History
I have owned two Losmandy G11 mounts, the first without and the second with the Gemini controller. Over a period of several years I have measured the tracking errors of these mounts and attempted to reduce them by various modifications. These modifications have included replacing and swapping worm gears, replacing the worm bearings with better quality ones, filing and smoothing the blocks, and numerous attempts at adjusting the worm position. I have never been able to get the periodic error (PE) below about 12 arcsecs pk to pk, and at times it has been up to about 20 arcsecs pk to pk. I also discovered the existence of an aperiodic variation with a period of 76 seconds, that has since been confirmed by many others, and has been attributed to the worm gear not sitting squarely in its blocks. Early in 2007 I saw a reference to the Optique et Vision company run by Franck Valbousquet; see their website at Ovision. At that time he was selling an upgrade kit consisting of a new worm and a pair of new bearings. He promised a peak to peak PE of less than 10 arcseconds, but only if he fitted the new worm and adjusted it himself. As I was already quite experienced in this field, and my location in England was quite a long way from his company in the south of France, I persuaded him to sell me a kit and fitted it myself. I'm fairly sure that no-one else outside France had tried one of these worms at that time, so I accepted that I was taking a bit of a risk.Method of Testing
All tests have been conducted by mounting my Celestron C11 280mm SCT on the G11 and imaging with a Toucam Pro webcam through a Celestron 2X Barlow. This gives an image scale of 0.2063 arcseconds per pixel, which allows fine resolution of the errors, limited only by seeing. The scope was focussed on a bright star near the meridian and near the celestial equator. If the star was a significant distance in declination from the equator then the movement in RA was multiplied by 1/cos(delta) to correct it. Data was captured using K3CCDTools software, which includes a nice 'Drift Explorer' facility. Data was captured as a csv file, usually using an exposure of typically 1/50th second at intervals of 0.5 seconds. This high sampling rate is useful for capturing any high frequency components of the tracking error, again limited by seeing. Subsequent analysis was done using Microsoft Excel spreadsheets. The columns and curves labeled 'Corrected RA' were created by estimating the drift due to polar misalignment (I don't have a permanent observatory) and applying the appropriate correction factor.First Worm
The first worm (serial no exhq3r4) was tested in March 2007. The results are shown below:
The full spreadsheet is available here.
It can be seen that the PE is about 14 arcseconds pk to pk, which is no improvement on the standard Losmandy worms.
I contacted Franck, expressing my disappointment, asking for a replacement.
I didn't receive one, and eventually wrote the whole exercise (and quite a lot of money) off to experience.
Second Worm
Early in January 2008 Franck contacted me to say that he was developing a newer one piece worm assembly. As I had been advocating this approach for some years myself I was quite excited at the prospect. In July 2008 I received one of the new one piece assemblies, and returned the previous worm. As soon as the weather allowed, which was 16th July, I fitted the new assembly and tested it. The results are shown below:
The full spreadsheet is available here.
It can be seen that the PE is about 30 arcsecond pk to pk, much the worst performance that I have ever seen in a G11.
I again forwarded the data to Franck, again expressing my disappointment.
He promised to send another assembly.
We also spoke on the phone and he was obviously determined to solve the problem.
Third Worm
Another worm assembly arrived soon after, and I was still keen to try it. The opportunity arose on 14th August 2008. By now I had refitted the original steel Losmandy worm, so that the mount would at least be usable, so I decide to measure this for comparison. The results are shown below:
The full spreadsheet is available here.
The PE is nearly 20 arcseconds pk to pk, and might have benefitted from a readjustment, especially as I had left the backlash quite high on this occasion.
Then without moving or changing any of the setup, I exchanged the Losmandy worm for the Ovision assembly.
By now I am quite experienced at fitting and adjusting these assemblies, so this was quite easy to accomplish, even in the dark by torchlight!
I then immediately tested the new worm.
The results are shown below:
The full spreadsheet is available here.
It can be seen that the PE has dropped dramatically to about 6 arcseconds peak to peak.
At last the mount is performing as I have been hoping for some time!
A Note on Assessment
Testing of this nature is limited by the effects of seeing, especially when the total excursion is below 10 arcseconds. I believe that it's important nevertheless to use a high imaging scale, and to use a high sampling rate. On the last graph I have applied a digital low pass filter to smooth out some of the seeing. If you download the full spreadsheet you can experiment with adjusting the time constant, in Cell I1. A low value, of say 0.1 second, will reveal the raw almost unsmoothed data, whereas a high value, say 10 seconds, will remove most of the effects of seeing, but also any high frequency noise that may have been present in the tracking. The graph shown is with a time constant of 5 seconds. I also believe that is important to present the data in the time domain as shown. While data in the frequency domain is interesting, especially when tracking down components at other than the fundamental, it doesn't give such a good indication of how the mount might perform when autoguiding.Conclusion
The latest Ovision worm that I tested is now performing well. It easily meets its claimed spec of less than 10 arcseconds pk to pk, and appears to give quite smooth tracking, although this is difficult to judge at this scale and with the seeing normally experienced here. It's also easy to fit. If you are interested in high quality imaging, especially with long focal lengths, then this upgrade is worthwhile, and I can recommend it.
John Moore 20th August 2008