These webpages were born when I still owned my G11, until the year 2002 or so. More recent models have seen significant improvements, thus a lot of the modifications and tuning described here are now superseeded or superfluous, or maybe even plainly wrong. They only apply to models produced before the year 2000 or so. I moved on to a larger mount a long time ago since my own needs changed, and therefore I cannot reflect the G11's current state in these web pages any more. I will keep these pages alive as they are now, for reference, and because I expect these old sturdy mounts to be around for a significant number of years yet to come. Should any pictures or text presented here about the G11 be of use for your own webpage, then please feel free to use it.
Thank you,
Mischa Schirmer, June 11, 2009
... it requires some care in order to perform well. Here is a list of things I liked and disliked.
Yet there are still as many negative points as positive ones. Some of them have been eliminated in more recent versions of the G11, and are marked as such. Many of these issues, however, can be permanently fixed by yourself.
See these instructions to learn how to remove the axes from their housings. You might want to use gasoline or similar liquids to remove all of the old grease from the bearings and from the axes. This can take quite a while for old mounts, since the grease can be very sticky. This could be due to some anti-corrosion stuff that was originally put into the bearings before reassembly. When greasing them later on, a chemical reaction ages the grease within two weeks by a decade.
You succeeded when you can easily rotate each of the needles in the bearings, and when you can rotate the bearing itself: stick the nail of your thumb in between the needles and try to push them into one direction. Once everything is moveable, clean it again. You are done when the bearing doesn't leave any dirty traces on some white tissue any more. Do this for both the two bearings of the RA and the DEC axis each. If you now put back the axis into its bearing before you put the new grease, and give it a strong push with your hand, then it should rotate freely up to five or ten times before coming to a rest.
What kind of grease can you use? Usually some titanium or lithium grease work fine, you get it in your local bicycle store. Any other grease of this sort will do as well. You might want to put it into the freezer for an hour before greasing your mount, in order to see whether it is still smooth at very low temperatures. You do not need much grease, the amount you can see in Fig. 1 below is just fine.
You should notice a dramatic difference in the performance of the mount when reassembled. It is much more sensitive to balance with open clutch knobs, which allows for better tracking. Furthermore, you don't need to move the mount with the slow 16x speed of the hand controller any more. With opened clutch knobs you can now accurately point the scope with a slight pressure of your finger since there is no more initial resistance. You will not lose the target in this process any more (the known "overshooting" problem). For visual observing you might even forget the need to tighten the axes.
 Fig. 1: The upper needle bearing |
 Fig. 2: The lower needle bearing |
Over time grease and oil moves to parts of the mount where there shouldn't be any (e.g. the couplings). This is a normal and unavoidable process, and is easily repaired. Just make sure that these surfaces are dry and free of any grease:
When pulling out the axes from their housings you will find a big white plastic ring on top of it (see Fig. 3). This is the coupling, which is made of some special material (nylon, I think). You can pull it off the axis. Do not use alcohol to clean it, but just plain water or some mild detergent for dish washing. They need to be absolutely free of any grease. The same holds for the surfaces above and below the coupling. As you can see in Fig. 1 there is some darker area outside the golden ring, having the same diameter as the white coupling. It is a fine layer of grease that slowly made its way onto the top surface of the worm gear. Thus the friction between the coupling and the metal surface is eliminated, and the clutch knobs will not work properly any more.
Losmandy offers couplings made of a different material, which guarantee even smoother operation than the nylon ones. Of course they must be free of any grease, too. They are called UHMW pads, but it seems Losmandy renamed them to clutch pads, code "CP11" at www.losmandy.com/pricelist.html. According to people on the Losmandy users group they are worth their money.
Mike Siniscalchi found a different explanation for the lacking friction when tightening the clutch knobs. See his website for more details.
The other surface that needs to be free of any grease and oil is seen in Fig. 2. It is the inner black ring immediately outside of the shiny needle bearing. Probably you will find some oil from the roller bearings and from the set of big washers (Fig. 3) there. Clean it, as well as the surface of the washer that comes next. Only use a tiny drop of oil for the roller bearings themselves when reassembling the mount.
Being in the process of greasing, you might also want to care for the worms and gears.
 Fig. 3: The white plastic coupling on the RA axis. |
This is probably the part of the G11 where you have to fiddle longest to find your personal optimum. Though this section might well be neglected by visual observes, photographers will find it important. The problem is that the worm is held in place only by two screws. Losen these screws and you can adjust the pressure of the two worm bearing blocks against the worm, the orientation and the distance of the worm with respect to the gear, and a couple of other things. It is not easy to find the sweet spot in this multi-dimensional parameter space that guarantees best tracking performance.
You want to achieve that you can rotate the worm with your thumb and index finger, grabbing it at the motor coupling. It must turn freely in both directions without getting stuck for at least one revolution. One also wants to minimize the backlash when changing the direction of rotation.
Let's start from a losened worm. Press the two block bearings with your right hand as strong as you can against each other, so that the worm can not move along its long axis. If it still does, you might want to put a thin washer between the motor coupling and the bearing block (see Fig. 5), so that there is no space left. At the same time, press the worm slightly against the gear with your right hand, and try to turn it with your left hand. If you press it to hard against the gair, the worm will get stuck. If the pressure isn't high enough, you will encounter noticeable backlash. Find the optimum pressure. You might want to press one of the bearing blocks slightly stronger against the gear than the other. If you think you found the optimum, tighten the two screws that hold the bearing blocks just so much that they keep the blocks safely in place. Check whether you can still freely turn the worm in both directions. Slightly open one of the screws if you want to change pressure of one of the bearing blocks. I'm talking about tenths of a millimeter here. When done, you can slowly but strongly tighten the two screws.
If after this exercise your worm still gets stuck, then you might want to adjust its height with respect to the gear. I did this by putting a double layer of paper beneath each of the two bearing blocks (Fig. 6).
Now attach the motor to the mount again, and check whether it can rotate the worm for a full revolution in both directions at 16x and 8x speed. You might want to repeat the whole exercise with the telescope attached to the mount. The additional forces may change the setup a bit, and you want to go for a slightly different adjustment: a worm that got stuck without telescope might be turned easily with the telescope in place, and vice versa. The effect of balancing can be dramatic in this respect, too. Note that every worm/gear combination needs to have a small amount of backlash! If there is no backlash at all, then the worm gets stuck and, in the worst case, damaged (as well as the gear). For photographic purposes, you want to go for a slightly larger backlash in RA. This eliminates sudden jumps of the guide star due to a binding of the worm. Since you guide at 0.5x or even 0.3x speed, you will not note any backlash since the worm always turns into the same direction. Backlesh in DEC can be compensated with the TVC setting.
 Fig. 4: The worm in its place. One can also see the two cubic worm bearing blocks and the big gear. |
 Fig. 5: An additional washer minimizes the axial movement of the worm. |
 Fig. 6: Adjusting the height of the worm with respect to the gear ('shimming') with a piece of paper. |
Periodic error correction (PEC) is often advertised as a cure for all problems. Well, it is not. The improvement gained in tracking strongly depends on the smoothness of the worm, and especially on the accuracy of the training.
What can you expect? Of course you can only correct the periodic components of the tracking error. Since you can not distinguish them from the non-periodic errors, your PEC will always contain some errors, in particular if you train it with your eye and not an autoguider. The more you can minimize these effects, the better your tracking will be. Make sure that you thoroughly cleaned and greased your worms and gears. A tiny bit of dust can move your guide star within seconds 10 arcseconds off the centre position. I think on average and with a reasonable good worm (+/- 15 arcsec periodic error) you can cut down the periodic error by a factor of two. With a worm with very smooth periodic error, and an excellent PEC training (such as with an autoguider, for instance), you can gain a factor of three or maybe even four in PE reduction. You will never get it to zero, and you will always have to guide your exposures. A properly balanced telescope, with slightly more weight on the eastern side, and a well ajusted worm is critical for good guiding with the G11.
In the following you can see the periodic error I measured with a C11 and a ST-8E. The focal length was about 2900mm in the imaging configuration, the pixel scale 0.64 arcsec per pixel. The G11 was equipped with the standard 'Celestron Digital Drive System'.
All fields lied on the celestial equator with DECL = 0. Thus the periodic error shown below is maximal. For larger declination values it decreases with a factor of cosine(decl). In order to get the tracks shown below, the mount was misaligned by several degrees in azimuth, making all stars in the field drifting slowly in declination (to the right in the images below). The periodic error in right ascension then shows up as a vertical oscillation.
The PEC was trained by the autoguider of the ST-8E when the mount was polar aligned. I then misaligned the mount, recorded the tracks with PEC on and then with PEC off. Each image was exposed 8 minutes, thus showing two revolutions of the worm. During the training of the PEC, corrections to the mount were sent every three seconds (exposure time of the guide star), thus averaging out seeing effects. The telescope was well balanced in RA and DEC, with a slight imbalance towards the east.
The periodic error of a mount is defined as the maximum deviation of the star from its nominal position. This value is indicated below in brackets, e.g. (±8.1). It was determined by measuring the largest peak-to-valley value within one revolution of the worm (4 minutes), and then dividing this value by 2.
I did this exercise in two nights, with the telescope teared down and set up again. In the first night, Fig. 7, it can be seen that one can mess up the tracking with PEC. In the second night (Fig. 8) I took a bit more care.
 Fig. 7: In this image the PEC was trained by the ST-8E with an aggressiveness = 10. This means that the correction the autoguider sends to the mount is exactly as determined by the calibration. Obviously this leads to a permanent over- and undercorrection, resulting in a even worse tracking performance. Note that in the untrained stellar tracks there is a period of about 1.5 minutes with an excellent tracking of about ±3 arcsec. It gets destroyed by the PEC. The stellar track with the enabled PEC is shorter because the untrained image was taken first, then the mount was polar aligned and the PEC trained, then I accidentally misaligned the mount by a smaller amount. However, this doesn't change the result. |
 Fig. 8: The same experiment again, but this time the PEC was trained with an aggressiveness = 7, meaning that the autoguider corrections sent to the mount were reduced by 30%. The improvement is significant. Wouldn't there be the outlier (marked with an arrow) with ±5 arcsec the result would be even better. This feature can also be seen at the beginning of the track, marked with the first yellow stick in the lower image. Why it is considerably smaller there I can't explain, probably some non-periodic feature jumped in. Without this outlier the remaining periodic error with PEC switched on would be ±2 arcsec and less during the first 4 minutes. In the remaining 4 minutes it is getting slighly worse, about ±3.5 arcsec. Note that the error looks different to the one I recorded in Fig. 7. |
Some mounts seem to run a bit too fast, as is illustrated in Fig. 9:
Fig. 9: An early record of the G11's periodic error
I found that the guide star was slowly drifting away from the center, something I noticed already when manually guiding some exposures. The amount of the drift is 1.5" per time minute, which means that the mount is 0.16% too fast, or that it would be 35 arcmin ahead after 24 hours. This is a very noticeable effect, and I have no clue where it comes from. I exchanged worms, drives, and even tried two other control units from other G11 mounts. Whatever I did, the effect remained. I tried to switch on/off the King rate or the solar speed, which didn't solve the problem. I also tried various batteries and wall transformers to rule out weird effects due to different voltages. To me it seems that the standard tracking speed that comes from the control unit is slightly off. An autoguider will cope with this problem automatically, hence it is not of a big concern in these days.
Losmandy offers a precision worm for the G11. They have a very smooth periodic error of about 5". In March 2003 I ordered 10 of these precision worm gears for a handful of german and austrian amateurs, who had tracking problems with their G11 mounts, too.
 Fig. 10a: Periodic error for the precision worm gear of Uli Schüly. We measured it as +/- 5.0 arcsec. As you can see the tracking is very smooth. Uli imaged with an ST-7 and a 10" SC at f/5. The bright star is Eta Aql at the celestial equator. Two-and-a-half revolutions of the worm can be seen (in the original image it is more). There are basically no non-periodic features evident. |
Mike Siniscalchi has polished his worms and the gears and hence improved their surface smoothness significantly, resulting in much better tracking performance.
What do we learn from this? I think that this G11 is just an average off-the-shelf product as most customers have. There are ones with much worse, but also much better worms. I have seen a G11 with an excellent tracking behaviour, less than 4 arcsec without PEC and a with a nice sinusoidal behaviour. If you are not satisfied with the tracking of your worm, then first exchange it with the DEC worm and see whether the latter one is better. If this doesn't help, then you might order one of Scott's precision worm gears. In most cases this should fix the tracking problems. If not, then the cheap Hurst stepper drives are the likely culprit. Exchange the RA drive with the DEC drive for a test, and see Rene Görlich's web page for further details and much more about tuning a G11. Some people say that for them the tracking problem emerged from the small worm bearings, which they got replaced.
A well balanced telescope (slightly more weight on the eastern side) is essential for good tracking. But depending on your imaging train, asymmetries in the mass distribution of the telescope and the position of your target on the sky, the balance of your telescope changes all the time. As does the stress acting on the mount. That's probably why the periodic error won't look exactly the same from night to night, especially when you are close to the G11's upper load limit.
One might argue about the use of PEC. As shown in Fig. 7 even an autoguider can clearly mess up the training of the PEC and therefore the tracking performance. Yet an autoguider definitely trains a PEC better than the human eye. In case the periodic error is really huge a manual training of the PEC makes sense (for conventional imaging with film, not for CCD). When the telescope is guided by an autoguider IMHO you can forget about the PEC. It is useless since the autoguider corrects deviations anyway. Besides, one would lose too much time verifying that the PEC was well trained.
Whether the G11 satisfies your needs for CCD photography depends of the focal length of your telescope and also of your demands. For me it quickly turned out that the G11 was insufficient. It is a great mount for visual observing, and also fine for photography with conventional 35mm film. It also works nice for CCD and focal lengths of, say, less than 500 mm.
For larger focal lengths the G11 is imho no longer satisfactory. The optional GEMINI GOTO functionality overcomes some of these issues, yet it does not eliminate some of the mechanical flaws (in particular the worms). For the price you pay for a G11 including the GEMINI system you can find mounts with higher payload capacity, and an already integrated GOTO system.
Though the G11 is a very robust mount, parts of it can be damaged if you don't be aware of the following issues. However, none of these problems will make your G11 die. You just have to buy a small replacement part for a fix.
 Fig. 11: The mount head hits the RA drive. |
 Fig. 12: The DEC encoder gears |
 Fig. 13: The RA encoder |
