Fixing the dreaded tilt and backspacing error in optical systems with objective analysis.

#...

Jure:
Thanks for the report Bill! Fantastic! My Photon Cage is in transit for a while now (Europe) and I can't wait to finally get rid of tilt on my ASI2600MM. Question: do you use a filter wheel in your imaging train? As far as I could see the Cage is constructed in such a way that EFW does not block access.

Here I am using the QHY CFW3-L with the dovetail adapter removed, and the Photon Cage threaded into the M54 threads on the wheel. I had Josh put male M54 threads on the Cage so I can do this.

11.02 Bill Long - Dark Matters Astrophotography #... · view views · 1 like
I would add, that Josh may want to think about offering a bottom plate with threaded holes to match the QHY/Pegasus/ZWO pattern on the wheels. This would let people bolt the cage to the wheel instead of using threads. The small problem with this, is that a QHY version would need a small lip on it to seat into the wheel. The Pegasus wheel does not need this lip. I also think, if I recall correctly, the ZWO wheel has a different hole pattern, but no lip is needed. Need to express this point though -- the Photon Cage is excellent. Simple, yet elegant design. Edited ... Like

11.02

#...

· 1 like

I would add, that Josh may want to think about offering a bottom plate with threaded holes to match the QHY/Pegasus/ZWO pattern on the wheels. This would let people bolt the cage to the wheel instead of using threads. The small problem with this, is that a QHY version would need a small lip on it to seat into the wheel. The Pegasus wheel does not need this lip. I also think, if I recall correctly, the ZWO wheel has a different hole pattern, but no lip is needed.

Need to express this point though -- the Photon Cage is excellent. Simple, yet elegant design.

Edited ...

0.00 Ron Kramer #... · view views
Jure: Thanks for the report Bill! Fantastic! My Photon Cage is in transit for a while now (Europe) and I can't wait to finally get rid of tilt on my ASI2600MM. Question: do you use a filter wheel in your imaging train? As far as I could see the Cage is constructed in such a way that EFW does not block access. I see you're a OZONE flyer. I jumped in late in this thread - is this CAGE a new octipi type adjuster? link to it? photo of it? Edited ... Like

0.00

#...

Jure:
Thanks for the report Bill! Fantastic! My Photon Cage is in transit for a while now (Europe) and I can't wait to finally get rid of tilt on my ASI2600MM. Question: do you use a filter wheel in your imaging train? As far as I could see the Cage is constructed in such a way that EFW does not block access.

I see you're a OZONE flyer. I jumped in late in this thread - is this CAGE a new octipi type adjuster? link to it? photo of it?

Edited ...

0.00 Ron Kramer #... · view views
Steve, OK. So this is a bit different than when I used ASTAP. It looks like Han is providing more data than previously. In the past ASTAP relied on a 5 zone analysis. (4 corners and the center). This would generate a table that had designations such as TL, TR, BL, BR (top left, etc...) and would generate a quadrilateral model such as I shared in the first post of this thread. Han has since created an Octolateral (is that a word) model which instead of having the sensor divided into 5 zones, now divides the sensor into 9 zones. So instead of the quadrilateral shape like I shared earlier, you get something like this: Each of these HFD numbers corresponds to the 9 zones that are being analyzed. If you refer to the ASTAP documentation there is a diagram of the zones being analyzed. It can be found at this address: https://www.hnsky.org/astap.htm Here is the diagram from that page showing the sensor zones: Finally, if you look at the screen shot you shared, in the column header is a small grid that corresponds to the numbers you were referencing. This grid has a little highlighted part so you know which zone the HFD value corresponds with. I've circled this in the image below for your reference: Hopefully this clears up what all this data means, and you can choose which columns you are interested in graphically comparing. The outer FRAME is often line is bent? Is this a move sensor in/out bend? Your image (ASTAP)) looks like it has a bit of a bowl downward (backward) in the center? For me tilt is pretty easy but getting the spacing is the really hard part. Like

0.00

#...

Steve,

OK. So this is a bit different than when I used ASTAP. It looks like Han is providing more data than previously. In the past ASTAP relied on a 5 zone analysis. (4 corners and the center). This would generate a table that had designations such as TL, TR, BL, BR (top left, etc...) and would generate a quadrilateral model such as I shared in the first post of this thread.

Han has since created an Octolateral (is that a word) model which instead of having the sensor divided into 5 zones, now divides the sensor into 9 zones.

So instead of the quadrilateral shape like I shared earlier, you get something like this:

Each of these HFD numbers corresponds to the 9 zones that are being analyzed. If you refer to the ASTAP documentation there is a diagram of the zones being analyzed. It can be found at this address: https://www.hnsky.org/astap.htm

Here is the diagram from that page showing the sensor zones:

Finally, if you look at the screen shot you shared, in the column header is a small grid that corresponds to the numbers you were referencing. This grid has a little highlighted part so you know which zone the HFD value corresponds with. I've circled this in the image below for your reference:

Hopefully this clears up what all this data means, and you can choose which columns you are interested in graphically comparing.

The outer FRAME is often line is bent? Is this a move sensor in/out bend?
Your image (ASTAP)) looks like it has a bit of a bowl downward (backward) in the center? For me tilt is pretty easy but getting the spacing is the really hard part.

20.43 Chris White- Overcast Observatory #... · view views
Ron Kramer: The outer FRAME is often line is bent? Is this a move sensor in/out bend? Your image (ASTAP)) looks like it has a bit of a bowl downward (backward) in the center? For me tilt is pretty easy but getting the spacing is the really hard part. Yes, the outer line is often bent like that because the points in between the corners along the edge of the censor are closer to the center than the corner points, and thus have a lower HFD than the corners. Like

20.43

#...

Ron Kramer:
The outer FRAME is often line is bent? Is this a move sensor in/out bend?
Your image (ASTAP)) looks like it has a bit of a bowl downward (backward) in the center? For me tilt is pretty easy but getting the spacing is the really hard part.

Yes, the outer line is often bent like that because the points in between the corners along the edge of the censor are closer to the center than the corner points, and thus have a lower HFD than the corners.

11.02 Bill Long - Dark Matters Astrophotography #... · view views
Ron Kramer: Jure: Thanks for the report Bill! Fantastic! My Photon Cage is in transit for a while now (Europe) and I can't wait to finally get rid of tilt on my ASI2600MM. Question: do you use a filter wheel in your imaging train? As far as I could see the Cage is constructed in such a way that EFW does not block access. I see you're a OZONE flyer. I jumped in late in this thread - is this CAGE a new octipi type adjuster? link to it? photo of it? ASG Astronomy Photon Cage - ASG Astronomy Like

11.02

ASG Astronomy Photon Cage - ASG Astronomy

#...

Ron Kramer:
Jure:
Thanks for the report Bill! Fantastic! My Photon Cage is in transit for a while now (Europe) and I can't wait to finally get rid of tilt on my ASI2600MM. Question: do you use a filter wheel in your imaging train? As far as I could see the Cage is constructed in such a way that EFW does not block access.

I see you're a OZONE flyer. I jumped in late in this thread - is this CAGE a new octipi type adjuster? link to it? photo of it?

0.00 DaveDE #... · view views
Bill Long, what telescope are you using the photon cage with? Thanks - Dave Edited ... Like

11.02 Bill Long - Dark Matters Astrophotography #... · view views
Bill Long, what telescope are you using the photon cage with? Thanks - Dave AP 130 GTX Like

22.64 John Hayes #... · view views · 3 likes
Paul Abbott: Chris, Thanks for the CN PM re: my astigmatism at the corners. From a hardware standpoint, I am going back to my FF from my FRC and eliminating dovetails from my optical train. I plan to keep working on tilt and back spacing with AI in Nina. I am sorry that Optics may not allow me to completely use my entire sensor since I will probably choose to crop out the peripheral bad stars. Maybe the FF will perform better that the FRC. The octopi still perplexes me: 1. Last night I adjusted top left and bottom right, ran AI and there was minimal change of the at those corners but the top right and bottom left changed dramatically! 2. I had a right-left tilt and adjusted the right side, ran AI and it showed change in the right bottom but no change in the right top. 3. On several occasions, I have opened the octopi to find one of the push screws (larger ones) with not even close to having contact with the plate. I assume that adjusting this screw would have no effect with small adjustments. 4. I am not clear how/whether the screw (corner) position designations in NINA depend on what side of the meridian I am on and how/whether rotation affects what corner NINA reports. 5. When the Tilt measurement in AI suggests 50 micron adjustment (slightly less than a half turn) and I do it, auto-focus and run AI again I get minimal change. 6. When I see a positive value for a Tilt Corner Measurement, NINA tells me to move the sensor away from the objective. This would mean turning the larger screw counterclockwise. Likewise a negative value would require turning the larger screw clockwise. Seems like I have to do the opposite to see improvement in the right direction. 7. I had a -49 micron backfocus error so I needed move the sensor away from the flattener. I put a 1/3 mm spacer in and the back focus error became -242 microns. I would have thought that the back focus error should have been positive. I am sure that I am doing something(s) wrong. Chris thanks for all your efforts in this area. Paul Paul, The Octopi works well to adjust tilt but it is seriously over-constrained and I have suggested a couple of design improvements to Keith. The lack of clear instructions makes it hard to use if you don't clearly understand the way that it provides orthogonal tilt adjustments. There are two ways to lock any one of the adjustment screws. The first is by locking the two push/pull screws for each axis against each other. The second way is to position the inner plate using the three screws opposite to the locked screw. This has the effect of applying a bending moment on the inner plate--and you don't want to do that. The proper way to adjust the system is to place two opposite screws in contact with the inner clamp plate and to adjust tilt across those two screws using the other two screws to push/pull the plate--teeter-totter style about that pivot line. If the unit had a spring-load in one direction, this process would be MUCH easier. The second set of screws would then be strictly for locking the clamp plate in position. Since the lock screws are not on a radial line defined by the adjustment screws, locking the plate also bends it, which is a problem. As I've said, the Octopi works, but a little pseudo-kinematic design would go a long way to make it more effective, more stable, and easier to use. As it is, the design is a bit "brute-force." It's a good first effort but hopefully Keith will eventually offer an improved design. John Like

22.64

#...

· 3 likes

Paul Abbott:
Chris,

Thanks for the CN PM re: my astigmatism at the corners. From a hardware standpoint, I am going back to my FF from my FRC and eliminating dovetails from my optical train.

I plan to keep working on tilt and back spacing with AI in Nina. I am sorry that Optics may not allow me to completely use my entire sensor since I will probably choose to crop out the peripheral bad stars. Maybe the FF will perform better that the FRC.

The octopi still perplexes me:

1. Last night I adjusted top left and bottom right, ran AI and there was minimal change of the at those corners but the top right and bottom left changed dramatically!

2. I had a right-left tilt and adjusted the right side, ran AI and it showed change in the right bottom but no change in the right top.

3. On several occasions, I have opened the octopi to find one of the push screws (larger ones) with not even close to having contact with the plate. I assume that adjusting this screw would have no effect with small adjustments.

4. I am not clear how/whether the screw (corner) position designations in NINA depend on what side of the meridian I am on and how/whether rotation affects what corner NINA reports.

5. When the Tilt measurement in AI suggests 50 micron adjustment (slightly less than a half turn) and I do it, auto-focus and run AI again I get minimal change.

6. When I see a positive value for a Tilt Corner Measurement, NINA tells me to move the sensor away from the objective. This would mean turning the larger screw counterclockwise. Likewise a negative value would require turning the larger screw clockwise. Seems like I have to do the opposite to see improvement in the right direction.

7. I had a -49 micron backfocus error so I needed move the sensor away from the flattener. I put a 1/3 mm spacer in and the back focus error became -242 microns. I would have thought that the back focus error should have been positive.

I am sure that I am doing something(s) wrong.

Chris thanks for all your efforts in this area.

Paul

Paul,
The Octopi works well to adjust tilt but it is seriously over-constrained and I have suggested a couple of design improvements to Keith. The lack of clear instructions makes it hard to use if you don't clearly understand the way that it provides orthogonal tilt adjustments. There are two ways to lock any one of the adjustment screws. The first is by locking the two push/pull screws for each axis against each other. The second way is to position the inner plate using the three screws opposite to the locked screw. This has the effect of applying a bending moment on the inner plate--and you don't want to do that. The proper way to adjust the system is to place two opposite screws in contact with the inner clamp plate and to adjust tilt across those two screws using the other two screws to push/pull the plate--teeter-totter style about that pivot line. If the unit had a spring-load in one direction, this process would be MUCH easier. The second set of screws would then be strictly for locking the clamp plate in position. Since the lock screws are not on a radial line defined by the adjustment screws, locking the plate also bends it, which is a problem. As I've said, the Octopi works, but a little pseudo-kinematic design would go a long way to make it more effective, more stable, and easier to use. As it is, the design is a bit "brute-force." It's a good first effort but hopefully Keith will eventually offer an improved design.

John

0.00 Paul Abbott #... · view views
Thanks so much for your response! Like

20.43 Chris White- Overcast Observatory #... · view views · 1 like
John Hayes: John Hayes John Hayes: Paul, The Octopi works well to adjust tilt but it is seriously over-constrained and I have suggested a couple of design improvements to Keith. The lack of clear instructions makes it hard to use if you don't clearly understand the way that it provides orthogonal tilt adjustments. There are two ways to lock any one of the adjustment screws. The first is by locking the two push/pull screws for each axis against each other. The second way is to position the inner plate using the three screws opposite to the locked screw. This has the effect of applying a bending moment on the inner plate--and you don't want to do that. The proper way to adjust the system is to place two opposite screws in contact with the inner clamp plate and to adjust tilt across those two screws using the other two screws to push/pull the plate--teeter-totter style about that pivot line. If the unit had a spring-load in one direction, this process would be MUCH easier. The second set of screws would then be strictly for locking the clamp plate in position. Since the lock screws are not on a radial line defined by the adjustment screws, locking the plate also bends it, which is a problem. As I've said, the Octopi works, but a little pseudo-kinematic design would go a long way to make it more effective, more stable, and easier to use. As it is, the design is a bit "brute-force." It's a good first effort but hopefully Keith will eventually offer an improved design. John John, I've been thinking about a way to modify my octopi so that it was spring loaded, but I think that the only way to do this would be to sacrifice some of the minimum backspacing consumption. I think it would require about 3mm of space to sneak a spring in under the plates and around the locking screws. It might also require drilling some dimples for the springs to nest into, which I'm not willing to do. I agree that it would be a nice upgrade as the push pull design is a little clunky. The bottom line though, is that it does work and it works well. Once you get it set for an imaging system it stays put and it doesnt require enless tinkering. I would like to comment about your flexing concerns. I dont think those screws are going to exert enough force to flex anything. The included wrench for adjusting is a 2mm, long shaft allen wrench with a flexy T-handle on top. As long as you stick with that wrench for tightening down any of the lock screws you arent going to push hard enough to flex the plates. I do think it would be an improvement to put the adjustment screw and the locking screw at the same hour angles at each of the 4 points. Since I only unlock two screws to make an adjustment, there is still some force on the fine thread adjustment screw and I can't move it. I need to loosen a 3rd locking screw. In any event, it's a great device for what it can accomplish, although like many things in this hobby, could see some evolution. For the record I have not suggested any changes to kieth. I may do that at some point, but it's low enough on my priority list that I havent. He does seem receptive to feedback, whether he would execute a change I dont know. Edited ... Like

20.43

#...

· 1 like

John Hayes:
John Hayes

John Hayes:
Paul,
The Octopi works well to adjust tilt but it is seriously over-constrained and I have suggested a couple of design improvements to Keith. The lack of clear instructions makes it hard to use if you don't clearly understand the way that it provides orthogonal tilt adjustments. There are two ways to lock any one of the adjustment screws. The first is by locking the two push/pull screws for each axis against each other. The second way is to position the inner plate using the three screws opposite to the locked screw. This has the effect of applying a bending moment on the inner plate--and you don't want to do that. The proper way to adjust the system is to place two opposite screws in contact with the inner clamp plate and to adjust tilt across those two screws using the other two screws to push/pull the plate--teeter-totter style about that pivot line. If the unit had a spring-load in one direction, this process would be MUCH easier. The second set of screws would then be strictly for locking the clamp plate in position. Since the lock screws are not on a radial line defined by the adjustment screws, locking the plate also bends it, which is a problem. As I've said, the Octopi works, but a little pseudo-kinematic design would go a long way to make it more effective, more stable, and easier to use. As it is, the design is a bit "brute-force." It's a good first effort but hopefully Keith will eventually offer an improved design.

John

John,

I've been thinking about a way to modify my octopi so that it was spring loaded, but I think that the only way to do this would be to sacrifice some of the minimum backspacing consumption. I think it would require about 3mm of space to sneak a spring in under the plates and around the locking screws. It might also require drilling some dimples for the springs to nest into, which I'm not willing to do. I agree that it would be a nice upgrade as the push pull design is a little clunky. The bottom line though, is that it does work and it works well. Once you get it set for an imaging system it stays put and it doesnt require enless tinkering.

I would like to comment about your flexing concerns. I dont think those screws are going to exert enough force to flex anything. The included wrench for adjusting is a 2mm, long shaft allen wrench with a flexy T-handle on top. As long as you stick with that wrench for tightening down any of the lock screws you arent going to push hard enough to flex the plates.

I do think it would be an improvement to put the adjustment screw and the locking screw at the same hour angles at each of the 4 points. Since I only unlock two screws to make an adjustment, there is still some force on the fine thread adjustment screw and I can't move it. I need to loosen a 3rd locking screw. In any event, it's a great device for what it can accomplish, although like many things in this hobby, could see some evolution. For the record I have not suggested any changes to kieth. I may do that at some point, but it's low enough on my priority list that I havent. He does seem receptive to feedback, whether he would execute a change I dont know.

Edited ...

20.43 Chris White- Overcast Observatory #... · view views
Paul Abbott: Chris, Thanks for the CN PM re: my astigmatism at the corners. From a hardware standpoint, I am going back to my FF from my FRC and eliminating dovetails from my optical train. I plan to keep working on tilt and back spacing with AI in Nina. I am sorry that Optics may not allow me to completely use my entire sensor since I will probably choose to crop out the peripheral bad stars. Maybe the FF will perform better that the FRC. The octopi still perplexes me: 1. Last night I adjusted top left and bottom right, ran AI and there was minimal change of the at those corners but the top right and bottom left changed dramatically! 2. I had a right-left tilt and adjusted the right side, ran AI and it showed change in the right bottom but no change in the right top. 3. On several occasions, I have opened the octopi to find one of the push screws (larger ones) with not even close to having contact with the plate. I assume that adjusting this screw would have no effect with small adjustments. 4. I am not clear how/whether the screw (corner) position designations in NINA depend on what side of the meridian I am on and how/whether rotation affects what corner NINA reports. 5. When the Tilt measurement in AI suggests 50 micron adjustment (slightly less than a half turn) and I do it, auto-focus and run AI again I get minimal change. 6. When I see a positive value for a Tilt Corner Measurement, NINA tells me to move the sensor away from the objective. This would mean turning the larger screw counterclockwise. Likewise a negative value would require turning the larger screw clockwise. Seems like I have to do the opposite to see improvement in the right direction. 7. I had a -49 micron backfocus error so I needed move the sensor away from the flattener. I put a 1/3 mm spacer in and the back focus error became -242 microns. I would have thought that the back focus error should have been positive. I am sure that I am doing something(s) wrong. Chris thanks for all your efforts in this area. Paul Paul, I'm sorry I missed your post here. I didnt catch it until John quoted you. I just wanted to acknowledge it at this time but I will give you a thoughtful reply when I get a chance. Like

20.43

#...

Paul Abbott:
Chris,

Thanks for the CN PM re: my astigmatism at the corners. From a hardware standpoint, I am going back to my FF from my FRC and eliminating dovetails from my optical train.

I plan to keep working on tilt and back spacing with AI in Nina. I am sorry that Optics may not allow me to completely use my entire sensor since I will probably choose to crop out the peripheral bad stars. Maybe the FF will perform better that the FRC.

The octopi still perplexes me:

1. Last night I adjusted top left and bottom right, ran AI and there was minimal change of the at those corners but the top right and bottom left changed dramatically!

2. I had a right-left tilt and adjusted the right side, ran AI and it showed change in the right bottom but no change in the right top.

3. On several occasions, I have opened the octopi to find one of the push screws (larger ones) with not even close to having contact with the plate. I assume that adjusting this screw would have no effect with small adjustments.

4. I am not clear how/whether the screw (corner) position designations in NINA depend on what side of the meridian I am on and how/whether rotation affects what corner NINA reports.

5. When the Tilt measurement in AI suggests 50 micron adjustment (slightly less than a half turn) and I do it, auto-focus and run AI again I get minimal change.

6. When I see a positive value for a Tilt Corner Measurement, NINA tells me to move the sensor away from the objective. This would mean turning the larger screw counterclockwise. Likewise a negative value would require turning the larger screw clockwise. Seems like I have to do the opposite to see improvement in the right direction.

7. I had a -49 micron backfocus error so I needed move the sensor away from the flattener. I put a 1/3 mm spacer in and the back focus error became -242 microns. I would have thought that the back focus error should have been positive.

I am sure that I am doing something(s) wrong.

Chris thanks for all your efforts in this area.

Paul

Paul, I'm sorry I missed your post here. I didnt catch it until John quoted you. I just wanted to acknowledge it at this time but I will give you a thoughtful reply when I get a chance.

20.43 Chris White- Overcast Observatory #... · view views
Paul Abbott: Chris, Thanks for the CN PM re: my astigmatism at the corners. From a hardware standpoint, I am going back to my FF from my FRC and eliminating dovetails from my optical train. I plan to keep working on tilt and back spacing with AI in Nina. I am sorry that Optics may not allow me to completely use my entire sensor since I will probably choose to crop out the peripheral bad stars. Maybe the FF will perform better that the FRC. The octopi still perplexes me: 1. Last night I adjusted top left and bottom right, ran AI and there was minimal change of the at those corners but the top right and bottom left changed dramatically! 2. I had a right-left tilt and adjusted the right side, ran AI and it showed change in the right bottom but no change in the right top. 3. On several occasions, I have opened the octopi to find one of the push screws (larger ones) with not even close to having contact with the plate. I assume that adjusting this screw would have no effect with small adjustments. I skipped the first two questions, because without having the octopi setup correctly, it's difficult to say whether you made a meaningful adjustment or not. Rather than describe them as larger screws or smaller, go with coarse thread or fine thread. The fine threads are the adjustment screws, the coarse threads are the locking screws. When you setup your octopi for the first time, I recommend that you take apart the two halves and then back out the coarse threads crews until the tips recede into the plate a little bit. Then bolt the two halves together and gently tighten clockwise the fine thread screws until they all snug up. Be gentle. Once those are snugged up, then tighten clockwise the coarse thread screws until snug. These can be a little tighter than the fine thread screws, but not gorilla tight. At this point, all 8 screws are making contact. The fine thread screws are contacting the upper place, and the coarse thread screws are contacting the lower plate. From this point here, when you make an adjustment only loosen two of the coarse thread screws at a time. You can either do diagonally opposing screws, or two adjacent screws. Just do two at a time so that the other two can hold the assembly in place so that you dont get any sloppy matings messing up your adjustment. From this zero point you start from , keep in mind that you can ONLY add spacing, i.e move the sensor away from the OTA. So it helps if your backspacing is slightly too short. From here, after you determine which fine thread screw to turn, loosen the two coarse screws I suggest above and turn the fine screw counter clockwise. This will ADD spacing at this point. Only adjust one fine screw per iteration. Then snug up the two coarse locking screws that you loosened. Rerun the analyis in NINA. 4. I am not clear how/whether the screw (corner) position designations in NINA depend on what side of the meridian I am on and how/whether rotation affects what corner NINA reports. This depends on how your camera is oriented. If your camera is oriented so that the top is the top and bottom is bottom (when standing behind the camera) then left is also left and right is also right. If your camera is upside down, then things invert... top left will be bottom right, etc... 5. When the Tilt measurement in AI suggests 50 micron adjustment (slightly less than a half turn) and I do it, auto-focus and run AI again I get minimal change. Dont use the reported data as an ABSOLUTE measurement. It is not. The distance in microns that is reported is an the delta between a specific corner's optimum focuser position and the AVERAGE optimum focuser position of all 4 corners. Consider it a relative measurement that can guide you in which direction to move the corner. With my system a 1/4 turn on the octopi fine thread screw corresponds to about 16 microns as reported by NINA. Just use these numbers as a DIRECTION guide, not an ABSOLUTE adjustment suggestion. 6. When I see a positive value for a Tilt Corner Measurement, NINA tells me to move the sensor away from the objective. This would mean turning the larger screw counterclockwise. Likewise a negative value would require turning the larger screw clockwise. Seems like I have to do the opposite to see improvement in the right direction. Correct, however if your camera is upside down (see above) then you would see this behavior. 7. I had a -49 micron backfocus error so I needed move the sensor away from the flattener. I put a 1/3 mm spacer in and the back focus error became -242 microns. I would have thought that the back focus error should have been positive. Your expectation is correct. I cant say why it would be opposite. Curious: Are you using a sensor ROI with NINA Hocus Focus? If it is set to 1.0 then you are using the entire sensor for analysis. NINA uses HFR measurements for all of this. With a perfect system HFR would be a good choice, however for a system with field curvature, it's not that robust. Eccentric and Astigmatic stars confuse HFR measurement and you will get funky results if you sensor is larger than the corrected field of your telescope. It sounds like this is the scenario that you are dealing with. Try to use an ROI for the analysis. Try 0.6 or 0.7 (guessing here) to see if that helps you get more consistent results. You should exclude astigmatic stars from the analysis if possible. Additionally, if you have astigmatic or very eccentric stars the backspacing estimate that NINA outputs will be WRONG. You will need to do some experimentation and final tweaks the old fashioned way.... make a change, focus, and visually inspect. If you turn every fine thread screw on the octopi a full rotation, you change the spacing by 125 microns (0.125mm). Hope this helps. I am sure that I am doing something(s) wrong. Chris thanks for all your efforts in this area. Paul Paul, see my replies above inBold Italics Like

20.43

#...

Paul Abbott:
Chris,

Thanks for the CN PM re: my astigmatism at the corners. From a hardware standpoint, I am going back to my FF from my FRC and eliminating dovetails from my optical train.

I plan to keep working on tilt and back spacing with AI in Nina. I am sorry that Optics may not allow me to completely use my entire sensor since I will probably choose to crop out the peripheral bad stars. Maybe the FF will perform better that the FRC.

The octopi still perplexes me:

1. Last night I adjusted top left and bottom right, ran AI and there was minimal change of the at those corners but the top right and bottom left changed dramatically!

2. I had a right-left tilt and adjusted the right side, ran AI and it showed change in the right bottom but no change in the right top.

3. On several occasions, I have opened the octopi to find one of the push screws (larger ones) with not even close to having contact with the plate. I assume that adjusting this screw would have no effect with small adjustments. I skipped the first two questions, because without having the octopi setup correctly, it's difficult to say whether you made a meaningful adjustment or not. Rather than describe them as larger screws or smaller, go with coarse thread or fine thread. The fine threads are the adjustment screws, the coarse threads are the locking screws. When you setup your octopi for the first time, I recommend that you take apart the two halves and then back out the coarse threads crews until the tips recede into the plate a little bit. Then bolt the two halves together and gently tighten clockwise the fine thread screws until they all snug up. Be gentle. Once those are snugged up, then tighten clockwise the coarse thread screws until snug. These can be a little tighter than the fine thread screws, but not gorilla tight. At this point, all 8 screws are making contact. The fine thread screws are contacting the upper place, and the coarse thread screws are contacting the lower plate. From this point here, when you make an adjustment only loosen two of the coarse thread screws at a time. You can either do diagonally opposing screws, or two adjacent screws. Just do two at a time so that the other two can hold the assembly in place so that you dont get any sloppy matings messing up your adjustment. From this zero point you start from , keep in mind that you can ONLY add spacing, i.e move the sensor away from the OTA. So it helps if your backspacing is slightly too short. From here, after you determine which fine thread screw to turn, loosen the two coarse screws I suggest above and turn the fine screw counter clockwise. This will ADD spacing at this point. Only adjust one fine screw per iteration. Then snug up the two coarse locking screws that you loosened. Rerun the analyis in NINA.

4. I am not clear how/whether the screw (corner) position designations in NINA depend on what side of the meridian I am on and how/whether rotation affects what corner NINA reports. This depends on how your camera is oriented. If your camera is oriented so that the top is the top and bottom is bottom (when standing behind the camera) then left is also left and right is also right. If your camera is upside down, then things invert... top left will be bottom right, etc...

5. When the Tilt measurement in AI suggests 50 micron adjustment (slightly less than a half turn) and I do it, auto-focus and run AI again I get minimal change. Dont use the reported data as an ABSOLUTE measurement. It is not. The distance in microns that is reported is an the delta between a specific corner's optimum focuser position and the AVERAGE optimum focuser position of all 4 corners. Consider it a relative measurement that can guide you in which direction to move the corner. With my system a 1/4 turn on the octopi fine thread screw corresponds to about 16 microns as reported by NINA. Just use these numbers as a DIRECTION guide, not an ABSOLUTE adjustment suggestion.

6. When I see a positive value for a Tilt Corner Measurement, NINA tells me to move the sensor away from the objective. This would mean turning the larger screw counterclockwise. Likewise a negative value would require turning the larger screw clockwise. Seems like I have to do the opposite to see improvement in the right direction. Correct, however if your camera is upside down (see above) then you would see this behavior.

7. I had a -49 micron backfocus error so I needed move the sensor away from the flattener. I put a 1/3 mm spacer in and the back focus error became -242 microns. I would have thought that the back focus error should have been positive. Your expectation is correct. I cant say why it would be opposite. Curious: Are you using a sensor ROI with NINA Hocus Focus? If it is set to 1.0 then you are using the entire sensor for analysis. NINA uses HFR measurements for all of this. With a perfect system HFR would be a good choice, however for a system with field curvature, it's not that robust. Eccentric and Astigmatic stars confuse HFR measurement and you will get funky results if you sensor is larger than the corrected field of your telescope. It sounds like this is the scenario that you are dealing with. Try to use an ROI for the analysis. Try 0.6 or 0.7 (guessing here) to see if that helps you get more consistent results. You should exclude astigmatic stars from the analysis if possible. Additionally, if you have astigmatic or very eccentric stars the backspacing estimate that NINA outputs will be WRONG. You will need to do some experimentation and final tweaks the old fashioned way.... make a change, focus, and visually inspect. If you turn every fine thread screw on the octopi a full rotation, you change the spacing by 125 microns (0.125mm).

Hope this helps.

I am sure that I am doing something(s) wrong.

Chris thanks for all your efforts in this area.

Paul

Paul, see my replies above inBold Italics

22.64 John Hayes #... · view views · 2 likes
John, I've been thinking about a way to modify my octopi so that it was spring loaded, but I think that the only way to do this would be to sacrifice some of the minimum backspacing consumption. I think it would require about 3mm of space to sneak a spring in under the plates and around the locking screws. It might also require drilling some dimples for the springs to nest into, which I'm not willing to do. I agree that it would be a nice upgrade as the push pull design is a little clunky. The bottom line though, is that it does work and it works well. Once you get it set for an imaging system it stays put and it doesnt require enless tinkering. I would like to comment about your flexing concerns. I dont think those screws are going to exert enough force to flex anything. The included wrench for adjusting is a 2mm, long shaft allen wrench with a flexy T-handle on top. As long as you stick with that wrench for tightening down any of the lock screws you arent going to push hard enough to flex the plates. I do think it would be an improvement to put the adjustment screw and the locking screw at the same hour angles at each of the 4 points. Since I only unlock two screws to make an adjustment, there is still some force on the fine thread adjustment screw and I can't move it. I need to loosen a 3rd locking screw. In any event, it's a great device for what it can accomplish, although like many things in this hobby, could see some evolution. For the record I have not suggested any changes to kieth. I may do that at some point, but it's low enough on my priority list that I havent. He does seem receptive to feedback, whether he would execute a change I dont know. Chris, I had a spring kit in my shop and I happened to have just the right (pretty stiff) compression springs to fit over the screws in compression. I ended up deciding to use the unit as designed but as long as you put the springs on the set of screws that have the most room, compression springs will work without having to modify anything. The better solution is to use extension springs in clearance holes through two of the plates with anchor pins at the outer surfaces of the plates. The spring suggestion is simply for improving the ease of initial alignment. I agree that once it's all locked down it is quite stable. Allow me to give you a different perspective concerning mechanical flexure. Here is a basic fact about materials: At some level, EVERYTHING is made of rubber. In most cases, mechanical flexure also makes everything act like a spring. You may not realize it but screws work as springs and that's where torque limits come from. Depending on the material and the dimensions of the part, the amount of flexure may be small and the spring constant may be very high, but it's still there. In fact, mechanical flexure and the resulting spring force are what locks opposing screws if you don't properly adjust the 4 tilt screws on the Octopus. Many tilt adjustment systems use three screws at 90-degrees to achieve orthogonal adjustment but on this sort of design with a big hole in the middle, that's hard to do. I'm in agreement with Keith's choice to use 4 screws but the screws must be properly adjusted to avoid binding against the spring force induced by mechanical flexure. Even with the small amount of flexure that can be induced in the thick plates in this system, it's possible to change alignment a noticeable (and unpredictable) amount by loosening one of the screws if the clamp plate is flexed--and that's VERY undesirable. That kind of flexure generated preload can create thermal instability as well. That's where springs would make it easier to properly adjust all the screws so that they are in equal contact with the clamp plate BEFORE the locking screws are engaged, which will minimize unwanted flexure. Let me emphasize: A core principle of opto-mechanical engineering is that EVERYTHING is made of rubber! That principle is a key driver for creating precise, mechanically and thermally stable designs. I've been having some discussions with Keith and he is indeed very receptive to suggestions. I won't discussed what he has shared with me but I'm bet that we might see a new and improved product at some point. John Like

22.64

#...

· 2 likes

John,

I've been thinking about a way to modify my octopi so that it was spring loaded, but I think that the only way to do this would be to sacrifice some of the minimum backspacing consumption. I think it would require about 3mm of space to sneak a spring in under the plates and around the locking screws. It might also require drilling some dimples for the springs to nest into, which I'm not willing to do. I agree that it would be a nice upgrade as the push pull design is a little clunky. The bottom line though, is that it does work and it works well. Once you get it set for an imaging system it stays put and it doesnt require enless tinkering.

I would like to comment about your flexing concerns. I dont think those screws are going to exert enough force to flex anything. The included wrench for adjusting is a 2mm, long shaft allen wrench with a flexy T-handle on top. As long as you stick with that wrench for tightening down any of the lock screws you arent going to push hard enough to flex the plates.

I do think it would be an improvement to put the adjustment screw and the locking screw at the same hour angles at each of the 4 points. Since I only unlock two screws to make an adjustment, there is still some force on the fine thread adjustment screw and I can't move it. I need to loosen a 3rd locking screw. In any event, it's a great device for what it can accomplish, although like many things in this hobby, could see some evolution. For the record I have not suggested any changes to kieth. I may do that at some point, but it's low enough on my priority list that I havent. He does seem receptive to feedback, whether he would execute a change I dont know.

Chris,
I had a spring kit in my shop and I happened to have just the right (pretty stiff) compression springs to fit over the screws in compression. I ended up deciding to use the unit as designed but as long as you put the springs on the set of screws that have the most room, compression springs will work without having to modify anything. The better solution is to use extension springs in clearance holes through two of the plates with anchor pins at the outer surfaces of the plates. The spring suggestion is simply for improving the ease of initial alignment. I agree that once it's all locked down it is quite stable.

Allow me to give you a different perspective concerning mechanical flexure. Here is a basic fact about materials: At some level, EVERYTHING is made of rubber. In most cases, mechanical flexure also makes everything act like a spring. You may not realize it but screws work as springs and that's where torque limits come from. Depending on the material and the dimensions of the part, the amount of flexure may be small and the spring constant may be very high, but it's still there. In fact, mechanical flexure and the resulting spring force are what locks opposing screws if you don't properly adjust the 4 tilt screws on the Octopus. Many tilt adjustment systems use three screws at 90-degrees to achieve orthogonal adjustment but on this sort of design with a big hole in the middle, that's hard to do. I'm in agreement with Keith's choice to use 4 screws but the screws must be properly adjusted to avoid binding against the spring force induced by mechanical flexure. Even with the small amount of flexure that can be induced in the thick plates in this system, it's possible to change alignment a noticeable (and unpredictable) amount by loosening one of the screws if the clamp plate is flexed--and that's VERY undesirable. That kind of flexure generated preload can create thermal instability as well. That's where springs would make it easier to properly adjust all the screws so that they are in equal contact with the clamp plate BEFORE the locking screws are engaged, which will minimize unwanted flexure. Let me emphasize: A core principle of opto-mechanical engineering is that EVERYTHING is made of rubber! That principle is a key driver for creating precise, mechanically and thermally stable designs.

I've been having some discussions with Keith and he is indeed very receptive to suggestions. I won't discussed what he has shared with me but I'm bet that we might see a new and improved product at some point.

John

11.02 Bill Long - Dark Matters Astrophotography #... · view views · 1 like
@John Hayes - this is awesome and it's great to see folks taking in your incredible knowledge to improve Astro products, and of course it's great you offer this to them out of your love for the hobby. I'm sure commissioning you to design something from the ground up wouldn't be cheap. Did those Pegasus masks work out? Edited ... Like

22.64 John Hayes #... · view views
@John Hayes - this is awesome and it's great to see folks taking in your incredible knowledge to improve Astro products, and of course it's great you offer this to them out of your love for the hobby. I'm sure commissioning you to design something from the ground up wouldn't be cheap. Did those Pegasus masks work out? The masks only "sort of" work. I think that the overall dimension are a bit too big. First, I had to file them down to get them to fit side-by-side on the carousel. Second they are too deep so that the filter rattle around. That may allow the filters to rotate when the filter wheel runs, which is obviously not good. I've been so busy with other things that I haven't had time to screw with them. Since I moved to a MacBook Pro M2, I've also lost some of my Windoz based 3D design tools so I have to pull out my old Mac to deal with re-designing them. On top of that, the Pegasus FW doesn't seamlessly connect to my PC. I have to manually disconnect and reconnect the USB cable to get it to work and that's obviously no good for a remote system. The entire FW on this new scope has been a total PIA! My former FL FWs were trouble free and that's what I expected from Pegasus but no joy. I wish that Astro gear JUST WORKED...right out of the box. I'm about to write a scathing letter to the folks at Pegasus. John Like

22.64

#...

@John Hayes - this is awesome and it's great to see folks taking in your incredible knowledge to improve Astro products, and of course it's great you offer this to them out of your love for the hobby.

I'm sure commissioning you to design something from the ground up wouldn't be cheap.

Did those Pegasus masks work out?

The masks only "sort of" work. I think that the overall dimension are a bit too big. First, I had to file them down to get them to fit side-by-side on the carousel. Second they are too deep so that the filter rattle around. That may allow the filters to rotate when the filter wheel runs, which is obviously not good. I've been so busy with other things that I haven't had time to screw with them. Since I moved to a MacBook Pro M2, I've also lost some of my Windoz based 3D design tools so I have to pull out my old Mac to deal with re-designing them. On top of that, the Pegasus FW doesn't seamlessly connect to my PC. I have to manually disconnect and reconnect the USB cable to get it to work and that's obviously no good for a remote system. The entire FW on this new scope has been a total PIA! My former FL FWs were trouble free and that's what I expected from Pegasus but no joy. I wish that Astro gear JUST WORKED...right out of the box. I'm about to write a scathing letter to the folks at Pegasus.

John

11.02 Bill Long - Dark Matters Astrophotography #... · view views
John Hayes: @John Hayes - this is awesome and it's great to see folks taking in your incredible knowledge to improve Astro products, and of course it's great you offer this to them out of your love for the hobby. I'm sure commissioning you to design something from the ground up wouldn't be cheap. Did those Pegasus masks work out? The masks only "sort of" work. I think that the overall dimension are a bit too big. First, I had to file them down to get them to fit side-by-side on the carousel. Second they are too deep so that the filter rattle around. That may allow the filters to rotate when the filter wheel runs, which is obviously not good. I've been so busy with other things that I haven't had time to screw with them. Since I moved to a MacBook Pro M2, I've also lost some of my Windoz based 3D design tools so I have to pull out my old Mac to deal with re-designing them. On top of that, the Pegasus FW doesn't seamlessly connect to my PC. I have to manually disconnect and reconnect the USB cable to get it to work and that's obviously no good for a remote system. The entire FW on this new scope has been a total PIA! My former FL FWs were trouble free and that's what I expected from Pegasus but no joy. I wish that Astro gear JUST WORKED...right out of the box. I'm about to write a scathing letter to the folks at Pegasus. John Very strange, I did not have that experience with any of the filter masks I used. I am curious to know the print settings used to print them. While I know you have access to likely very nice 3D printers, I am willing to print you a set with the settings I used and send them to you. Let me know, and I will box some up and send them to you. Should only take a day or so to get from Snohomish to Oregon. Hey now on the "Windoz" stuff friend, I earn my keep there. I have also not seen this issue with the USB connection on the Indigo filter wheel, the problem I had with it, was that it was not centering filters correctly, even though it has an encoder, and it seems mine was lacking some anodizing in one of the holes. Pretty massive QC failure, easy to fix -- but man why is that allowed to happen? I know many Greeks that would be completely upset if they knew "Made in Greece" was used on products anything less than perfect. Like

11.02

#...

John Hayes:
@John Hayes - this is awesome and it's great to see folks taking in your incredible knowledge to improve Astro products, and of course it's great you offer this to them out of your love for the hobby.

I'm sure commissioning you to design something from the ground up wouldn't be cheap.

Did those Pegasus masks work out?

The masks only "sort of" work. I think that the overall dimension are a bit too big. First, I had to file them down to get them to fit side-by-side on the carousel. Second they are too deep so that the filter rattle around. That may allow the filters to rotate when the filter wheel runs, which is obviously not good. I've been so busy with other things that I haven't had time to screw with them. Since I moved to a MacBook Pro M2, I've also lost some of my Windoz based 3D design tools so I have to pull out my old Mac to deal with re-designing them. On top of that, the Pegasus FW doesn't seamlessly connect to my PC. I have to manually disconnect and reconnect the USB cable to get it to work and that's obviously no good for a remote system. The entire FW on this new scope has been a total PIA! My former FL FWs were trouble free and that's what I expected from Pegasus but no joy. I wish that Astro gear JUST WORKED...right out of the box. I'm about to write a scathing letter to the folks at Pegasus.

John

Very strange, I did not have that experience with any of the filter masks I used. I am curious to know the print settings used to print them. While I know you have access to likely very nice 3D printers, I am willing to print you a set with the settings I used and send them to you. Let me know, and I will box some up and send them to you. Should only take a day or so to get from Snohomish to Oregon.

Hey now on the "Windoz" stuff friend, I earn my keep there.

I have also not seen this issue with the USB connection on the Indigo filter wheel, the problem I had with it, was that it was not centering filters correctly, even though it has an encoder, and it seems mine was lacking some anodizing in one of the holes. Pretty massive QC failure, easy to fix -- but man why is that allowed to happen? I know many Greeks that would be completely upset if they knew "Made in Greece" was used on products anything less than perfect.

0.00 Paul Abbott #... · view views
Thanks Chris and John. This is all very helpful. I will try the techniques suggested. Paul Like

Gary Imm #... · view views
Excellent work here, Chris, thank you! Like

20.43 Chris White- Overcast Observatory #... · view views
John Hayes: John, I've been thinking about a way to modify my octopi so that it was spring loaded, but I think that the only way to do this would be to sacrifice some of the minimum backspacing consumption. I think it would require about 3mm of space to sneak a spring in under the plates and around the locking screws. It might also require drilling some dimples for the springs to nest into, which I'm not willing to do. I agree that it would be a nice upgrade as the push pull design is a little clunky. The bottom line though, is that it does work and it works well. Once you get it set for an imaging system it stays put and it doesnt require enless tinkering. I would like to comment about your flexing concerns. I dont think those screws are going to exert enough force to flex anything. The included wrench for adjusting is a 2mm, long shaft allen wrench with a flexy T-handle on top. As long as you stick with that wrench for tightening down any of the lock screws you arent going to push hard enough to flex the plates. I do think it would be an improvement to put the adjustment screw and the locking screw at the same hour angles at each of the 4 points. Since I only unlock two screws to make an adjustment, there is still some force on the fine thread adjustment screw and I can't move it. I need to loosen a 3rd locking screw. In any event, it's a great device for what it can accomplish, although like many things in this hobby, could see some evolution. For the record I have not suggested any changes to kieth. I may do that at some point, but it's low enough on my priority list that I havent. He does seem receptive to feedback, whether he would execute a change I dont know. Chris, I had a spring kit in my shop and I happened to have just the right (pretty stiff) compression springs to fit over the screws in compression. I ended up deciding to use the unit as designed but as long as you put the springs on the set of screws that have the most room, compression springs will work without having to modify anything. The better solution is to use extension springs in clearance holes through two of the plates with anchor pins at the outer surfaces of the plates. The spring suggestion is simply for improving the ease of initial alignment. I agree that once it's all locked down it is quite stable. Allow me to give you a different perspective concerning mechanical flexure. Here is a basic fact about materials: At some level, EVERYTHING is made of rubber. In most cases, mechanical flexure also makes everything act like a spring. You may not realize it but screws work as springs and that's where torque limits come from. Depending on the material and the dimensions of the part, the amount of flexure may be small and the spring constant may be very high, but it's still there. In fact, mechanical flexure and the resulting spring force are what locks opposing screws if you don't properly adjust the 4 tilt screws on the Octopus. Many tilt adjustment systems use three screws at 90-degrees to achieve orthogonal adjustment but on this sort of design with a big hole in the middle, that's hard to do. I'm in agreement with Keith's choice to use 4 screws but the screws must be properly adjusted to avoid binding against the spring force induced by mechanical flexure. Even with the small amount of flexure that can be induced in the thick plates in this system, it's possible to change alignment a noticeable (and unpredictable) amount by loosening one of the screws if the clamp plate is flexed--and that's VERY undesirable. That kind of flexure generated preload can create thermal instability as well. That's where springs would make it easier to properly adjust all the screws so that they are in equal contact with the clamp plate BEFORE the locking screws are engaged, which will minimize unwanted flexure. Let me emphasize: A core principle of opto-mechanical engineering is that EVERYTHING is made of rubber! That principle is a key driver for creating precise, mechanically and thermally stable designs. I've been having some discussions with Keith and he is indeed very receptive to suggestions. I won't discussed what he has shared with me but I'm bet that we might see a new and improved product at some point. John John, If everything is made of rubber, then the goal is to reduce the real world impact if this detriment. I.e.... get it to a point where it doesnt matter. This could be achieved from a design or use approach. My point here is, while there may be some flex with the push pull design it's not significant enough to matter. There are other considerations that have a far greater impact... the flex of the upper plate in the Octopi body is not one of them. If you are getting problems with this plate flexing then you are tightening your screws too much. I'm all for updating designs to make them more user friendly, and part of that is to reduce the chance a user might have an issue with routine use, so please dont think I'm arguing against progress. I'm not. Keith sent me a message last night, and I think it's very cool that he is working to improve the product. Can I ask, how you are determining "predictable" with your adjustments? Is this with visual inspection? Which I do think is the gold standard. The HFR/HFD analysis solutions are not very robust at the level we are working at once you get to a certain quality of correction. They are excellent at getting you to like 95% but that extra 5% must be done the old fashioned way. I'm asking because I have two Octopi devices and have been using it extensively since March. I have made hundreds of adjustments in all of my testing and I have found them to be extremely predictable. Edited ... Like

20.43

#...

John Hayes:
John,

I've been thinking about a way to modify my octopi so that it was spring loaded, but I think that the only way to do this would be to sacrifice some of the minimum backspacing consumption. I think it would require about 3mm of space to sneak a spring in under the plates and around the locking screws. It might also require drilling some dimples for the springs to nest into, which I'm not willing to do. I agree that it would be a nice upgrade as the push pull design is a little clunky. The bottom line though, is that it does work and it works well. Once you get it set for an imaging system it stays put and it doesnt require enless tinkering.

I would like to comment about your flexing concerns. I dont think those screws are going to exert enough force to flex anything. The included wrench for adjusting is a 2mm, long shaft allen wrench with a flexy T-handle on top. As long as you stick with that wrench for tightening down any of the lock screws you arent going to push hard enough to flex the plates.

I do think it would be an improvement to put the adjustment screw and the locking screw at the same hour angles at each of the 4 points. Since I only unlock two screws to make an adjustment, there is still some force on the fine thread adjustment screw and I can't move it. I need to loosen a 3rd locking screw. In any event, it's a great device for what it can accomplish, although like many things in this hobby, could see some evolution. For the record I have not suggested any changes to kieth. I may do that at some point, but it's low enough on my priority list that I havent. He does seem receptive to feedback, whether he would execute a change I dont know.

Chris,
I had a spring kit in my shop and I happened to have just the right (pretty stiff) compression springs to fit over the screws in compression. I ended up deciding to use the unit as designed but as long as you put the springs on the set of screws that have the most room, compression springs will work without having to modify anything. The better solution is to use extension springs in clearance holes through two of the plates with anchor pins at the outer surfaces of the plates. The spring suggestion is simply for improving the ease of initial alignment. I agree that once it's all locked down it is quite stable.

Allow me to give you a different perspective concerning mechanical flexure. Here is a basic fact about materials: At some level, EVERYTHING is made of rubber. In most cases, mechanical flexure also makes everything act like a spring. You may not realize it but screws work as springs and that's where torque limits come from. Depending on the material and the dimensions of the part, the amount of flexure may be small and the spring constant may be very high, but it's still there. In fact, mechanical flexure and the resulting spring force are what locks opposing screws if you don't properly adjust the 4 tilt screws on the Octopus. Many tilt adjustment systems use three screws at 90-degrees to achieve orthogonal adjustment but on this sort of design with a big hole in the middle, that's hard to do. I'm in agreement with Keith's choice to use 4 screws but the screws must be properly adjusted to avoid binding against the spring force induced by mechanical flexure. Even with the small amount of flexure that can be induced in the thick plates in this system, it's possible to change alignment a noticeable (and unpredictable) amount by loosening one of the screws if the clamp plate is flexed--and that's VERY undesirable. That kind of flexure generated preload can create thermal instability as well. That's where springs would make it easier to properly adjust all the screws so that they are in equal contact with the clamp plate BEFORE the locking screws are engaged, which will minimize unwanted flexure. Let me emphasize: A core principle of opto-mechanical engineering is that EVERYTHING is made of rubber! That principle is a key driver for creating precise, mechanically and thermally stable designs.

I've been having some discussions with Keith and he is indeed very receptive to suggestions. I won't discussed what he has shared with me but I'm bet that we might see a new and improved product at some point.

John

John,

If everything is made of rubber, then the goal is to reduce the real world impact if this detriment. I.e.... get it to a point where it doesnt matter. This could be achieved from a design or use approach. My point here is, while there may be some flex with the push pull design it's not significant enough to matter. There are other considerations that have a far greater impact... the flex of the upper plate in the Octopi body is not one of them. If you are getting problems with this plate flexing then you are tightening your screws too much. I'm all for updating designs to make them more user friendly, and part of that is to reduce the chance a user might have an issue with routine use, so please dont think I'm arguing against progress. I'm not. Keith sent me a message last night, and I think it's very cool that he is working to improve the product.

Can I ask, how you are determining "predictable" with your adjustments? Is this with visual inspection? Which I do think is the gold standard. The HFR/HFD analysis solutions are not very robust at the level we are working at once you get to a certain quality of correction. They are excellent at getting you to like 95% but that extra 5% must be done the old fashioned way. I'm asking because I have two Octopi devices and have been using it extensively since March. I have made hundreds of adjustments in all of my testing and I have found them to be extremely predictable.

Edited ...

22.64 John Hayes #... · view views · 1 like
Chris, I aligned my camera using the FWHMEccentricty tool in PI to characterize the field aberrations. I found that it is A) very repeatable and B) very sensitive to misalignment. The unpredictable behavior happens when you have all the screws locked down (no…not over-tightened) with one of the four positioning screws not in good contact with the clamp plate. Three points define a plane so you have to be very careful to make sure that all four adjustment screws are in contact to avoid this problem. And since you don’t know which screw isn’t properly positioned, loosening the lock screws causes the orientation of the clamp plate to “spring” back to an unpredictable angle. The plates are pretty stiff so the error isn’t large but it’s easily measurable—and a bit frustrating when you are simply trying to make a small tweak in only one direction. One side problem that can occur with the “out of position” screw is that when you tighten the lock-screws, it can put a lot of force on the other screws as you snug the lock screws, which is bad for those super fine thread screws. At the end of the day, we are in complete agreement that “the goal is to reduce the real world impact.” This is where implementing a pseudo-kinematic design will really help. John Like

22.64

#...

· 1 like

Chris,
I aligned my camera using the FWHMEccentricty tool in PI to characterize the field aberrations. I found that it is A) very repeatable and B) very sensitive to misalignment. The unpredictable behavior happens when you have all the screws locked down (no…not over-tightened) with one of the four positioning screws not in good contact with the clamp plate. Three points define a plane so you have to be very careful to make sure that all four adjustment screws are in contact to avoid this problem. And since you don’t know which screw isn’t properly positioned, loosening the lock screws causes the orientation of the clamp plate to “spring” back to an unpredictable angle. The plates are pretty stiff so the error isn’t large but it’s easily measurable—and a bit frustrating when you are simply trying to make a small tweak in only one direction. One side problem that can occur with the “out of position” screw is that when you tighten the lock-screws, it can put a lot of force on the other screws as you snug the lock screws, which is bad for those super fine thread screws. At the end of the day, we are in complete agreement that “the goal is to reduce the real world impact.” This is where implementing a pseudo-kinematic design will really help.

John

0.00 Ron Kramer #... · view views
I've been working on this for 3-4 years with the first RASA. I talked with Baader for months to spawn their tilt ring and they sent me a free beta to report on. My feedback didn't stop them from selling it a few months later = (. I gave it away to someone in need. I found the octopi in development and friended Keith. I got the first beta again and tuned up my RASA in a hour. (Using CCDI) I reported my con's to Keith and he took all feedback and a year later released the final? version. He sent me one as a suprise for my help and it solved all my reported cons but then spawned new one. (the new threaded ring replacement part makes it impossible to measure camera distance so that getting "close" before testing is impossible. With this version I have spent about 14 hours without success. I then went Into ASTAP and H-F which looked cool. Hoc-Foc for some reason isn't working for me and the rasa. George said it's my Celestron driver (for focuser) ? I didn't get proper results with it but it LOOKS FANTASTIC if if works on your system. I mostly went back to CCDI. = ( overpriced I do not recommend it at anything over 15.00. Another IMPORTANT thing to note. DETERMINE which corner is which... I tried to put my hand in front of a corner of the rasa but could not determine which corner was which. The RASA at 2.2 does not see my hand at all. Keith suggested I put the rasa out of focus (like collimating) where we get that large DONUT. (AH - I thought - yes I can see my cable guide when I do that!) Trying this and placing a piece of paper in front of each adjustment corner was CLEARLY VISIBLE! this was a huge help. In a couple hours I finally had good results. But, wanting it better is a dangerous concept. I lost it bad a couple times and finally had decent result. NOW SCARED to try and get it better. The other trick I learned since I could not "measure a starting point" that was at least decent for spacing and tilt. Because of this CCDI and others could not detected the stars at all in the corners with the most error. The Trick I found was to set it out of focus a littel so that I could see stars as small sharp donuts. Comparing donuts in corners I could see which 2 corners were the worst and I set NINA in a exposure loop of 2 seconds as I tuned the Octopi screws in real time. I did this until I could visually see all corner donuts were visually pretty equal in size. From here the software tools could finally see all the stars in the entire field that would allow them to work their magic. NOTE to others out there. I would never do this with my refractors. Both are high quality (80mm Stellarvue / 2500.00) scope and a (152mm carbon fiber - Both scopes APO). They are both tack sharp corner to corner. But SCT's REFLECTORS and RASA! all NEED this alignment properly done. Especially the Rasa due to the F2.2 - aperature... making its depth of in focus range extremely shallow down to 1/10 of a mm. Or that of a human hair. A SCT at F10 would be much more foregiving. Anyway - Imagine the hair pulling I did before any of these tools exsisted and there was no hardware to address tuning the hardware. I tried shims and spacers and any and all tricks. We are so blessed to have these great minds for the software and hardware we have today! BEFORE using CCDI Here is a final. CCDI using the Octopi. (with connection option). ALSO AFTER - I'd like to pull out that 3% tilt - but it is like the forbidden fruit. DON'T GO THERE. = ) My goal is the range between max and min. I like it at least .5 HFR. Here I have it lower than 0.20 ASTAP - BEFORE TUNING AFTER TUNING ONE MORE THING. I mentioned having problems with HOC-FOC. It is clearly the best of them all - but why would mine not work properly? I could not use it and it often errored during the focus run. I would also often "TIME OUT" so I had to set focus time out at 1000 seconds. Since I could not use it - I thought I'd check it once I was happy with my results to see what it says. SHOCKINGLY it said my perfect tuning was grossly OUT OF WACK. See below. The sensor looks like a BALLOON. Unless I'm reading it wrong it would have me throw my near perfect tuning into oblivion. Near final. OH - ADDITIONAL NOTE. Not mentioned here is the filter you use will likely change your (talking to all in the thread) tuning. (spacing) Once I nailed it. I installed my NBZ and knew it would be off but how much? On F8 scopes not much if any. On fast scopes it's alot of error. I had to raise the sensor evenly on all 4 adjustments. It took quite a while to get the extra .3mm of spacing I needed and then to tune out the tilt again as well. The last hour I had NBZ installed which really reduces the light (narrow band only). I had to increase exposure from 7 seconds to 25 seconds to get barely adequate stars but eventually won. Which makes me wonder? Should I have tuned with the L-PRO instead? the filters being the same thickness but the L-pro lets in a LOT more light... then when tuned in, just swap the L-pro for the NBZ? (would that be wise?) Like

0.00

#...

I've been working on this for 3-4 years with the first RASA. I talked with Baader for months to spawn their tilt ring and they sent me a free beta to report on. My feedback didn't stop them from selling it a few months later = (. I gave it away to someone in need.

I found the octopi in development and friended Keith. I got the first beta again and tuned up my RASA in a hour. (Using CCDI) I reported my con's to Keith and he took all feedback and a year later released the final? version. He sent me one as a suprise for my help and it solved all my reported cons but then spawned new one. (the new threaded ring replacement part makes it impossible to measure camera distance so that getting "close" before testing is impossible. With this version I have spent about 14 hours without success. I then went Into ASTAP and H-F which looked cool.

Hoc-Foc for some reason isn't working for me and the rasa. George said it's my Celestron driver (for focuser) ? I didn't get proper results with it but it LOOKS FANTASTIC if if works on your system. I mostly went back to CCDI. = ( overpriced I do not recommend it at anything over 15.00.

Another IMPORTANT thing to note. DETERMINE which corner is which... I tried to put my hand in front of a corner of the rasa but could not determine which corner was which. The RASA at 2.2 does not see my hand at all. Keith suggested I put the rasa out of focus (like collimating) where we get that large DONUT. (AH - I thought - yes I can see my cable guide when I do that!) Trying this and placing a piece of paper in front of each adjustment corner was CLEARLY VISIBLE! this was a huge help. In a couple hours I finally had good results. But, wanting it better is a dangerous concept. I lost it bad a couple times and finally had decent result. NOW SCARED to try and get it better.

The other trick I learned since I could not "measure a starting point" that was at least decent for spacing and tilt. Because of this CCDI and others could not detected the stars at all in the corners with the most error. The Trick I found was to set it out of focus a littel so that I could see stars as small sharp donuts. Comparing donuts in corners I could see which 2 corners were the worst and I set NINA in a exposure loop of 2 seconds as I tuned the Octopi screws in real time. I did this until I could visually see all corner donuts were visually pretty equal in size. From here the software tools could finally see all the stars in the entire field that would allow them to work their magic.

NOTE to others out there. I would never do this with my refractors. Both are high quality (80mm Stellarvue / 2500.00) scope and a (152mm carbon fiber - Both scopes APO). They are both tack sharp corner to corner. But SCT's REFLECTORS and RASA! all NEED this alignment properly done. Especially the Rasa due to the F2.2 - aperature... making its depth of in focus range extremely shallow down to 1/10 of a mm. Or that of a human hair. A SCT at F10 would be much more foregiving.

Anyway - Imagine the hair pulling I did before any of these tools exsisted and there was no hardware to address tuning the hardware. I tried shims and spacers and any and all tricks. We are so blessed to have these great minds for the software and hardware we have today!

BEFORE using CCDI

Here is a final. CCDI using the Octopi. (with connection option).

ALSO AFTER - I'd like to pull out that 3% tilt - but it is like the forbidden fruit. DON'T GO THERE. = )
My goal is the range between max and min. I like it at least .5 HFR. Here I have it lower than 0.20

ASTAP - BEFORE TUNING

AFTER TUNING

ONE MORE THING. I mentioned having problems with HOC-FOC. It is clearly the best of them all - but why would mine not work properly?
I could not use it and it often errored during the focus run. I would also often "TIME OUT" so I had to set focus time out at 1000 seconds.
Since I could not use it - I thought I'd check it once I was happy with my results to see what it says.
SHOCKINGLY it said my perfect tuning was grossly OUT OF WACK. See below. The sensor looks like a BALLOON.
Unless I'm reading it wrong it would have me throw my near perfect tuning into oblivion.

Near final. OH - ADDITIONAL NOTE. Not mentioned here is the filter you use will likely change your (talking to all in the thread) tuning. (spacing)
Once I nailed it. I installed my NBZ and knew it would be off but how much? On F8 scopes not much if any. On fast scopes it's alot of error.
I had to raise the sensor evenly on all 4 adjustments. It took quite a while to get the extra .3mm of spacing I needed and then to tune out the tilt again as well. The last hour I had NBZ installed which really reduces the light (narrow band only). I had to increase exposure from 7 seconds to 25 seconds to get barely adequate stars but eventually won. Which makes me wonder? Should I have tuned with the L-PRO instead? the filters being the same thickness but the L-pro lets in a LOT more light... then when tuned in, just swap the L-pro for the NBZ? (would that be wise?)

0.00 Ron Kramer #... · view views
P.S. to my rather long post above. There has been another contender for tilting/spacing hardware. I'm torn on mentioning it because he raped and pillaged Keith about this Octopi and I actually kicked him out of several of my Facebook groups for stealing his work. However, since I've seen the final product I don't really feel it's a rip-off of Keith's design. It is quite different. I've not checked it out in-depth, but from what I see. 1. the camera is not hidden as with the octopi and it appears I could measure the 72.8 backspacing I need to start with. 2. It seems (I could be wrong) that the spacing adjustment and tilt adjustments are separate. That would be awesome because adjusting tilt with the Octopi can kill spacing and spacing adjustments can ruin your perfect (zero) tilt work. On the downside, I believe "THE CAGE" does not fix the RASA's camera connection issues as the Octopi V2 does. (what I suggested to Keith) as it's a major Celestron RASA design flaw. Still, it's out there as an option. I'd love to hear about it firsthand. Looking at it - I still think Octopi is best. I don't know the price difference. The design flaw is not rectified and this photo even shows it installed on the included RASA 42 or 48mm connection plates. BOTH of which I threw in the trash when I opened the box of the new RASA. And that issue is Edited ... Like

0.00

#...

P.S. to my rather long post above. There has been another contender for tilting/spacing hardware. I'm torn on mentioning it because he raped and pillaged Keith about this Octopi and I actually kicked him out of several of my Facebook groups for stealing his work. However, since I've seen the final product I don't really feel it's a rip-off of Keith's design. It is quite different. I've not checked it out in-depth, but from what I see.
1. the camera is not hidden as with the octopi and it appears I could measure the 72.8 backspacing I need to start with.
2. It seems (I could be wrong) that the spacing adjustment and tilt adjustments are separate. That would be awesome because adjusting tilt with the Octopi can kill spacing and spacing adjustments can ruin your perfect (zero) tilt work.

On the downside, I believe "THE CAGE" does not fix the RASA's camera connection issues as the Octopi V2 does. (what I suggested to Keith) as it's a major Celestron RASA design flaw. Still, it's out there as an option. I'd love to hear about it firsthand. Looking at it - I still think Octopi is best. I don't know the price difference.

The design flaw is not rectified and this photo even shows it installed on the included RASA 42 or 48mm connection plates. BOTH of which I threw in the trash when I opened the box of the new RASA.

And that issue is

Edited ...

0.90 Paul Wilson #... · view views
I have the RASA 8 and the ASI2600s. I have very bad tilt problems, like a lot of guys, and I have tried to track down where they are coming from. Ha, so far with no success. But my experiments have pointed away from back focus issues and camera sensor tilt. I even pointed the scope straight up to take the weight of the camera off the corrector plate. All to no avail so far. My collimation might be slightly off, the unfocused stars however are mostly round, but I am getting extreme tilt. Collimation is on the list! A credible source is talking to me about some possible mirror slipping within the RASA itself. This is not something I can readily confirm myself, I lack the training and the equipment. But just saying, if it turned out at some point (sooner I hope) that I had a scope internal problem causing optical train distortion, this isn't something I would want to fix with an octopi. This would be a trip to Torrance Calif and the Celestron repair facility. Meanwhile, experiments here at home will continue and more evaluation. Like

0.90

Paul Wilson

#...

I have the RASA 8 and the ASI2600s.

I have very bad tilt problems, like a lot of guys, and I have tried to track down where they are coming from. Ha, so far with no success.

But my experiments have pointed away from back focus issues and camera sensor tilt. I even pointed the scope straight up to take the weight of the camera off the corrector plate.
All to no avail so far. My collimation might be slightly off, the unfocused stars however are mostly round, but I am getting extreme tilt. Collimation is on the list!

A credible source is talking to me about some possible mirror slipping within the RASA itself. This is not something I can readily confirm myself, I lack the training and the equipment.
But just saying, if it turned out at some point (sooner I hope) that I had a scope internal problem causing optical train distortion, this isn't something I would want to fix with an octopi. This would be a trip to Torrance Calif and the Celestron repair facility.
Meanwhile, experiments here at home will continue and more evaluation.

0.00 Ron Kramer #... · view views
I'm not real familiar with the 8". It's a different animal from the 11. Shoot us a picture of your train from the side. VERY unlikely culmination. Lots of 11 people say that I warn them from messing with the factory setup. Spacing and tilt will effect the culmination. So do not mess with it until you have no tilt or spacing issues. My first guess is spacing. It has to be correct down to the thickness of a hair. Before the fancy tuners were available I played with spacing using an adjustable spacing ring. You can also (if you dare, I did). Just screw one of your threaded connections in and out. It's probably also tight, but then you can unscrew a connection (I mean like just loose and a 18th of a turn... those threads are coarse). Unscrewing a connection in the train a hair will add more spacing. Of course unscrewing to far will allow sag/tilt. I have a couple of adjustable spacers for my APOS. They're pretty cheap at around 20 bucks? they allow you some play in spacing adjustment. SET it up to be in the middle of the adjustable spacer. (add remove other spacers to get it in the middle). Then you can dial it fore and aft. I have a 2600 also - it's not nearly as difficult to deal with as the ASI094 (full frame) the 2600 is 1/2 the 094 frame. The larger the sensor the further is REACHES OUTWARD into and out of the light cone. So twice the size extends my error out twice as far. Keith sells a version of the Octopi for the 8". Much smaller and should be much cheaper. It might be the best way to go. What do your images look like? My initial problem with RASA V1 was... Like

0.00