanced". A virtually "free floating" OTA. I had no idea it could be like this! Can't wait to power up and test! But, must finish this assembly and alignment first.

While performing the OTA alignment, the longest focal length eye piece I had available was 26 mm, so the resulting light cone spot at the RA bearing retaining screw (see MH's procedure) was about as big as a nickel. So I had to "eyeball" the centering of this spot around the screw head (not too precise, but OK). At an auto parts supply store, I found the "black moly" grease for the DEC worm gear. Re-installing the DEC drive motor assembly was un-eventful. I used a few drops of locktite on the DEC setting circle spacer to lock the setting circle in place.

OK! Time to power up and test. First, static balance was set with the Meade sliding weights. Now apply the power and boot up! With the DEC drive cover off it's time to check out the DEC drive visually to see what's going on. Hmm! I can hardly hear the DEC drive motor running now when I push the keypad N/S buttons. Nice! No more DEC grinding sound! Lets look at the the DEC worm action. Before the bearing modification, the DEC worm would flip out of position and the spring supported plate would dip down about 1/8th inch or so at the onset of each new command. Hmm! Difficult to see it moving at all now. Substantial improvement here. Never thought I would see the worm so still! (later when mounted on the pier, I was able to set the backlash at 75 and have no (yes zero) backlash.

Did all this improve the pointing accuracy?? You better believe it. Now with Tpoint its better than +/- 2 arcminutes, and it puts the target inside the ST7 frame at f/10 100% of the time (54 successes out of 54 Slews). I can't believe that this modification could make that much difference, but it shurly does! Caution -- Further extended testing is in progress and the 100% factor may eventually drop a percentage point or so with this. My latest mapping run on 01/05/99 had another 100% success rate for 47 mapped points. However, I feel current results may be too optimistic until I have additional data to support it.


First I want to thank Michael Hart for his very complete set of instructions for performing this modification. Secondly I want to thank Dick Greiner (Doc G) for making Michael's and others instructions available from his renowned website. Before I started this project, I printed out all of the pertinent articles from Doc G's site and referred to them frequently while working on my scope.

Last, reflecting back on this modification, I can see no reason for my earlier fears. This is a pretty straightforward procedure. I do believe that anyone with a reasonable amount of mechanical skills can perform this modification and be successful with the results. If you have a 12" LX 200 (or other similar scope) with a high friction set of DEC bearings, please consider this modification. You will be very pleased with the results.



TPoint is an excellent telescope pointing program is the work of Patrick Wallace who many consider to be the worlds leading expert on telescope pointing. It is my understanding that most of the worlds major telescopes like the giants up on Mauna Kea and elsewhere use TPoint and the services of Patrick Wallace to set them up and map in their pointing software. As I write this Patrick is now pointing - in the new Gemini 1 (8.3 meter) telescope on top of Mauna Kea (14,000 ft).

Software Bisque has licensed TPoint from PW and adapted it to work in conjunction with TheSky Level 4. That is the version that I am using and writing about here. At this point I want to extend a very large "Thank You" to Patrick Wallace who has been extremely helpful to me personally, as I have been learning to use TPoint with my LX200 12" telescope. When I began to learn about and use TPoint in early November of 1998, I was struggling and had many questions which generated my cries for help on MAPUG-Astronomy (Meade Advanced Products Users Group). Patrick jumped to my rescue and has been my TPoint mentor ever since, sometimes on a daily basis.

I want to make a very strong statement right off here. TPOINT WORKS!

Two factors may cause problems using it! First, your LX200 (particularly if it is the 12" model) may have a high friction DEC bearing problem (and/or sloppy DEC drive) as mine did originally. Secondly, the implementation by Software Bisque, in particular some parts of the instruction manual, can be confusing and cause much frustration.

As I progress here, I will outline what I hope is an easy to understand (if I do this correctly) procedure for using TPoint with your LX200 from night to night.

First, however, you must solve (if you have it) the high friction DEC problem with the bearing modification described above. I can't say this too strongly. If you have a high friction DEC set of bearings, and/or a sloppy DEC drive system with a lot of uncontrollable backlash , then TPoint will not help much!! TPoint requires repeatable errors (not random errors) in order to function efficiently and to significantly improve your pointing accuracy!

Look above and see my LX200 set up. Attached to the pier mounted scope is a Van Slyke Versaport II Flip Mirror Box, then the SBIG AO7 adaptive optics device and finally the ST7 CCD camera. Not shown, but always used is a Kendrick Dew Shield. With the Dew Shield attached, the Meade balance weights are positioned to get as good a balance as possible. Most importantly, I was (and still am) operating at f/10.

TPoint requires that you "Map- in" a significant number of stars (say 25 to 50 is the norm here) and this "mapping" requires slewing to the "targets" and then centering the targets on the CCD imaging chip before pressing the "MAP" button. I do this using the CCD frame on my computer monitoring screen, upon which I have drawn with a fine point magic marker a small "+" at the center of the frame on the monitor. I then use the W,A,S, & Z keys as for the normal method of scope movement using CCDOPS (DOS version 3.78) to center the target star. TheSky level 4 must be used here. TheSky level 4 is a $50 upgrade if you have one of the lower levels, but you will also obtain and then use the new improved 2CD level 4 version 5.0 ( I like this version a lot).

Some of you may want to do the "Dummy Run" described below first. I point this out now so that you do not overlook it. Link here


When I first started using TPoint, the pointing results were very variable and confusing. The results were showing a success rate ( target slewed to ending up inside the CCD frame) of 54% to 68 %, while statistically TPoint was showing an RMS of 47 to 48 and PSD of 48 to 54. These values of RMS and PSD are considered pretty good on an LX 200, but this turned out to be a red herring (see the later remarks on not using SYNC).

While "mapping" I began to notice (visually, in the FMB, trying to find and "rough center" my targets that didn't make it into the CCD frame), that the errors were sometimes rather consistently and always off about the same amount to the north for a while, say 3 to 6 or so slews and then again rather consistently off to the south for another batch of slews, while maintaining very good and consistent East-West locations (very tight grouping E-W).

My basic intuition was telling me then, that perhaps my high friction and sloppy DEC drive was causing a lot of my TPoint problems. My first corrective action was to replace my Gel Cell battery power supply (with the 12/18 Mead converter which I always had suspected of being a "soft" power supply compared with the Mead 120/18 volt "stiff" power supply) with the Meade 12/18 volt supply.. Upon doing this, the DEC motor drive sounded much "peppier" and less "wimpy". Simultaneously the TPoint scatter diagram changed shape dramatically, from a larger spread in RA to a larger spread in DEC. TPoint was reporting RMS values like 119 to 130 and the PSD to 131 to 139. On my best slew runs the success rate was at times as high as 92% to 98%. I felt that this was very encouraging. But repeat runs sometimes produced success rates in only the 50% bracket.

Further experimentation convinced me that I should implement the DEC bearing modification above, if there was ever to be any hope of having a system on which TPoint could show what it can do.


Described below are procedures and techniques that work for me, with my particular LX200 12" f/10 set up. Your situation may be different and require some changes, but if you are looking for a way to start, then this should be helpful.

TheSky Documents -- Understanding "Documents" in TheSky software was interesting and became very useful for me. When you open TheSky; the ""; is the title that usually opens as the default document. To open a new "document" you must "save as" the "" under a different name say When you do this, you then are able to set up TPoint with a "clean" empty data sheet to start a mapping run. If you need in the future a "clean" empty TPoint data sheet, for a new set of circumstances that might require another new mapping run, then you can use this method to set up other documents like, etc.

Setting Up Tpoint -- Once you have obtained TPoint from Software Bisque, follow their instructions for installing it on page 8 of their manual, and after installation, launch it once to initialize it.

Now, open up TheSky, and either set up a new sky document or if you have previously done this, then open up your new document.

In the Edit menu, click on "Insert New Object", then select "TPoint Model" from the list, and check the "Display as Icon" box and "OK". TPoint is now working in your new "SKY" document and will be seen as the gray Icon on the upper left of your "sky" window.

Now link your LX 200 to the sky. Double click on the TPoint Icon to open up Tpoint, select the "Model" menu and Observing Parameters. You should find your LX 200 listed there as well as your local site data. Set the "pressure" & "height" to "0". (This is important!). Click " OK" and in the "File" menu, select "Exit and return to the sky". Now right click on the TPoint icon, and then select "show as Icon" which already has a check on it to change to the "scatter diagram view" which will show no data until after you map the first 6 stars.

You are now ready to start mapping in the stars as data points. Select a fairly bright star and slew to it. Use whatever means you normally use to "center" that star in your CCD target frame on your monitor. Switching back to your monitors view of "TheSky", click on the "Map" button in the "Object information box".

CAUTION -- NEVER CLICK ON SYNC WHEN YOU ARE USING TPOINT, as "SYNC" screws up the works immediately! I find that if I map the first 6 stars or so as stars that are rather close to each other, that this helps TPoint to start controlling your scope, and almost immediately this will improve your pointing accuracy. Then proceed to map in a total of 25 to 50 stars, spread around the particular portion of the sky that you normally use. Your pointing accuracy should be very good now and you should be getting about 100% of your targets in your CCD frame even at f/10 if that is where you work.

Your TPoint SKY RMS should be somewhere between 70 and 90 and the PSD in the same range.

If you wish, continue now to slew to your desired targets and do some imaging! TPoint should be working for you just fine. In my case, at the end of the evenings imaging, I must un-link, turn off the LX200 (after first returning to "0" RA and "0" DEC, and turn off my computer. That means that to start again the next imaging evening, I must do a "Short Mapping Run"


During any observing and imaging session, I keep a rather detailed log of what I am doing and what it happening. I found it very helpful to do the same with my TPoint mapping runs.

My note taking during a mapping run (or subsequent observing/imaging run) with TPoint consists of the following as a minimum:

  1. I list each star as mapped.
  2. Aside of the star's name, if after I slew to the star, it is in the CCD frame I draw a little rectangle simulating the CCD frame and very approximately place a "dot" where the star fell into the frame before centering.
  3. If the star did not fall into the CCD target frame, instead of the rectangle, I draw an extended "plus" sign (+), and put a dot in the plus sign approximately at the location of where this star fell in my FMB reticled eyepiece.
  4. As I progress through this session, I use the above information to "keep score" of the successes, so at the end I can compute the % of successes. I can also spot "trends" that provide clues to possible problems.


TPoint will provide you with much information about your Telescope setup. One of these that I have been making use of, is the term MA or the Polar Axis Azimuth Error. This is reported in arc seconds, which, with a little Trig you can use to compute how much you need to crank into your "Wedge" to make the required correction. After making the correction, running another TPoint mapping run will verify your results.

Return To Beginning



  1. Open your previous Sky document that contains the previous TPoint model that you wish to use again.
  2. You must then do a "Short Mapping Run" of at least 6 new star data points. You may refer to pages 24 & 25 in the TPoint manual, which are two of the most confusing pages in the book, while you follow the procedure below. I will try to explain why you do what you do here.
  3. Now link your LX 200 to TheSky. Double click on the TPoint Icon to open up TPoint, select the "Model" menu and Observing Parameters. You should find your LX 200 listed there as well as your local site data. Set the "pressure" and "height" to "0". Click " OK" and in the "File" menu, select "Exit and return to the sky".
  4. Double click on the TPoint Icon to open TPoint.In the Model menu select "Short Mapping Run". Click on "Fix and Mask" and OK.
  5. This fixes NP, ME, & MA, leaving IH, ID, & CH un fixed and empty, ready to receive the data from the new 6 points you are going to map. IH, ID, and CH will then develop values as you map the 6 new points, while the coefficients of NP, ME, & MA will remain fixed. These will be helping TPoint point your scope. Close TPoint by "Exit and Return to the Sky" from the file menu.
  6. Now right click on the TPoint icon, and then select "Show as Icon" which already has a check on it to change to the "scatter diagram view". Now, as you did previously, map in 6 or so new stars.
  7. You now have six or more new points at the bottom of the spread sheet.
  8. Double click on the TPoint icon so that the full spread sheet occupies the screen.
  9. Click on Model / Fit Data or just the FIT box. This brings up a Fit Data Box and at, the top left, 6 Terms, all ticked. NP, ME, & MA will appear "fixed" (brackets around the coefficients), while IH, ID, & CH appear as "not fixed" (no brackets around the coefficients).
  10. Now click on the More Terms box and this brings up a Terms box. Select the terms you want to un-fix (NP, ME, and MA) and remove the check mark from the box "fixed term" for NP, ME, and MA.
  11. You will see as you select the various terms, those that are "unfixed" to start with will not have a check mark in the boxes, while those that are "fixed" to start with, will have the check mark in the boxes.
  12. In other words you must leave the Use box ticked for these three terms (NP, ME, and MA) and
    un-tick the Fixed Term box.
  13. You must now leave only the Use box ticked for the other three terms IH, ID and CH. Got it?
  14. The "use box" will now be ticked for all 6 terms. None will be "fixed" at this point. You will be using these 6 terms for the rest of the evening with whatever coefficients they have developed. IH, ID,
    & CH will have the new coefficients from the 6 points mapped, and NP, ME, and MA will have the coefficients from your old previous mapping run.
  15. This assumes you are just using six terms in all which is my normal situation.
  16. Now close the Terms box. You then OK the Fit Data box. You collapse the TPoint spread sheet to an icon ("Exit and return to TheSky") and you then just carry on slewing to whatever you are interested in for the rest of that evening.

At this point, I usually pour a "sherry", relax and slew to my hearts content, hopefully hitting all my targets nearly center on. If you have reached this point successfully, you are well on your way to mastering TPoint and enjoying its use.


A "dummy run" provides an end-to-end test of the system, proving that TPoint can reconstruct a specified model from a set of fake star observations.

It is probably a good idea to do such a run early on, to practice and set up benchmarks for comparison if you later make changes to your scope and accessories.

The "Dummy" TPoint run can be done any time including in daylight if you want to, as actual stars are not required (keep your end cap on for this). A "dummy" run is performed by using a new .sky document with a freshly installed Tpoint that has a clean "empty" data sheet.

Before we start, we want to set all 6 terms IH, ID, CH, MP, ME, & MA to zero (or other known values) and "fix" them at those values.

The routine for doing a "dummy run" is as follows:

If you want to enter model coefficients without any pointing data at any time you can use this method.

  1. Start TheSky.
  2. Go to Edit, Insert New Object, and select a TPoint Model to get a model into TheSky.
  3. Add six "bogus" mapping points to the new model. (This is the work around to allow you to enter the Fit dialog in TPoint. The easiest way to do this is to connect to the Telescope Simulator, identify an object and press the Map button on the object identification dialog six times.)
  4. Double click the TPoint Icon in TheSky to fully open TPoint.
  5. Mask the six bogus points by clicking on each row and choosing Data, Mask (or click the Mask button on the toolbar).
  6. Choose Model, Fit Data, More Terms to allow you to fix and add coefficient values (you must fix all terms you use so their entered value is maintained). Here now enter "0" for all of the 6 terms (or what ever values you want to use) and fix all 6.
  7. Choose Close, OK, then File, Exit and Return to TheSky. TheSky will then be applying coefficients just entered.

In Software Bisque's implementation of TPoint, the mapping data and the model are contained in one document -- the TPoint Model. The model that is inserted into TheSky is the one that is being used (as soon as there are six or more mapped points, by default the six most common coefficients kick in, which one can alter as needed). You can shut off the model from TheSky by clearing the checkmark under Telescope, Setup, Enable TheSky Modeling.

  1. Now re-select the LX200 as your scope to link to and link your LX 200 to TheSky. Double click on the TPoint Icon to open up Tpoint, select the "Model" menu and Observing Parameters. You should find your LX200 listed there as well as your local site data. Set the "pressure" & "height" to "0". Click " OK" and in the "File" menu, select "Exit and return to the sky".
  2. Now right click on the TPoint icon, and then select "show as Icon" which already has a check on it to change to the "scatter diagram view" which will show no data until after you map the first 6 stars.
  3. Then, using "TheSky" slew your LX200 to 25 or so "Stars" as if the stars were "out" at night. Map each (without adjusting anything).
  4. Look at your scatter data, and other plots. The RMS & PSD should be around 30 (limited by the LX 200 encoder resolution) if all is working ok. Now un - fix the coefficients and fit the model to the data.
  5. The coefficient values will change slightly because of "noise" but the RMS shouldn't decrease much.
  6. That's it for the "Dummy" run. Print and save your results for later comparisons if you make changes to your setup.

At this point in time , now, after about 7 weeks experience working with TPoint, I have completed 23 mapping runs. I would be very pleased to answer any questions that readers might have. Please write!

Return to Beginning

Comments from Hugh McKerrell, Lochranza, Isle of Arran, Scotland.


Frank Loch's experience with TPoint parallels my own in that we both started out with it at the same time (November last) and we both conclude that it really does work. However, although we both use a 12 inch lx200 and an ST7 camera, our two systems could hardly be more different.

Frank uses his LX200 at f10 whereas I use mine with the Optec f3.3 focal reducer. My CCD area is thus 25' x 15' whereas Frank's is 8' x 5'. And although my slewing would seem to be far less critical in fact we both ended up with remarkably similar precisions of about 1.5 minutes for both the RMS and PSD of TPoint. Frank also switches off his power supply after use while I keep mine on with an un-interruptible power supply and a Sleeper plugged into the CCD socket. Thus Frank has to do the Short Mapping Run after starting TPoint whereas, in principle, I do not.

I ran 60 test slews under TPoint control and got every one well within the confines of my CCD chip. Indeed the PSD of 1.5 minutes is just about right for what I found. About two thirds of the results were within 1.5 minutes of the centre of the chip, a quarter were within 3.0 minutes of the centre and just one in twenty were further from the centre than that. Moreover, the sixtieth slew was spot on, within a minute of the centre. Compare the results without TPoint when, with just six slews and no ynchronising, the star was already off the chip. So, unequivocally, TPoint does work.

But, that said, I can only echo Frank's experience with the Software Bisque manual. This really needs clarification and re-writing from the section after the short mapping run. The only guidance you are given is that your newly mapped six points are to be submitted to TPoint.

Fortunately Patrick Wallace was able to advise and clarify. So if you are thinking of using TPoint then read Frank's instructions carefully for the exact method of linking your short mapping run to your basic TPoint model.

Go to Home Index for Doc G's Info Site