Chip tilt test jig.

[Laurin Dave]

Hi Adam

I'd fix the problem at source and get a new M63 to M48 adapter made up the correct length to screw direct between the FW and the flattener with a bit of slack for back focus adjustment via spacers.. 

Dave

[symmetal]

I'm not sure I follow where the bad base plate is now. 🤔 In your original photo the base plate was screwed to the rear of the focuser and was also included in the setup on the jig. In your latest image with the FF is that the same base plate now to the left of the FF, and the FF is now screwed to the focuser or is the separate bad baseplate still between the FF and the focuser?

If it's the former then in this case the bad base plate is just a M63 to 48 adapter. Could one of these take the place of the bad baseplate instead or am I missing something?

If the bad baseplate is still attached to the focuser, then yes, correcting it by the camera tilt will cause the FF to camera distance to vary over the image which is not what you want. The correction for the bad baseplate would need to be done before the FF either by the 'tilted' rings or a proper tilt adjuster fitted to the bad baseplate directly.

Alan

[tooth_dr]

Hi Alan

Hard to explain in words !

The large plate screws onto the focuser itself. It has an integrated M63 to M48 adapter which appears to be glued in situ. 
 

The inside of this plate has an m63 male thread and is designed to purposely hold the reducer so it can live inside the drawtube. The outside is M48 to attach normal camera stuff.

 

 

 

[tooth_dr]

1 hour ago, Laurin Dave said:

Hi Adam

I'd fix the problem at source and get a new M63 to M48 adapter made up the correct length to screw direct between the FW and the flattener with a bit of slack for back focus adjustment via spacers.. 

Dave

It would need to be the entire focuser end plate by the looks of things as the m48 m63 adapter is fixed to the focuser end plate. 

[Laurin Dave]

Yes, just seen that from the photo..  I'd get onto APM for a new one

[symmetal]

Ah. I get it now. So the FF is before the bad base plate in the imaging train. It's hidden inside the focuser. In that case changing the camera tilt to counteract the base plate tilt would be OK as the sensor would end up orthoganal to the FF although slightly shifted to one side of the central optical axis.

As I mentioned before, this means you have to determine the sensor cover slip reflection spot, as all the other reflection spots will end up describing circles as you rotate the camera, and only the cover slip reflection will be stationary when the camera tilt cancels out the base plate tilt. The whole image train including the base plate, needs to be on the jig of course. But not with FF attached to base plate. 😁

As Laurin Dave suggests, if you show the problem to APM, they may send you a new base plate, either free or at low cost as it's clearly a problem with the plate.

Edit. If it's the whole central insert in the base plate that's tilted then that would mean the FF attachment is also tilted, so the FF is never aligned centrally to the optical axis. 😬 This means canceling the plate tilt with the camera tilt will mean good scope focus across the image but a change of FF back focus across the image and so misshapen stars. Adjusting camera tilt to be orthogonal to the FF may give round stars but the focus will change across the image. 

You could try the tilting rings on the FF side of the bad plate to make the FF on axis, and then a matching tilting ring on the other side of the plate to cancel the plate tilt. This needs M63 tilt rings though on the FF side which probably aren't available unless specially made.

Trying to get all those lined up is not an easy task. 😲

Good luck!

Alan

Edited December 3, 2021 by symmetal

[Stuart1971]

Just built my own tilt test jig, based on Olly’s idea, 

so questions, where at the sensor to point the laser for best results…as I assume if pointed at the centre it won’t show much in the way if movement…? Or am I missing something…

and where can I get a suitable laser pointer to do the job, as my Baader 1mw collimating laser is not powerful enough and is a bit big and bulky…

And re the laser reflection dots, is it the cover slip reflection or one of the pixel reflections I am looking to work on, as I have read conflicting ideas on this…?

 

Edited October 1, 2022 by Stuart1971

[symmetal]

It doesn't make any difference where on the sensor the laser is pointed as long as you get a visible reflection off it. The sensor and cover slip glass are optically flat, one would hope, 😉 and the angle of reflection is the same from any part of the sensor.

Laser pens like this are commonly used as they are cheap and plenty available from ebay. Just check it's a 1mW one, and has Class 1 on the label. Red or green is fine though red seems most commonly used. I made a 3d printed holder to keep the pen at an 10 degree angle and it can then be easily moved around for the clearest reflection if you have a filter wheel, tilt adjuster etc. also fitted to the jig which reduces the available sensor area visible to the laser.

It's the flat cover slip reflection which you want to adjust on. It's generally the brightest reflection spot, though there may be two or three reflection spots close together. Rotating the camera and see what group of bright dots move the least points you to where to start. Moving the laser pen a little should make the pixel reflections move so they can be eliminated. If you can't tell which is the right spot you have to choose one, adjust the tilt so it doesn't move when the camera assembly is rotated, and try it on the scope for a test image. If it looks good you're OK. On my ASI6200 it took two tries and I had to choose the second spot before the tilt was gone. 🙂

Alan

Edited October 1, 2022 by symmetal

[Stuart1971]

51 minutes ago, symmetal said:

It doesn't make any difference where on the sensor the laser is pointed as long as you get a visible reflection off it. The sensor and cover slip glass are optically flat, one would hope, 😉 and the angle of reflection is the same from any part of the sensor.

Laser pens like this are commonly used as they are cheap and plenty available from ebay. Just check it's a 1mW one, and has Class 1 on the label. Red or green is fine though red seems most commonly used. I made a 3d printed holder to keep the pen at an 10 degree angle and it can then be easily moved around for the clearest reflection if you have a filter wheel, tilt adjuster etc. also fitted to the jig which reduces the available sensor area visible to the laser.

It's the flat cover slip reflection which you want to adjust on. It's generally the brightest reflection spot, though there may be two or three reflection spots close together. Rotating the camera and see what group of bright dots move the least points you to where to start. Moving the laser pen a little should make the pixel reflections move so they can be eliminated. If you can't tell which is the right spot you have to choose one, adjust the tilt so it doesn't move when the camera assembly is rotated, and try it on the scope for a test image. If it looks good you're OK. On my ASI6200 it took two tries and I had to choose the second spot before the tilt was gone. 🙂

Alan

Thanks very much, just a couple of things that I can’t get my head around, I have the QHY268c with the IMX571 sensor, surely if I aim the laser certain parts of this, when I rotate the camera, the  beam could go off the edge of the sensor and then come back on..? so does that mean ideally it needs to be aimed near the centre…? 

I am trying to visualise the sensor being tilted, say long edge to long edge, but if the beam was pointed bang in the middle, then would the dot not then move when rotated as the centre would be pretty much still….or am I missing something….??

I have  taken the laser out of a Baader collimator I had, as I don’t use it, it’s a 1mw but it’s a class 2 red beam, I assume this will be fine as otherwise Baader would not have sold it for this purpose…?

sorry to be dim…

 

[symmetal]

2 hours ago, Stuart1971 said:

Thanks very much, just a couple of things that I can’t get my head around, I have the QHY268c with the IMX571 sensor, surely if I aim the laser certain parts of this, when I rotate the camera, the  beam could go off the edge of the sensor and then come back on..? so does that mean ideally it needs to be aimed near the centre…? 

I am trying to visualise the sensor being tilted, say long edge to long edge, but if the beam was pointed bang in the middle, then would the dot not then move when rotated as the centre would be pretty much still….or am I missing something….??

I have  taken the laser out of a Baader collimator I had, as I don’t use it, it’s a 1mw but it’s a class 2 red beam, I assume this will be fine as otherwise Baader would not have sold it for this purpose…?

sorry to be dim…

 

Yes, if aimed near the sensor edge it could go off the edge, so it needs to be inside the diameter of a circle that covers the sensor, so aiming near the sensor centre will avoid this happening, my point being in response to your initial question that if it was aimed at the centre it wouldn't show the tilt error which isn't the case. 😉

If the sensor is tilted as you say, then as the camera is rotated, the centre of the sensor would stay in the same place vertically, but the angle the sensor cover slip surface plane presents to the laser won't stay the same, which is all we're concerned with, and so the reflected dot will rotate about a centre point. If you move the laser a little so it's pointing at a different point on the sensor the change in cover slip angle angle presented to the laser won't change, though the cover slip will move up and down a little as the camera is rotated so the reflected dot will describe a larger circle, but this will only be apparent at large tilt angles. As we're only talking about correcting tilt errors of a small fraction of a degree this change in reflected circle diameter is not really noticeable. Once the tilt is elimated then the cover slip presents a constant angle to the laser when the camera is rotated no matter which point on the sensor the laser is pointed at, so the reflected dot doesn't move. Hope that explains it better.🙂

Sorry, the 1mW I use is labelled Class 2 and not Class 1 as I stated earlier. Class 1 lasers are rated at below 0.39mW and Class 2 are below 1 mW so you will be fine with your Baader 1mW laser. Class 3R are rated at 1 to 5mW. ebay banned some laser pen sellers for selling 1mW Class 2 lasers which were actually Class 3 10mW lasers. Class 3B actually allows power up to 500mW. The ebay 1mW Class 1 stated lasers probably are marked Class 2 but that's OK. Laser classes. It's all a bit confusing. 

Alan

Edited October 1, 2022 by symmetal

[Stuart1971]

17 hours ago, symmetal said:

Yes, if aimed near the sensor edge it could go off the edge, so it needs to be inside the diameter of a circle that covers the sensor, so aiming near the sensor centre will avoid this happening, my point being in response to your initial question that if it was aimed at the centre it wouldn't show the tilt error which isn't the case. 😉

If the sensor is tilted as you say, then as the camera is rotated, the centre of the sensor would stay in the same place vertically, but the angle the sensor cover slip surface plane presents to the laser won't stay the same, which is all we're concerned with, and so the reflected dot will rotate about a centre point. If you move the laser a little so it's pointing at a different point on the sensor the change in cover slip angle angle presented to the laser won't change, though the cover slip will move up and down a little as the camera is rotated so the reflected dot will describe a larger circle, but this will only be apparent at large tilt angles. As we're only talking about correcting tilt errors of a small fraction of a degree this change in reflected circle diameter is not really noticeable. Once the tilt is elimated then the cover slip presents a constant angle to the laser when the camera is rotated no matter which point on the sensor the laser is pointed at, so the reflected dot doesn't move. Hope that explains it better.🙂

Sorry, the 1mW I use is labelled Class 2 and not Class 1 as I stated earlier. Class 1 lasers are rated at below 0.39mW and Class 2 are below 1 mW so you will be fine with your Baader 1mW laser. Class 3R are rated at 1 to 5mW. ebay banned some laser pen sellers for selling 1mW Class 2 lasers which were actually Class 3 10mW lasers. Class 3B actually allows power up to 500mW. The ebay 1mW Class 1 stated lasers probably are marked Class 2 but that's OK. Laser classes. It's all a bit confusing. 

Alan

Well, after a bit of messing about I think I managed to see how much tilt I have in my QHY268c camera

First of all I left the filter in and the laser produced 3 dots, as per (first picture below), and I tracked the wrong dot as then I removed the filter as I realised it was an UV/IR block filter, and then dot i was tracking disappeared, and left me two, (second image below), so had to start again, so I tracked both and plotted with dots, i only have an AR window on my camera so if you have a UV/IR window you will probably see three dots.

so then I rotated the camera and one of the two dots moved in a very small circle, the other, which I am assuming is the cover slip reflection and the one I need to concentrate on, moved in roughly a 4mm diameter circle, (see pictures and video below) not sure how much tilt that relates too, but a single piece of A4 paper under the edge was about right to move the dot to the centre of the 4mm circle, as this would mean the tilt should be eliminated, I don’t have my tilt adapter yet, so can’t sort it permanently ☹️☹️
 

Am I correct in assuming I have chosen the correct dot to concentrate on, or does it not really matter…as when tilt is corrected both dots will remain static…?

Edited October 2, 2022 by Stuart1971

[symmetal]

Glad to see you've got it working. It difficult to say which is the right dot. The other one is very likely from the AR window as you say. If you deliberately tilt the camera assembly slightly both dots will move and the dot reflected off the cover slip should move further as the total light path is longer. If you still can't tell you'll have to choose one dot and centre that and take a test image. If your image looks good with no tilt effects then you chose the right dot.

It depends on how well the camera is put together as to whether both dots are stationary when the tilt is corrected. If the AR glass is accurately parallel to the sensor then yes, they will both be stationary. I've found in some cameras that's the case, while in others it isn't. If they are both stationary then you're good to go but if not you'll have to choose one and do a test image.

When doing the tilt adjustment, for checking how much the dot moves, mark a cross on the piece of paper, and place it under the dot you've chosen and see how far it moves from the cross while you rotate the camera. If you can make a tilt adjustment while still on the rig, adjust it to bring the dot half way back to the cross, recenter the cross and repeat the process. If you have to remove the assembly to make adjustment, stick a label by each tilt adjustment so you know which one you're adjusting each time. You'll soon get the hang of which to adjust by visually assessing the position of the dot relative to the tilt adjustment positions.

Yes, the thickness of a piece of paper can be the amount of adjustment needed to fix the tilt.

Alan

[Stuart1971]

6 minutes ago, symmetal said:

Glad to see you've got it working. It difficult to say which is the right dot. The other one is very likely from the AR window as you say. If you deliberately tilt the camera assembly slightly both dots will move and the dot reflected off the cover slip should move further as the total light path is longer. If you still can't tell you'll have to choose one dot and centre that and take a test image. If your image looks good with no tilt effects then you chose the right dot.

It depends on how well the camera is put together as to whether both dots are stationary when the tilt is corrected. If the AR glass is accurately parallel to the sensor then yes, they will both be stationary. I've found in some cameras that's the case, while in others it isn't. If they are both stationary then you're good to go but if not you'll have to choose one and do a test image.

When doing the tilt adjustment, for checking how much the dot moves, mark a cross on the piece of paper, and place it under the dot you've chosen and see how far it moves from the cross while you rotate the camera. If you can make a tilt adjustment while still on the rig, adjust it to bring the dot half way back to the cross, recenter the cross and repeat the process. If you have to remove the assembly to make adjustment, stick a label by each tilt adjustment so you know which one you're adjusting each time. You'll soon get the hang of which to adjust by visually assessing the position of the dot relative to the tilt adjustment positions.

Yes, the thickness of a piece of paper can be the amount of adjustment needed to fix the tilt.

Alan

Thanks…

I just wish I could tell which dot is the correct one, as one moves in a much bigger circle than the other, does that tell me anything…as if it’s the dot that is not moving much at all, then the tilt is very minimal, they seem the same brightness too, I thought the one that did not move as much may have been one of the set of dots from the pixel array, as it seemed to move about the same amount as all the others further back, but this would mean that there was no reflection from the AR window….🤔🤔🤯

[symmetal]

The pixel array reflections, which are dimmer, don't tell much about tilt as they have curved microlenses in front the pixels. If your tilt currently shown in your images is small I would start with the dot that doesn't move much.

On some cameras the cover slip reflection is much brighter than the other dots so is easy to spot, but on other cameras like you've found, the dots have a similar brightness.

Alan

[Stuart1971]

Just now, symmetal said:

The pixel array reflections, which are dimmer, don't tell much about tilt as they have curved microlenses in front the pixels. If your tilt currently shown in your images is small I would start with the dot that doesn't move much.

On some cameras the cover slip reflection is much brighter than the other dots so is easy to spot, but on other cameras like you've found, the dots have a similar brightness.

Alan

Yes I have just come to the same conclusion, they seem pretty much the same brightness, also why are there no pixel array dots near to these two…🤔 they only seem to show further away towards where the laser is situated, wonder if I have to steep angle on the laser beam, and it should be more upright…

I think it is the other dot, as the one I thought has a slight red flare around it, which would make more sense it’s from the AR window…🤔 do you think…?

[Stuart1971]

Would you agree that the centre dot here is much more pronounced and round and more symmetrical then the other two which have a bit of flare, the top one turned out to be from the filter…so that left the other two, which I think I chose the wrong one….

I am assuming that the reflection from the slip cover, should be a more perfect dot due to it coming from the closest to the sensor itself…correct…?

Just not sure why the AR window dot should move in a bigger circle, then the cover slip dot, as it closer to the base…🤔

Edited October 2, 2022 by Stuart1971

[symmetal]

The centre dot does look to be the best candidate to go for. The AR window and filter likely shows reflections from both surfaces which blurs the result a little causing the flares.

The pixel array does seem to present unexpected results. Some cameras show a pixel dot close by the bright dots and some don't. You would think moving the laser slightly would make the pixel reflections move the same and at one point a pixel would be directly under the bright dots but that doesn't appear to happen like that in reality. 🤔

Alan

[Stuart1971]

1 minute ago, symmetal said:

The centre dot does look to be the best candidate to go for. The AR window and filter likely shows reflections from both surfaces which blurs the result a little causing the flares.

The pixel array does seem to present unexpected results. Some cameras show a pixel dot close by the bright dots and some don't. You would think moving the laser slightly would make the pixel reflections move the same and at one point a pixel would be directly under the bright dots but that doesn't appear to happen like that in reality. 🤔

Alan

In which case I have very little tilt, probably about a 2-3mm circle, would this make much of a difference to the stars…I guess it would, just wondering really how much tilt is acceptable and will not really affect the stars on one side of the image….🤔🤔

I get it’s best to have none, but is there a point where it will not manifest in the images….

[symmetal]

It depends on the size of your sensor. For an APS-C size you would likely need it better than 1mm movement I would say for that size of jig. For my full size ASI6200 I had to get it to where there was no noticeable movement to avoid elongated stars. It still showed some coma in the corners but that's not due to tilt. 🙂

Alan

[Stuart1971]

This guy says you should ignore the window reflections and use the pixel array reflection to set the tilt….🤔🤔

https://www.macobservatory.com/blog/2021/11/28/how-to-remove-sensor-tilt-from-your-astrophotography-camera-while-indoors

[GalaxyGael]

Having a good angle between the laser and the sensor, and a suitable distance will allow you to see the periodic reflections from the sensor. If you have not done so, its useful to do this in near complete dark so you can identify all the dots, and then let some light in so you know where you are!

In a good sensor the cover slip that is just over the sensor (you cannot see this easily) is perfectly parallel and since it is a slightly brighter reflection it should be fine. I use the dimmer actual sensor reflection, even on an OSC sensor. My aim was to get what records the data to have no tilt just in case the cover slip had some (but its moot really in the end). It is easy to identify as this dot will be in line and periodically spaced to the rows of other reflections from the sensor. Focus on the dot that has a close-by partner dot, which is the cover slip. The other (third) one further away is likely the AR or Ir-cut glass, which you can ignore. Depending on the angle of the laser, you sometimes might not see this reflection anyway. I did not notice any adverse effect from microlenses, bayer matrix arrays or the texture common to larger CCD pixels (CMOS are flat), so I always use the central periodic dimmer reflection of the sensor itself. Seeing the square array of dots really helps to identify it unambiguosly and that square array of dots will rotate around the central dot. Once the central dot stays put, you are good to go.

For gauging the movement, 2-3 mm is quite a large circle, but will depend on your optical system as to how much that will cause mis-shapen stars. It can be annoying to make so many small adjustments, but that sensor central dot can be made to almost rotate on point. Mine has the tiniest of movement, it used to make a huge circle since ASI did not place the sensor flat in the first place. As a reference, when ideally flat, my ASI2600Mc tilt plate has a clear gap on one side that you can look through :). yours might be much better.

Happy tweaking with a small hex key in the dark !!!

ASIDE: eventually I had to put mine on a shelf in the closet and get my knee pads on.

[ollypenrice]

F r

[ollypenrice]

F ratio also plays a big part in the precision required. A fast F ratio has a steep light cone which means a shallow depth of field. We see this in daytime photography all the time: for a deep focal range you need to stop down to a slow F stop. Even though we are shooting at infinity in AP the same rules apply. A slow system will be more tolerant of a tilted chip since critical focus only has to fall somewhere onto a single pixel. It doesn't matter where on that pixel it falls. When critical focus from one side doesn't find the same pixel as critical focus from the other you lose resolution.

Olly

[Stuart1971]

51 minutes ago, GalaxyGael said:

Having a good angle between the laser and the sensor, and a suitable distance will allow you to see the periodic reflections from the sensor. If you have not done so, its useful to do this in near complete dark so you can identify all the dots, and then let some light in so you know where you are!

In a good sensor the cover slip that is just over the sensor (you cannot see this easily) is perfectly parallel and since it is a slightly brighter reflection it should be fine. I use the dimmer actual sensor reflection, even on an OSC sensor. My aim was to get what records the data to have no tilt just in case the cover slip had some (but its moot really in the end). It is easy to identify as this dot will be in line and periodically spaced to the rows of other reflections from the sensor. Focus on the dot that has a close-by partner dot, which is the cover slip. The other (third) one further away is likely the AR or Ir-cut glass, which you can ignore. Depending on the angle of the laser, you sometimes might not see this reflection anyway. I did not notice any adverse effect from microlenses, bayer matrix arrays or the texture common to larger CCD pixels (CMOS are flat), so I always use the central periodic dimmer reflection of the sensor itself. Seeing the square array of dots really helps to identify it unambiguosly and that square array of dots will rotate around the central dot. Once the central dot stays put, you are good to go.

For gauging the movement, 2-3 mm is quite a large circle, but will depend on your optical system as to how much that will cause mis-shapen stars. It can be annoying to make so many small adjustments, but that sensor central dot can be made to almost rotate on point. Mine has the tiniest of movement, it used to make a huge circle since ASI did not place the sensor flat in the first place. As a reference, when ideally flat, my ASI2600Mc tilt plate has a clear gap on one side that you can look through :). yours might be much better.

Happy tweaking with a small hex key in the dark !!!

ASIDE: eventually I had to put mine on a shelf in the closet and get my knee pads on.

Thanks for that

My scope is f5.3 TakFSQ85, camera is QHY268c same as the ASI2600mc

well I can’t see any pixel reflections near the AR window reflection at all, (just a few around the laser base about 6” away) so I can’t use that…I have just two dots as shown above, one from the AR window and one from the cover window, and don’t know which is which as they are the same brightness, one moves in a circle about 2-3mm across and the other one 4mm across…

 

so I have not had the luck you had in sorting it out….

[symmetal]

The sensors are manufactured with high precision so the cover slip would be parallel to the sensor array, so whether you choose the cover slip or a close pixel dot, both should not move when the tilt has been corrected and the camera is rotated. To see the close pixel dot clearly you would need likely need a larger laser angle which may not be possible if you have a filter wheel and separate tilt adjuster on the jig.

If the cover slip and AR coating are parallel and you have tilt then the bright dot which describes the largest circle when rotated would be the cover slip as the light path has travelled a greater distance. As you don't know if they're parallel that doesn't really help.

Alan

Edited October 2, 2022 by symmetal