Chip tilt test jig.

[ollypenrice]

Here's my take on this helpful video - 

- seen recently in a link on here. In our case the scope is a RASA 8 which has the camera sitting on a flat circular plate, so a simple re-orientation was in order to do away with the equivalent of a draw tube in the video...

The laser holder is just a bit of alloy angle with one side cut off to allow it to bend at the bottom. Cable ties hold the laser on and a loose cable tie slides over the button to keep it lit. You can just bend the bracket till it moints to mid-chip.

 

Three slim wooden plates 120 degrees apart stop the camera moving when rotated but for best results I pushed the camera holder away from one of these three and towards the other two when rotating it. A strip of masking tape round the camera let me number the tilt adjustment screws so I could keep a record of each tweak I made.

It worked like a charm, assuming I picked the right laser reflection to track while turning. Initially it described a small circle but just the second tweak had it rotating on the spot.

Will this improve our corners? We hope so!  The whole thing was rustled up out of offcuts, the only expense being the 44mm hole-maker to let the laser see the chip.

Olly

 

[wimvb]

On 31/08/2021 at 15:45, ollypenrice said:

Will this improve our corners? We hope so!

How was the proverbial pudding (wherein the proof lies)?

[tooth_dr]

On 31/08/2021 at 14:45, ollypenrice said:

Here's my take on this helpful video - 

- seen recently in a link on here. In our case the scope is a RASA 8 which has the camera sitting on a flat circular plate, so a simple re-orientation was in order to do away with the equivalent of a draw tube in the video...

The laser holder is just a bit of alloy angle with one side cut off to allow it to bend at the bottom. Cable ties hold the laser on and a loose cable tie slides over the button to keep it lit. You can just bend the bracket till it moints to mid-chip.

 

Three slim wooden plates 120 degrees apart stop the camera moving when rotated but for best results I pushed the camera holder away from one of these three and towards the other two when rotating it. A strip of masking tape round the camera let me number the tilt adjustment screws so I could keep a record of each tweak I made.

It worked like a charm, assuming I picked the right laser reflection to track while turning. Initially it described a small circle but just the second tweak had it rotating on the spot.

Will this improve our corners? We hope so!  The whole thing was rustled up out of offcuts, the only expense being the 44mm hole-maker to let the laser see the chip.

Olly

 

I really hope this works!  I've a sensor I need to check and I'm itching to knock up something like this.

[tooth_dr]

@ollypenrice did you adjust the face of the camera each time, and set it back on again?  Do you mind briefly explaining how you adjusted it if this was the case

[gorann]

Very interesting! If you get an improvement I may try the same. Just one question, is there no risk of damaging the sensor by shining a laser at it?

Also, I assume the top and bottom plate of that plywood construction must be completely parallell.

Edited September 6, 2021 by gorann

[ollypenrice]

10 hours ago, wimvb said:

How was the proverbial pudding (wherein the proof lies)?

 

10 hours ago, tooth_dr said:

@ollypenrice did you adjust the face of the camera each time, and set it back on again?  Do you mind briefly explaining how you adjusted it if this was the case

 

1 hour ago, gorann said:

Very interesting! If you get an improvement I may try the same. Just one question, is there no risk of damaging the sensor by shining a laser at it?

Also, I assume the top and bottom plate of that plywood construction must be completely parallell.

We think it worked. At least, after the adjustment we got by far the best corner and edge stars we've had since using the RASA 8. There is no doubt at all that my reflection went from describing a circle to turning on the spot. Was it the right reflection???  Who knows? 

The camera has three pairs of antagonistic (push-pull) screws 120 degrees apart. I simply fitted the RASA camera holder, sat the camera on the top of the box above the hole and rotated it. To adjust it I lifted it off, removed the RASA plate and tweaked the screws by the smallest imaginable amount. (Note, I had three little wooden blocks glued to the top of the box to oblige the camera to turn in a perfect circle but the Celestron camera plate wasn't truly round so it went tight-slack during a revolution. The Artsky plate proved better but still not perfectly round so I kept a little finger pressure pushing the camera against the same two of my three blocks. That worked OK.) I put some masking tape round the camera and numbered each pair of screws so I could write down each adjustment and undo it if necessary. As it happens the second adjustment nailed it. Nice surprise! 

Parallel surfaces? To make the two sides of the box I screwed two pieces of ply together face to face and cut them in my large chop saw as if they were one. This means they are going to be close to identical in length. However, my thinking is that it doesn't matter as much as you might think. If they are not parallel the mis-directed beam reflection from a tilted chip will not describe a perfect circle but an eccentric one. This doesn't matter because you're not trying to make it describe a perfect circle but to rotate on the spot. Once it's doing so the degree of eccentricity over the size of a spot reflection arising from slight out-of-parallel effects will be negligible and below the sensitivity of the test. So 'reasonably parallel' is all you need.

Chip damage? I didn't use an astro green laser pointer, just a mild lecturer's red one. The test is recommended by Starlight Xpress and is used in their camera production so it must be harmless.

Olly

Edited September 6, 2021 by ollypenrice

[gorann]

I actually worried less about your large set-up where the top and bottom must be quite parallell than that of Daniel Sundstöm in the video. He says nothing about how he made sure it was all parallell. Any tilt in the screen where the laser spots are seen (or in the holder of the camera) will cause a circular movement and be interpreted as a tilt in the sensor even if there is none.

Edited September 6, 2021 by gorann

[wimvb]

@gorann, actually it is the reflecting surface (sensor) that determines the location and movement of the  spot. It's similar to a laser collimator test rig. If during a full rotation of the sensor, the spot does not change, the sensor is perpendicular to the rotation axis. You only need to make sure that the rotation axis is perpendicular to the front plate of the camera.

[gorann]

1 hour ago, wimvb said:

@gorann, actually it is the reflecting surface (sensor) that determines the location and movement of the  spot. It's similar to a laser collimator test rig. If during a full rotation of the sensor, the spot does not change, the sensor is perpendicular to the rotation axis. You only need to make sure that the rotation axis is perpendicular to the front plate of the camera.

Yes, after thinking more about it I came to the same conclusion. I will see if I have an old cupboard I can take apart and make one of the Olly design.

[wimvb]

4 hours ago, gorann said:

Yes, after thinking more about it I came to the same conclusion. I will see if I have an old cupboard I can take apart and make one of the Olly design.

Or this way:

https://www.sxccd.com/support/maintenance-information/aligning-the-ccd-in-an-sxvr-camera/

[gorann]

On 07/09/2021 at 12:39, wimvb said:

Or this way:

https://www.sxccd.com/support/maintenance-information/aligning-the-ccd-in-an-sxvr-camera/

Yes, that is the one our Swedish friend in the video got inspired by. Will probably give Olly's version a try when the clouds are back.

[Adam J]

On 06/09/2021 at 08:10, gorann said:

Very interesting! If you get an improvement I may try the same. Just one question, is there no risk of damaging the sensor by shining a laser at it?

I have done exactly this technique myself on a modified DSLR and the answer is yes there is a danger if you use a too higher power laser, a low power 1mw red pointer is fine, shine a 500mw blue laser and at it and you are going to burn a hole in the sensor. 

Adam 

[symmetal]

I'd like to try this on my ASI6200 but the camera is fixed to the filter wheel with 4 screws and to remove the camera means removing the filter wheel front cover and one of the 2" filters. Currently I have a Gerd-Neumann M48 tilt adjuster fitted before the filter wheel which I've adjusted looking at images but I think it could be better than it is. It may be that the FLT98 alignment is slightly out in covering full frame but it would be nice to know if this is the case.

Can anyone see why the Olly's method wouldn't work if I kept the filter wheel and tilt adjuster attached while making the test. There will be two extra reflections from the filter surfaces and the laser may have to be a bit more on axis in order to reach the sensor. Also the filter wheel will make it lop-sided so will have to ensure the tilt adjuster front surface is kept in flat contact with the testing surface while rotating it. I could attach a counterbalance weight to correct the filter wheel inbalance for the test I suppose. 😀

Alan

 

[wimvb]

13 minutes ago, symmetal said:

There will be two extra reflections from the filter surfaces and the laser may have to be a bit more on axis in order to reach the sensor.

So how will you know which reflection is which?

[tooth_dr]

1 hour ago, wimvb said:

So how will you know which reflection is which?

Maybe remove one filter to do the test?

[symmetal]

1 hour ago, wimvb said:

So how will you know which reflection is which?

The brightest reflection which is close to the plane occupied by the pixels is the sensor cover plate which seems to be the one to use. The filter will be so far in front its reflections will be way off to one side I would hope. 🤔

36 minutes ago, tooth_dr said:

Maybe remove one filter to do the test?

That would certainly help but means removing the front plate of the filter wheel. Currently there is no sign of dust on the filter flats so I would prefer not to risk getting some. 😊

Alan

[wimvb]

A bit of reflection on the subject (poor pun) leads me to the conclusion that the reflection spot on the screen closest to the incident laser beam is from the surface closest to the laser. The reflection furthest from the incident laser beam is from the deepest reflecting surface (= sensor). Unless there are multiple reflections, ie the beam reflecting back and forth between two surfaces. And this is something which I hope, will never occur.

Edited September 8, 2021 by wimvb

[Adam J]

3 hours ago, wimvb said:

So how will you know which reflection is which?

The reflection from the sensor is the strongest. But you also get a grid of reflections from the micro lens array. So its more complex than that. But still the primary sensor reflection is brightest by some margin. 

Edited September 8, 2021 by Adam J

[tooth_dr]

Thanks for sharing your design @ollypenrice 👍🏻
“often copied, never equalled” 😝

[wimvb]

Where can I get shares in IKEA?

[tooth_dr]

47 minutes ago, wimvb said:

Where can I get shares in IKEA?

Just wasn’t as handy put together 😂

 

[ollypenrice]

9 hours ago, tooth_dr said:

Thanks for sharing your design @ollypenrice 👍🏻
“often copied, never equalled” 😝

Very good! How are you going to stop lateral movement of the camera as you rotate it?

Let us know how you get on untangling the reflections!

Olly

[tooth_dr]

15 minutes ago, ollypenrice said:

Very good! How are you going to stop lateral movement of the camera as you rotate it?

Let us know how you get on untangling the reflections!

Olly

Hi Olly

Going to 3D print a holder for the camera. I’ve a new printer coming tomorrow so could not do it last night. 

Edited September 14, 2021 by tooth_dr

[ollypenrice]

1 minute ago, tooth_dr said:

Hi Olly

Going to 3D print a holder for the camera. I’ve a new printer coming tomorrow so could not do it last night. 

Should be ideal.  3D printers are just science fiction where I live... 

Olly

[wimvb]

I wonder if a well collimated laser collimator would work to adjust the sensor tilt. It might need a considerable extension to work. But the idea is the same as ordinary collimation; get a reflected light beam at the correct location, and stay there when rotating the optical elements.