SkyViking
23-04-2011, 11:47 PM
I have lately been working on my first ATM project: An equatorially mounted Serrurier truss Newtonian OTA. I earlier received some great advice in this thread: http://www.iceinspace.com.au/forum/showthread.php?t=67713, and as a result I settled for the Serrurier truss design. Thanks to all for your previous inputs.
My major inspiration came from this beautifully constructed Newtonian: http://www.aao.gov.au/local/www/sl/sl-tels.html#ni-tel
For many more images of the project see the gallery at http://www.pbase.com/rolfolsen/10_inch_serrurier_truss_tube_newton ian
This project was intended to be a cheap upgrade I could work on while saving for a new CCD cam. To give the upcoming new cam justice I didn't want to mount it on my existing OTA which is a 17 year old simple sonotube, squeezed out of shape by the secondary spider supports and generally emitting a slight mouldy odour - probably due to the many years of dew and moisture absorbtion I guess. It was definitely time for something new.
While I intend to purchase a better mirror (a conical Royce mirror seems tempting) I decided to build a 10" to keep costs down, since my current telescope is of this size and so I could simply reuse the optics for the time being. Also, I was not sure how much load I could safely put on the G-11 mount in practice while still maintaining smooth tracking for photography purposes. While a 12" would be great I felt I'd rather have a stable tracking 10" scope than a 12", or larger, which may cause problems for the mount. Furthermore, the 10" is just about the right size for my observatory, any larger and I think it might start to become too bulky to maneuver around in the dark - otherwise I would have to build a larger observatory...
The design consists of three parts: The mirror cell, the central box/brace and the secondary cage.
The central brace was the first part I built, since this was the most critical component as it must be very strong and rigid in order to carry the telescope and eliminate flexure. I made this out of aluminium square tubing and some pieces of 40x40x3mm aluminium angle extrusions. I cut 12 pieces of square tubing at 45 degree anges and riveted them to the angles. The protuding angles would then later function as truss connectors. This structure alone was surprisingly rigid and ensured that everything stayed in place, but as this component would hold the entire weight of the OTA I asked a local metal smith to weld all the joints on it. The final resulting central brace was very light, beautifully simple, and yet extremely sturdy.
Next thing to make was the mirror cell and secondary cage. I decided to make these from plywood as it would be much easier for me than using aluminium, and it seemed to be the long accepted tried and tested solution of ATM's. For the secondary cage I cut two identical rings with an inner diameter of 30cm from the plywood. For the mirror cell I realised the ring could be smaller (saving a bit of weight), since being close to the mirror it would not interfere with the optical path, so I cut this ring with an inner diameter of 27cm to acommodate the 25cm mirror (10 inches) plus a bit of wriggle room for adjustments. The adjustable mirror cell was then made by cutting a trifoil shape of the plywood and mounting this on the ring with three spring loaded bolts.
The two rings for the secondary cage were connected with 4 round 1" alu tube pieces. I thought about how to connect the hollow tubes for a while before deciding to simply slip them over a wooden round and mounting them with a large screw at each end (see photos for details). The cage turned out to be very rigid, better than I expected.
To attach the focuser I turned to some 3mm aluminium plate I had lying around. I bent this to follow the inner OTA radius and made holes for attaching the focuser. I then chiseled out groowes on each cage ring and screwed the plate onto these, then mounted the focuser onto the plate.
Now I had to weigh the components and make sure the central brace would be placed at the balance point which is the purpose of the Serrurier truss design.
I hanged a board, of the same length as the OTA would be, from the ceiling and then hanged the mirror cell, with mirror attached, to one end. Then I hanged the secondary cage and the secondary holder + mirror to the other end. To make up for the weight of the CCD camera I plan to purchase I added a bag with a book of similar weight to the secondary end. I then shifted the attachement of the board around until the two ends were in perfect balance, and I could then simply measure the distances on the board to know how to place the components when assembling the trusses.
I laid out the components in the correct positions and measured the required length of the first truss tube for the upper half. I then cut that tube plus the other 7 tubes to the same length. Then I put the first tube in place and drilled a hole all the way through both the tube and the alu angle extrusion on the central brace which would be its attachment point. The tube was then fastened with a bolt with washers and lock nut. I repeated this for each tube, carefully measuring that the central brace and the secondary cage were aligned at all times. This whole process was then repeated when attaching the mirror cell.
For the spider I considered a number of different options. My old spider had 4 vanes made of thin metal bands. Even though the vanes are very thin the reality is that they are more often than not slightly twisted and so produce rather large 'fat' diffraction spikes. Also, no matter how thin this kind of spider vane is, it will only look thin to a light ray that enters exactly along the optical axis of the telescope. From any other angle it will inevitably look much fatter than a thin steel wire and thus produce much more prominent diffraction spikes. So I decided to instead make a offset wire spider. I got some inspiration from Mark Cowan's minimalist wire spider design (http://www.raddobs.com/atm/wire%20spider/wirespider.htm) but I decided to go for an absolute minimalist design where the entire positioning and collimation of the secondary is controlled only by the wires.
For the wire itself I used 9 gauge (0.23 mm) guitar steel strings. After toying with various ideas for fastening the strings to the secondary cage I simply went for guitar machine heads. I did this to ensure I would obtain smooth and exact control of the spider tensioning and positioning, and because it was just easier and more elegant than any home made solution I could think of. I went for 4 wire vanes rather than 3 simply because I like the cross-shaped diffraction pattern with 4 spikes better than the 6 I would get from having 3 vanes. Another reason was that there for a cross is more space between each of the 4 spikes which could be handy in case I will be imaging something that lies very close to a bright star.
I attached the machine heads to small wooden blocks that can pivot around a screw. This way I got full control of the angle of each individual string, which is important in order to line up the strings completely and minimize the diffraction.
The mirror cell was completed by re-using part of the cell from my old telescope. The reason for this was that it was straight forward and would provide a stable and secure platform for the mirror without me having to make any additional parts. The old cell consisted of an aluminium plate with a 1cm high edge around the mirror. At the bottom of the plate were three cork pads to support the mirror, and on the sides were attached three clips to hold the mirror in place. These clips extended over the front of the mirror so to minimise diffraction effects I decided to get rid of them and instead simply mount the mirror on the plate with silicone. I removed the cork pads and put three blobs of silicone on the plate plus some spacers and then I lowered the mirror onto the blobs and let it rest on the spacers. Once the silicone had cured I then removed the spacers so the mirror was entirely held in place by only the silicone. I also added three blobs through the existing screw holes on the edge which were left from the clips. This fixated the mirror horizontally. Finally I wrapped black electrical tape around the outside of the plate right up to the edge of the mirror. I did this to further fix the mirror in place but also to enable my mirror dust cover to easily slip onto the aluminium plate, which has a slightly larger diameter than the mirror.
The dust cover was made from a black styrene sheet. I cut out a circle to fit the mirror holder and a long strip to wrap around in order to make a cap for the mirror. I welded the two parts together using some liquid plastic glue I had left from my model building days. This glue was some 20 years old, but still worked perfectly, leaving the resulting dust cap very strong and durable.
Yesterday I got the first chance to test the performance of this OTA. The results are excellent, and more than I hopped for in my wildest dreams! I guess this just shows how well the original Serrurier truss design works.
Primarily I was looking to minimise focus drift and flexure with this project, and both have been accomplished to the utmost degree. Over the course of 2 hours I saw absolutely no shift in focus while imaging the Jewel Box cluster and Spica continually. This has always plagued me with my previous sonotube OTA but I'm happy to conclude that so far this has been completely eleminated. It also appears that the OTA holds collimation extremely well in all orientations. I have used the scope for two nights now and experienced absolutely no degration in collimation since completion of the project.
Another major improvement is the diffraction pattern which was pretty bad in my old OTA. To illustrate the difference here is first an image of Betelgeuse taken with the old OTA:
http://www.pbase.com/rolfolsen/image/134141847/original
Notice the fat diffraction spikes and just how much of the light is diverted away from the star image itself. Then for comparison this is Spica imaged with the new OTA and wire spider:
http://www.pbase.com/rolfolsen/image/134141853/original
The diffraction spikes are much thinner and fainter! The starlight is clearly more concentrated in a single point, as it should be.
Finally, I took a quick test image of the Jewel Box cluster, just to see what it would look like now:
http://www.pbase.com/rolfolsen/image/134141855/original
There are practically no diffraction spikes visible. While the presence of spikes is a matter of taste, I personally prefer to know that the precious light gathered by the optics is not spread out all over the field, since this decrades contrast and resolution and could make the difference in capturing a faint object or not.
So overall I could not be more pleased, and I look forward to be imaging with this new OTA.
Again, thanks to all who gave me advice regarding this project.
For many more images of the project see the gallery at http://www.pbase.com/rolfolsen/10_inch_serrurier_truss_tube_newton ian
My major inspiration came from this beautifully constructed Newtonian: http://www.aao.gov.au/local/www/sl/sl-tels.html#ni-tel
For many more images of the project see the gallery at http://www.pbase.com/rolfolsen/10_inch_serrurier_truss_tube_newton ian
This project was intended to be a cheap upgrade I could work on while saving for a new CCD cam. To give the upcoming new cam justice I didn't want to mount it on my existing OTA which is a 17 year old simple sonotube, squeezed out of shape by the secondary spider supports and generally emitting a slight mouldy odour - probably due to the many years of dew and moisture absorbtion I guess. It was definitely time for something new.
While I intend to purchase a better mirror (a conical Royce mirror seems tempting) I decided to build a 10" to keep costs down, since my current telescope is of this size and so I could simply reuse the optics for the time being. Also, I was not sure how much load I could safely put on the G-11 mount in practice while still maintaining smooth tracking for photography purposes. While a 12" would be great I felt I'd rather have a stable tracking 10" scope than a 12", or larger, which may cause problems for the mount. Furthermore, the 10" is just about the right size for my observatory, any larger and I think it might start to become too bulky to maneuver around in the dark - otherwise I would have to build a larger observatory...
The design consists of three parts: The mirror cell, the central box/brace and the secondary cage.
The central brace was the first part I built, since this was the most critical component as it must be very strong and rigid in order to carry the telescope and eliminate flexure. I made this out of aluminium square tubing and some pieces of 40x40x3mm aluminium angle extrusions. I cut 12 pieces of square tubing at 45 degree anges and riveted them to the angles. The protuding angles would then later function as truss connectors. This structure alone was surprisingly rigid and ensured that everything stayed in place, but as this component would hold the entire weight of the OTA I asked a local metal smith to weld all the joints on it. The final resulting central brace was very light, beautifully simple, and yet extremely sturdy.
Next thing to make was the mirror cell and secondary cage. I decided to make these from plywood as it would be much easier for me than using aluminium, and it seemed to be the long accepted tried and tested solution of ATM's. For the secondary cage I cut two identical rings with an inner diameter of 30cm from the plywood. For the mirror cell I realised the ring could be smaller (saving a bit of weight), since being close to the mirror it would not interfere with the optical path, so I cut this ring with an inner diameter of 27cm to acommodate the 25cm mirror (10 inches) plus a bit of wriggle room for adjustments. The adjustable mirror cell was then made by cutting a trifoil shape of the plywood and mounting this on the ring with three spring loaded bolts.
The two rings for the secondary cage were connected with 4 round 1" alu tube pieces. I thought about how to connect the hollow tubes for a while before deciding to simply slip them over a wooden round and mounting them with a large screw at each end (see photos for details). The cage turned out to be very rigid, better than I expected.
To attach the focuser I turned to some 3mm aluminium plate I had lying around. I bent this to follow the inner OTA radius and made holes for attaching the focuser. I then chiseled out groowes on each cage ring and screwed the plate onto these, then mounted the focuser onto the plate.
Now I had to weigh the components and make sure the central brace would be placed at the balance point which is the purpose of the Serrurier truss design.
I hanged a board, of the same length as the OTA would be, from the ceiling and then hanged the mirror cell, with mirror attached, to one end. Then I hanged the secondary cage and the secondary holder + mirror to the other end. To make up for the weight of the CCD camera I plan to purchase I added a bag with a book of similar weight to the secondary end. I then shifted the attachement of the board around until the two ends were in perfect balance, and I could then simply measure the distances on the board to know how to place the components when assembling the trusses.
I laid out the components in the correct positions and measured the required length of the first truss tube for the upper half. I then cut that tube plus the other 7 tubes to the same length. Then I put the first tube in place and drilled a hole all the way through both the tube and the alu angle extrusion on the central brace which would be its attachment point. The tube was then fastened with a bolt with washers and lock nut. I repeated this for each tube, carefully measuring that the central brace and the secondary cage were aligned at all times. This whole process was then repeated when attaching the mirror cell.
For the spider I considered a number of different options. My old spider had 4 vanes made of thin metal bands. Even though the vanes are very thin the reality is that they are more often than not slightly twisted and so produce rather large 'fat' diffraction spikes. Also, no matter how thin this kind of spider vane is, it will only look thin to a light ray that enters exactly along the optical axis of the telescope. From any other angle it will inevitably look much fatter than a thin steel wire and thus produce much more prominent diffraction spikes. So I decided to instead make a offset wire spider. I got some inspiration from Mark Cowan's minimalist wire spider design (http://www.raddobs.com/atm/wire%20spider/wirespider.htm) but I decided to go for an absolute minimalist design where the entire positioning and collimation of the secondary is controlled only by the wires.
For the wire itself I used 9 gauge (0.23 mm) guitar steel strings. After toying with various ideas for fastening the strings to the secondary cage I simply went for guitar machine heads. I did this to ensure I would obtain smooth and exact control of the spider tensioning and positioning, and because it was just easier and more elegant than any home made solution I could think of. I went for 4 wire vanes rather than 3 simply because I like the cross-shaped diffraction pattern with 4 spikes better than the 6 I would get from having 3 vanes. Another reason was that there for a cross is more space between each of the 4 spikes which could be handy in case I will be imaging something that lies very close to a bright star.
I attached the machine heads to small wooden blocks that can pivot around a screw. This way I got full control of the angle of each individual string, which is important in order to line up the strings completely and minimize the diffraction.
The mirror cell was completed by re-using part of the cell from my old telescope. The reason for this was that it was straight forward and would provide a stable and secure platform for the mirror without me having to make any additional parts. The old cell consisted of an aluminium plate with a 1cm high edge around the mirror. At the bottom of the plate were three cork pads to support the mirror, and on the sides were attached three clips to hold the mirror in place. These clips extended over the front of the mirror so to minimise diffraction effects I decided to get rid of them and instead simply mount the mirror on the plate with silicone. I removed the cork pads and put three blobs of silicone on the plate plus some spacers and then I lowered the mirror onto the blobs and let it rest on the spacers. Once the silicone had cured I then removed the spacers so the mirror was entirely held in place by only the silicone. I also added three blobs through the existing screw holes on the edge which were left from the clips. This fixated the mirror horizontally. Finally I wrapped black electrical tape around the outside of the plate right up to the edge of the mirror. I did this to further fix the mirror in place but also to enable my mirror dust cover to easily slip onto the aluminium plate, which has a slightly larger diameter than the mirror.
The dust cover was made from a black styrene sheet. I cut out a circle to fit the mirror holder and a long strip to wrap around in order to make a cap for the mirror. I welded the two parts together using some liquid plastic glue I had left from my model building days. This glue was some 20 years old, but still worked perfectly, leaving the resulting dust cap very strong and durable.
Yesterday I got the first chance to test the performance of this OTA. The results are excellent, and more than I hopped for in my wildest dreams! I guess this just shows how well the original Serrurier truss design works.
Primarily I was looking to minimise focus drift and flexure with this project, and both have been accomplished to the utmost degree. Over the course of 2 hours I saw absolutely no shift in focus while imaging the Jewel Box cluster and Spica continually. This has always plagued me with my previous sonotube OTA but I'm happy to conclude that so far this has been completely eleminated. It also appears that the OTA holds collimation extremely well in all orientations. I have used the scope for two nights now and experienced absolutely no degration in collimation since completion of the project.
Another major improvement is the diffraction pattern which was pretty bad in my old OTA. To illustrate the difference here is first an image of Betelgeuse taken with the old OTA:
http://www.pbase.com/rolfolsen/image/134141847/original
Notice the fat diffraction spikes and just how much of the light is diverted away from the star image itself. Then for comparison this is Spica imaged with the new OTA and wire spider:
http://www.pbase.com/rolfolsen/image/134141853/original
The diffraction spikes are much thinner and fainter! The starlight is clearly more concentrated in a single point, as it should be.
Finally, I took a quick test image of the Jewel Box cluster, just to see what it would look like now:
http://www.pbase.com/rolfolsen/image/134141855/original
There are practically no diffraction spikes visible. While the presence of spikes is a matter of taste, I personally prefer to know that the precious light gathered by the optics is not spread out all over the field, since this decrades contrast and resolution and could make the difference in capturing a faint object or not.
So overall I could not be more pleased, and I look forward to be imaging with this new OTA.
Again, thanks to all who gave me advice regarding this project.
For many more images of the project see the gallery at http://www.pbase.com/rolfolsen/10_inch_serrurier_truss_tube_newton ian