Debayering a DSLR's Bayer matrix.

[Gina]

5m subs at ISO 6400 is beginning to show noise   CF showing -7C.

Anyway, I think all this proves the sensor is fine at -8.5C which is good enough, so I think I can move on to the next stage - removing the cover glass

[Gina]

I now have a rig set up for removing the cover glass.  I have arranged cooling of the sensor using the cold finger, TEC and CPU cooler with the sensor assembly with CF held onto the TEC etc. with wire.  (Elastic bands are unsuitable with a blow lamp around ).  The cooler is placed on the copper box from the 450D project to raise it off the table.  The cover glass is tipped slightly forward of vertical to help it fall away from the gold wires.

My heart will be in my mouth shortly

[russellhq]

I now have a rig set up for removing the cover glass.  I have arranged cooling of the sensor using the cold finger, TEC and CPU cooler with the sensor assembly with CF held onto the TEC etc. with wire.  (Elastic bands are unsuitable with a blow lamp around ).  The cooler is placed on the copper box from the 450D project to raise it off the table.  The cover glass is tipped slightly forward of vertical to help it fall away from the gold wires.

My heart will be in my mouth shortly

Slow and steady wins the race

Here's my video again for some encouragement.

[Gina]

Thank you Russel - that's very helpful - little bits of heating, long and slow

[Gina]

Here's a couple of photos of my rig.

Edited May 2, 2014 by Gina

[Gina]

Cover glass removal in progress   Procedure on hold ATM due to district nurse tending to Tim.  Glass has cracked but the glue has turned white so I expect it will come off.  I'll tip it up more though to avoid any pieces of glass touching the gold wires.   Not bothered about cracked glass as I already have two spare ones.  Fortunately they're all the same across the Canon models I have

[Gina]

Well the cover glass is all off now - in lots of bits   Close inspection shows that all the gold wires are intact   I have now replaced the filter frame complete with filters to protect the sensor and gold wires while I reassemble the test rig and see if the sensor still works.  The cooling caused some condensation on the sensor so the image won't be perfect but the water marks will clean off after I've applied the epoxy resin - I'm not going anywhere near those fragile gold wires!

PHEW!!!  That was hairy

Edited May 2, 2014 by Gina

[Thalestris24]

Well the cover glass is all off now - in lots of bits   Close inspection shows that all the gold wires are intact   I have now replaced the filter frame complete with filters to protect the sensor and gold wires while I reassemble the test rig and see if the sensor still works.  The cooling caused some condensation on the sensor so the image won't be perfect but the water marks will clean off after I've applied the epoxy resin - I'm not going anywhere near those fragile gold wires!

PHEW!!!  That was hairy

Rather you than me!

Good luck

Louise

[Gina]

Thank you  

The 1100D has the most difficult cover glass to remove of all the Canon DSLRs I reckon   I have worked on 300D, 350D, 450D, 1000D and 1100D, and the 1100D is the worst by far.  With some such as the 450D, the cover glass comes off without heating quite easily.

Edited May 2, 2014 by Gina

[Gina]

I have the camera back together but without the cooling as there's no point in testing with cooling unless the sensor works without it   ATM I'm having a problem with the camera.  I can't get it to work properly with APT - takes an exposure sometimes but then says "Busy" and won't do anything   I've checked the ribbon cable connections and they seem alright.  I'll do some more checking - I can't blame the sensor quite yet.

[Gina]

Tried with a known-to-work camera and getting the same problem with APT   So this seems to indicate it's not the sensor in the other camera.  Possible culprits :- APT settings (not been altered), laptop, USB cable, "the nut behind the wheel" (me).

[Gina]

EOS Utility works fine with the working camera.  But the one with the cover glass-less sensor takes an exposure, displays a black picture and says Err 70 (which means sensor fault).  So it seems I've succeeded in destroying yet another sensor   I guess one of those dreaded gold wires must have suffered even though they look fine.  I don't think the sensor got overheated with all that cooling and ice on the cold finger.

Edited May 2, 2014 by Gina

[Thalestris24]

I wouldn't give up on it yet - there could be several reasons for the error. Maybe this might help:

http://www.magiclantern.fm/forum/index.php?topic=4105.0

Edit: oh ignore that, it's for Magic Lantern... Sorry.

I noticed you've posted here with error 70 before, Gina!

Louise

Edited May 2, 2014 by Thalestris24

[Gina]

Thank you Louise   Yes, you're right, that error can also be due to a bad connection or broken ribbon cable etc. so I'll carry on checking.

[Gina]

I took the cold finger out and swapped the ribbon cable for the one out of the working camera, checked all the ribon connections and it still gives "Err 70" after displaying a black image.  So I think I have to assume the sensor is faulty. 

I've had another close inspection of the sensor and gold wires and it certainly looks like all the gold wires are intact with no sign of disconnection of any of them but there are a couple of scratches on the surface of the sensor that show as little gold lines.  I wouldn't have thought that little bits of glass touching the sensor would go very deep but who knows - a candidate for sensor destruction I guess   I can't see that it was the heat that did it in view of the cooling - there was ice on the cold finger.  My only though is - could it be too much thermal gradient?

[Thalestris24]

Hi Gina

Just to be sure - have you tried resetting the camera settings?

Being cmos, the sensors are likely to be static sensitive so easily damaged that way - a possibility, maybe..

Louise

[Gina]

Thank you Louise

I've checked all the settings.  Now with the same camera and ribbon cable, just swapping sensor assemblies, the untouched one is working fine with EOS Utility but the one I removed the cover glass from is giving Err 70 and a black/blank image.  I have yet to comfirm this with the other camera.

Edit...  Same now confirmed for the other camera - untouched sensor works - glass-less one doesn't.

I conclude that I have proved beyon all reasonable doubt that the sensor sans glass is dud

That is as far as I am going today.

I now have two cameras with untouched working sensors that I have available for AP (in addition to the one I use for general photography).  If I want to keep one for OSC that leaves just one for debayering.  The "Last Chance Saloon" beckons

Edited May 3, 2014 by Gina

[Thalestris24]

It's defo a high risk endeavour which is why I paid someone else to do it for me! Whether the mono 1100d is actually worth the cost remains to be seen!

Good luck again

Louise

[russellhq]

Bad luck Gina, maybe it had something to do with the thermal gradient as you say. I did both my sensors like the one I did in the video. Both worked OK and the glass came off in one piece on both.

Also, I did a little experiment today to see if I could re-set the epoxy that was left on the sensor package back to the cover glass using heat. I put it in the oven and took it up in temperature slowly from ambient to 250C. Epoxy didn't melt and all that happened was the board started decomposing. That's that idea ruled out now!

[Gina]

Bad luck Gina, maybe it had something to do with the thermal gradient as you say. I did both my sensors like the one I did in the video. Both worked OK and the glass came off in one piece on both.

Also, I did a little experiment today to see if I could re-set the epoxy that was left on the sensor package back to the cover glass using heat. I put it in the oven and took it up in temperature slowly from ambient to 250C. Epoxy didn't melt and all that happened was the board started decomposing. That's that idea ruled out now!

Thank you Russel   

I think I'll have another go but with a lot less cooling.  I thought the thermal gradient would help with the glass removal but it didn't - I think it may have caused the glass to crack.  Last time I tried this I just had the cold finger attached to a large heat sink - no active cooling.

Did the sensor work after you "cooked" it?

[russellhq]

Thank you Russel   

I think I'll have another go but with a lot less cooling.  I thought the thermal gradient would help with the glass removal but it didn't - I think it may have caused the glass to crack.  Last time I tried this I just had the cold finger attached to a large heat sink - no active cooling.

Did the sensor work after you "cooked" it?

I used a dead sensor for this experiment. Here's a couple pictures of the front and back, if this had been a working sensor I'm not sure if it would still have been working after the cooking:

Russell

[Gina]

That looks well baked - nice and brown

[russellhq]

Yip! I was hoping it would melt around 120-130C. Not a chance as it turned out

[Gina]

No, I think the heat just destroys the epoxy.

[Gina]

Another day - another attack on a sensor

I decided NOT to use the cold finger but to cool the frame and metal shield - only slightly.  I used Russel's method and almost immediately I got white showing at the bottom.  I continued carefully flicking the flame over the glass and the whiteness spread all across the bottom and up the sides but the top just would not turn white.  After a couple of minutes of careful heating there was a "ping" and two thirds of the glass from the bottom flew off.  I stopped at that stage and tried to prise off the rest of the glass without success. 

So being aware that the part of the sensor and gold wires were now exposed, I put the filter frame back and put the sensor back in the camera for testing.  It still works