Making a LHN Telescope
Optics of Lurie-Houghton Newtonian
(L)urie-(H)oughton (N)ewtonian telescope typically consists of a full size corrector located on the upper end of the tube, a spherical primary and a flat secondary mirror. The corrector usually is designed to have two lenses: A bi-convex(positive) front and a bi-concave(negative) rear lens. The lenses have a small air gap between each other. Also the lens glass material can be the same, ie. a good choice is, for example, Schott's BK7 or Pilkington's BCS grown glass. Actually the correct name for this kind of telescope should be "Houghton-Newtonian". I have read that people are interested in to make corrector lenses of normal plate glass. My opinion is that why to risk and spend enormously time for unsure glass material if there are am. good materials available? Of coarse a good one always costs more. That's the life. To design and calculate a LHN telescope setup a good tool is OSLO EDU or MODAS analysis software. Both can be free downloaded from the net. Also very useful web-pages are Vladimir Sacek's "Amateur Telescope Optics".
A very good, informative web-page of designing and making a LHN telescope is Rick Scott's "Lurie-Houghton Telescope Project". What are then the advantages and disadvantages of the LHN telescope? And is there any common sense to make one? A good normal Newtonian telescope performs just perfectly and is more than enough for most of us. I think now we are starting to talk about a topic which is not black and white thing. But if you try to find out pure physical facts, some of them can be listed:
Advantages:
- rather fast telescope F/4...F/5 (typically)
- large field of view
- curvature of field is almost flat
- coma free
- there is astigmatism but not a big issue
- corrector seals the tube end reducing tube turbulence
- no secondary spider vanes
- tube is "dust sealed"
- optics is easier to make, all surfaces are spherical
(L)urie-(H)oughton (N)ewtonian telescope typically consists of a full size corrector located on the upper end of the tube, a spherical primary and a flat secondary mirror. The corrector usually is designed to have two lenses: A bi-convex(positive) front and a bi-concave(negative) rear lens. The lenses have a small air gap between each other. Also the lens glass material can be the same, ie. a good choice is, for example, Schott's BK7 or Pilkington's BCS grown glass. Actually the correct name for this kind of telescope should be "Houghton-Newtonian". I have read that people are interested in to make corrector lenses of normal plate glass. My opinion is that why to risk and spend enormously time for unsure glass material if there are am. good materials available? Of coarse a good one always costs more. That's the life. To design and calculate a LHN telescope setup a good tool is OSLO EDU or MODAS analysis software. Both can be free downloaded from the net. Also very useful web-pages are Vladimir Sacek's "Amateur Telescope Optics".
A very good, informative web-page of designing and making a LHN telescope is Rick Scott's "Lurie-Houghton Telescope Project". What are then the advantages and disadvantages of the LHN telescope? And is there any common sense to make one? A good normal Newtonian telescope performs just perfectly and is more than enough for most of us. I think now we are starting to talk about a topic which is not black and white thing. But if you try to find out pure physical facts, some of them can be listed:
Advantages:
- rather fast telescope F/4...F/5 (typically)
- large field of view
- curvature of field is almost flat
- coma free
- there is astigmatism but not a big issue
- corrector seals the tube end reducing tube turbulence
- no secondary spider vanes
- tube is "dust sealed"
- optics is easier to make, all surfaces are spherical
Disadvantages:
- telescope is heavier because of the corrector
- obstruction is easily 30..33% depending on the requested FOV
- light loss because of two corrector lenses
- there is for sure a lot of work with the optics (5 surfaces)
These are the first things which come to my mind. My reasons to choose making LHN telescope was that it really is a good telescope type and I liked to practice and learn my ATM skills with lenses too. I had already made a couple of paraboloidal mirrors, so trying out lens making was evident. Optics of the LHN telescope is quite forgiving. I made at first the primary mirror trying to match the radius of curvature as close as possible to the calculated one. It doesn't matter if you do not reach the RoC exactly. The new RoC value can always be fed to OSLO file and correct the lenses' values, no problem. Maybe the final F/ratio will only change a little bit. During playing with my LHN setup in OSLO I noticed that R1 (front lens's first surface), R3 (rear lens's first surface) and R5 (primary) are the most critical ones. Ie. RoC of R1 and R3 should not only be within +/- "something" millimeters tolerance, they should also have exactly the same but opposite RoC value (R1 is convex, R3 is concave). But the good thing is that R1/R3 together can vary about +/- 5..7 mm of the calculated values. If the R1 (convex) is interference tested against polished surface of R3 the overall 3 fringe difference is about the maximum acceptable value. The primary mirror's acceptable value is app. +/- 3...4 mm if the corrector lenses are made first. So, I would say that it's better to make first the primary, adjust the values of the corrector in OSLO and then make the corrector lenses within that +/- 5...7 mm tolerance.
R2/R4 tolerances behave the same way as R1/R3. You should get both RoCs as close to the same value as possible but together they can vary and this time much more than R1/R3. For example in my setup a good value for R2/R4 is 3000 mm (R2=convex, R4=concave). OSLO Lt clearly shows that the value for RoCs could vary easily +/- 100 mm even more without noticeable destroying the final spot diagrams! If you have a spherometer at least 50% of lens diameter it should not be very difficult to achive much more accurate results than that. Actually if you grind lens pairs together checking concave's RoC is easy. Then there is that little air gap between lenses. At least what I tried to figure out with OSLO it seems that it's enough to keep it small. If between lenses is placed three pieces thin tapes or post stamp papers is enough. Then the gap on lenses edges are about 0,05...0,1 mm. Lenses thickness is once again a glad thing in LHN corrector case. Even if refractors' lenses thickness has to be quite accurate for LHN corrector lenses 1...2 mm difference (from designed) is nothing. Thiner is better but as we know a good starting point for lens disks is 1:10. Finally there is in LHN telescope setup the corrector's place compared to the other optical elements. This may start to sound monotonic but also the corrector's place on the optical axis is not so accurate. Meaning that you need not to design any specific high accurate secondary supporting system to locate the corrector exactly in the pre-calculated place! It's good to remember that I am writing these things from an ATM optics maker's point of view not from pure optics designer's view. Optics designers would try to make a perfect telescope which has a perfect spot diagrams. Yes, by giving independent values for each RoC of lenses is possible to make a little bit better LHN telescope especially Houghton-Cassegrain case. But a very good design starting point is in "Telescope Optics: Evaluation and Design" - book (page 126). The shown setup for LHN is the-all- spherical design (with R1/R3 and R2/R4 RoC pairs) and the results are more than adequate.
Making the lenses
In the first picture I am rough grinding R3 on R1. It's a nice thing in LHN corrector lenses that you can design the setup so that R1/R3 and R2/R4 surface pairs have exactly the same radius. Then you can rough grind those surface pairs on each other saving time a lot. Of coarse you can fine grind them in the same way or use a grinding machine and independent tools for each surface, as I did. In this point I have to mension "a pitch tile tool", ie. a grinding tool covered with tiles which are clued by using normal pitch. I made those tools after reading Mr. Roger Ceragioli's article "Refractor Construction Page" My opinion is that those tools work much better than normal epoxy tools. All the time the fine grinding action was good and the tools stayed in really good contact. That is a method which can be recommended honestly at least for lens making.
I had a little bit experience of trepaning mirrors, so making the lenses center holes for connecting the secondary went quite nicely. As I had read about crowns and flints that they are much more breakable than Pyrex or Duran50 glass, it seemed to be true. I was able to use a drill press which had a centering base. Then you can center a biscuit cutter exactly to the center of a lens. If you do not touch to the settings you can drill 3..4 mm from one side and turn over the lens and drill now until you reach the first side's pre-drilling. That way you can quite safely drill through a lens without a fear that you will break pieces off the edge of the center hole. A crown glass really is easily breakable!
I think that it's better to trepan lenses after you have ended the 220 carbo stage or so. That's the same no matter are you making a lens or a mirror. The reason is that after 220 carbo you have already grinded off the material which is needed to reach the RoC (especially in concave case). If you now make the center holes the required bevels in the edge of the hole and the core will last to the end of the fine grinding. Without a fear of taking the core off, re-beveling and plastering the core back...Oh boy what a mess and work! Okay, this is a quite small thing but still...please believe me. :)
My lenses are Pilkington's BSC crown glass which is equivalent to well-known BK7. I got the glass as a 15 mm thick molded glass slab (400x240 mm). Because of my work I got two 150 mm round glass disks water jet cutted. The cutted edge surfaces were quite smooth and both disks were in the same size enough to start grinding. I thought a lot how to edge lenses with my home tools. Then I figured out that the only way is to connect them to my grinding machine's shaft through the lens holes with a bolt after the lenses were trepaned. So, just after 220 carbo stage both lenses got their exact diameters. I bolted both lenses together to the shaft with a bolt and a large steel and rubber plate. I adjusted the lenses until they rotated centerly as possible. Then I used a diamond wet stone with a stand and grinded the lens edges to the same diameter. It took about 10 minutes. Finally I used a piece of steel plate and 20 um al-oxide to fine grind the edges.
After trepaning and edging the cores have to be clued back with Plaster of Paris. I think that stuff is well-known all over the world. Here is one method which I have used: Take a piece of plastic contact sheet and clue it to the backside of the lens or mirror. Now lay down the core from other side of the lens and quide it through the hole with three pieces of wood or plastic sticks. The tichness of the sticks should be the same as the gap between the core and center hole. The core should stick to the backside plastic tape rather well. Now you can pour thin, watery Plaster of Paris around the core. Normally I have layed down the lens on it's pair lens, only the plastic contact sheet between them. If you now make the plastering you can be quite sure that the core is almost in the same height as the lens itself. Finally after the plaster is dryed you have to take out a little bit of the plaster from both sides of the lens. Let's say app. 3 mm. The gaps can be filled carefully with candle vax or equal. If you don't do that there is a big risk that the core get loose. Candle vax is maybe not the best possible material. I tried also sanitate silicone. It was good stuff but making the seal was quite difficult. Anyway, the purpose is to make the plaster waterproof.
In the beginning of the carbo 220 stage at the latest you have to start to measure the wedge of the lenses. I made a wedge tester just from left- over plywood. In Texereau's "How to Make a Telescope" - book on page 200 is described well how de-wedging should be done. If you have done by manual method a mirror you learn de-wedging once you have tried it. The question is just manually grind mirror without rotating the upper disk in your hands. The thickest edge is placed towards you. Just little bit harder pressing during grinding. After 5 minutes you have to make 15 minutes normal grinding to ensure the surfaces are spherical enough before wedge testings. This way it has to be continued with both lenses until in the finest grades the wedge is about 0.025 mm. I left the both lenses to value 0.03 mm. In LHN corrector case it seems that the wedge can be much bigger than 0.025 mm. Anyway, Mr. Ceragioli gives in his article good values for wedge which can be and is realistic to achieve in different grades. So, I made de-wedging allways in the beginning of the new grinding stage and then continued until the surface marks were gone. Of coarse in the end I checked the wedge value before moving finer stage. I faced a surface quality problem with 2 surfaces, so I had to come back to coarse stages. With machine it seemed to be so that the wedge did not change even if I took 3 the finest grades again.
Polishing the lenses is quite the same thing compared to mirror polishing. I had polished a few mirrors with my machine so the concave surfaces R3/R4 were quite easy job. Only it's good to remember that thin lenses have to be supported well with, for example, a concrete base. Between the support base and the lens I used 2 mm rubber plate. A couple of turns of plastic tape around base/lens is adequate to keep them together. R1/R2 of the front lens are convex. I have used my Mirror-O-Matic style machine for some time and typical for this kind of machine is to use sub-diameter tools for grinding and polishing. 75% tool is good value and it always works well for a concave surface. Now I had a good situation to try out would a small tool work also for a convex surface. Actually I used full-sized-tool for R1 and 75% tool for R2. Both styles worked well. I achieved rather good spherical figures and app. that 3 fringes match for concave pairs. Of coarse it's good to remember that all surfaces have to be polished about half way so that you can see through a lens while making interference testing. And if a correction need is detected a half-way-polished surface is much easier to correct. I have had always little problems to get a perfectly polished optical surface without pits. Ie. many times I have had to come back a few stages and fine grind again. In these corrector lens cases I happily noticed that 2 surfaces came out at first trial very nicely. Either the crown optical glass is easier to grind or the new tile-pitch-tool worked so much better. I would say the tool was the biggest reason. At least I like to recommend it's use.
Finally after 5 months I got the whole telescope optics ready for testing. Now (08-Jan-2005) the primary is in Helsinki for aluminaizing but before sending it I tested the telescope twice. In star testing from 5 to 10 rings visible intra and extra images were very much equal. The outer most ring's texture, thickness and brightness looked good in my eyes. I am not an expert, can only compare to my normal Newtonian telescopes. Also if I came very close to focus point the donut was round indicating no astigmatism. The subjective "snap" test also worked as it should do. Ie. a star or the surface of the Moon snapped right in one point in the best focus. Actually I was a little bit suprised, so well the telescope performed! Well, as I have read the zero power corrector seems to work properly producing just opposite direction spherical aberrations for the spherical primary. No lateral colour as in a refractor, just pure nice and sharp images.
With Parks Erfle 20 mm eyepiece the telescope produces magnification 32x. I looked the full moon and estimated FOV was 3.5 full moons. The secondary size is 1.83" (46.5 mm) and the obstruction then 32.7 %. I look forward to continue my testings after the primary mirror comes back from aluminaizing.
Well, the Lurie-Houghton Newtonian telescope seems to perform as well as I have read. If consider the work amount which is huge compared to an ordinary Newtonian LHN telescope splits amateurs' thoughts I'm sure. If you are willing to spend a lot of your time and also money for optics making it's a very good choice. If observing is the main point and hobby, buying or making a good ordinary Newtonian is the right way to go. But if your hobby is making telescope optics as mine is, do not fear to try out LHN optics. In principle there is combined two telescopes with their good features in one package: a refractor and a reflector!
- telescope is heavier because of the corrector
- obstruction is easily 30..33% depending on the requested FOV
- light loss because of two corrector lenses
- there is for sure a lot of work with the optics (5 surfaces)
These are the first things which come to my mind. My reasons to choose making LHN telescope was that it really is a good telescope type and I liked to practice and learn my ATM skills with lenses too. I had already made a couple of paraboloidal mirrors, so trying out lens making was evident. Optics of the LHN telescope is quite forgiving. I made at first the primary mirror trying to match the radius of curvature as close as possible to the calculated one. It doesn't matter if you do not reach the RoC exactly. The new RoC value can always be fed to OSLO file and correct the lenses' values, no problem. Maybe the final F/ratio will only change a little bit. During playing with my LHN setup in OSLO I noticed that R1 (front lens's first surface), R3 (rear lens's first surface) and R5 (primary) are the most critical ones. Ie. RoC of R1 and R3 should not only be within +/- "something" millimeters tolerance, they should also have exactly the same but opposite RoC value (R1 is convex, R3 is concave). But the good thing is that R1/R3 together can vary about +/- 5..7 mm of the calculated values. If the R1 (convex) is interference tested against polished surface of R3 the overall 3 fringe difference is about the maximum acceptable value. The primary mirror's acceptable value is app. +/- 3...4 mm if the corrector lenses are made first. So, I would say that it's better to make first the primary, adjust the values of the corrector in OSLO and then make the corrector lenses within that +/- 5...7 mm tolerance.
R2/R4 tolerances behave the same way as R1/R3. You should get both RoCs as close to the same value as possible but together they can vary and this time much more than R1/R3. For example in my setup a good value for R2/R4 is 3000 mm (R2=convex, R4=concave). OSLO Lt clearly shows that the value for RoCs could vary easily +/- 100 mm even more without noticeable destroying the final spot diagrams! If you have a spherometer at least 50% of lens diameter it should not be very difficult to achive much more accurate results than that. Actually if you grind lens pairs together checking concave's RoC is easy. Then there is that little air gap between lenses. At least what I tried to figure out with OSLO it seems that it's enough to keep it small. If between lenses is placed three pieces thin tapes or post stamp papers is enough. Then the gap on lenses edges are about 0,05...0,1 mm. Lenses thickness is once again a glad thing in LHN corrector case. Even if refractors' lenses thickness has to be quite accurate for LHN corrector lenses 1...2 mm difference (from designed) is nothing. Thiner is better but as we know a good starting point for lens disks is 1:10. Finally there is in LHN telescope setup the corrector's place compared to the other optical elements. This may start to sound monotonic but also the corrector's place on the optical axis is not so accurate. Meaning that you need not to design any specific high accurate secondary supporting system to locate the corrector exactly in the pre-calculated place! It's good to remember that I am writing these things from an ATM optics maker's point of view not from pure optics designer's view. Optics designers would try to make a perfect telescope which has a perfect spot diagrams. Yes, by giving independent values for each RoC of lenses is possible to make a little bit better LHN telescope especially Houghton-Cassegrain case. But a very good design starting point is in "Telescope Optics: Evaluation and Design" - book (page 126). The shown setup for LHN is the-all- spherical design (with R1/R3 and R2/R4 RoC pairs) and the results are more than adequate.
Making the lenses
In the first picture I am rough grinding R3 on R1. It's a nice thing in LHN corrector lenses that you can design the setup so that R1/R3 and R2/R4 surface pairs have exactly the same radius. Then you can rough grind those surface pairs on each other saving time a lot. Of coarse you can fine grind them in the same way or use a grinding machine and independent tools for each surface, as I did. In this point I have to mension "a pitch tile tool", ie. a grinding tool covered with tiles which are clued by using normal pitch. I made those tools after reading Mr. Roger Ceragioli's article "Refractor Construction Page" My opinion is that those tools work much better than normal epoxy tools. All the time the fine grinding action was good and the tools stayed in really good contact. That is a method which can be recommended honestly at least for lens making.
I had a little bit experience of trepaning mirrors, so making the lenses center holes for connecting the secondary went quite nicely. As I had read about crowns and flints that they are much more breakable than Pyrex or Duran50 glass, it seemed to be true. I was able to use a drill press which had a centering base. Then you can center a biscuit cutter exactly to the center of a lens. If you do not touch to the settings you can drill 3..4 mm from one side and turn over the lens and drill now until you reach the first side's pre-drilling. That way you can quite safely drill through a lens without a fear that you will break pieces off the edge of the center hole. A crown glass really is easily breakable!
I think that it's better to trepan lenses after you have ended the 220 carbo stage or so. That's the same no matter are you making a lens or a mirror. The reason is that after 220 carbo you have already grinded off the material which is needed to reach the RoC (especially in concave case). If you now make the center holes the required bevels in the edge of the hole and the core will last to the end of the fine grinding. Without a fear of taking the core off, re-beveling and plastering the core back...Oh boy what a mess and work! Okay, this is a quite small thing but still...please believe me. :)
My lenses are Pilkington's BSC crown glass which is equivalent to well-known BK7. I got the glass as a 15 mm thick molded glass slab (400x240 mm). Because of my work I got two 150 mm round glass disks water jet cutted. The cutted edge surfaces were quite smooth and both disks were in the same size enough to start grinding. I thought a lot how to edge lenses with my home tools. Then I figured out that the only way is to connect them to my grinding machine's shaft through the lens holes with a bolt after the lenses were trepaned. So, just after 220 carbo stage both lenses got their exact diameters. I bolted both lenses together to the shaft with a bolt and a large steel and rubber plate. I adjusted the lenses until they rotated centerly as possible. Then I used a diamond wet stone with a stand and grinded the lens edges to the same diameter. It took about 10 minutes. Finally I used a piece of steel plate and 20 um al-oxide to fine grind the edges.
After trepaning and edging the cores have to be clued back with Plaster of Paris. I think that stuff is well-known all over the world. Here is one method which I have used: Take a piece of plastic contact sheet and clue it to the backside of the lens or mirror. Now lay down the core from other side of the lens and quide it through the hole with three pieces of wood or plastic sticks. The tichness of the sticks should be the same as the gap between the core and center hole. The core should stick to the backside plastic tape rather well. Now you can pour thin, watery Plaster of Paris around the core. Normally I have layed down the lens on it's pair lens, only the plastic contact sheet between them. If you now make the plastering you can be quite sure that the core is almost in the same height as the lens itself. Finally after the plaster is dryed you have to take out a little bit of the plaster from both sides of the lens. Let's say app. 3 mm. The gaps can be filled carefully with candle vax or equal. If you don't do that there is a big risk that the core get loose. Candle vax is maybe not the best possible material. I tried also sanitate silicone. It was good stuff but making the seal was quite difficult. Anyway, the purpose is to make the plaster waterproof.
In the beginning of the carbo 220 stage at the latest you have to start to measure the wedge of the lenses. I made a wedge tester just from left- over plywood. In Texereau's "How to Make a Telescope" - book on page 200 is described well how de-wedging should be done. If you have done by manual method a mirror you learn de-wedging once you have tried it. The question is just manually grind mirror without rotating the upper disk in your hands. The thickest edge is placed towards you. Just little bit harder pressing during grinding. After 5 minutes you have to make 15 minutes normal grinding to ensure the surfaces are spherical enough before wedge testings. This way it has to be continued with both lenses until in the finest grades the wedge is about 0.025 mm. I left the both lenses to value 0.03 mm. In LHN corrector case it seems that the wedge can be much bigger than 0.025 mm. Anyway, Mr. Ceragioli gives in his article good values for wedge which can be and is realistic to achieve in different grades. So, I made de-wedging allways in the beginning of the new grinding stage and then continued until the surface marks were gone. Of coarse in the end I checked the wedge value before moving finer stage. I faced a surface quality problem with 2 surfaces, so I had to come back to coarse stages. With machine it seemed to be so that the wedge did not change even if I took 3 the finest grades again.
Polishing the lenses is quite the same thing compared to mirror polishing. I had polished a few mirrors with my machine so the concave surfaces R3/R4 were quite easy job. Only it's good to remember that thin lenses have to be supported well with, for example, a concrete base. Between the support base and the lens I used 2 mm rubber plate. A couple of turns of plastic tape around base/lens is adequate to keep them together. R1/R2 of the front lens are convex. I have used my Mirror-O-Matic style machine for some time and typical for this kind of machine is to use sub-diameter tools for grinding and polishing. 75% tool is good value and it always works well for a concave surface. Now I had a good situation to try out would a small tool work also for a convex surface. Actually I used full-sized-tool for R1 and 75% tool for R2. Both styles worked well. I achieved rather good spherical figures and app. that 3 fringes match for concave pairs. Of coarse it's good to remember that all surfaces have to be polished about half way so that you can see through a lens while making interference testing. And if a correction need is detected a half-way-polished surface is much easier to correct. I have had always little problems to get a perfectly polished optical surface without pits. Ie. many times I have had to come back a few stages and fine grind again. In these corrector lens cases I happily noticed that 2 surfaces came out at first trial very nicely. Either the crown optical glass is easier to grind or the new tile-pitch-tool worked so much better. I would say the tool was the biggest reason. At least I like to recommend it's use.
Finally after 5 months I got the whole telescope optics ready for testing. Now (08-Jan-2005) the primary is in Helsinki for aluminaizing but before sending it I tested the telescope twice. In star testing from 5 to 10 rings visible intra and extra images were very much equal. The outer most ring's texture, thickness and brightness looked good in my eyes. I am not an expert, can only compare to my normal Newtonian telescopes. Also if I came very close to focus point the donut was round indicating no astigmatism. The subjective "snap" test also worked as it should do. Ie. a star or the surface of the Moon snapped right in one point in the best focus. Actually I was a little bit suprised, so well the telescope performed! Well, as I have read the zero power corrector seems to work properly producing just opposite direction spherical aberrations for the spherical primary. No lateral colour as in a refractor, just pure nice and sharp images.
With Parks Erfle 20 mm eyepiece the telescope produces magnification 32x. I looked the full moon and estimated FOV was 3.5 full moons. The secondary size is 1.83" (46.5 mm) and the obstruction then 32.7 %. I look forward to continue my testings after the primary mirror comes back from aluminaizing.
Well, the Lurie-Houghton Newtonian telescope seems to perform as well as I have read. If consider the work amount which is huge compared to an ordinary Newtonian LHN telescope splits amateurs' thoughts I'm sure. If you are willing to spend a lot of your time and also money for optics making it's a very good choice. If observing is the main point and hobby, buying or making a good ordinary Newtonian is the right way to go. But if your hobby is making telescope optics as mine is, do not fear to try out LHN optics. In principle there is combined two telescopes with their good features in one package: a refractor and a reflector!
Copyright©2008 Aki Lötjönen
A few photos with Houghton-Newtonian telescope: