Debayering a DSLR's Bayer matrix.

[libraryman]

Cheers, is it free flowing as in the video?

[pixueto]

Cheers, is it free flowing as in the video?

No, it has a slightly higher viscosity. Careful when applying it. Start by adding a very small quantity as it goes on to the sensor very easily.  I made a bit of a mess and I had to clean the excess with isopropanol. I had to use sensor swabs to keep it at bay -I suppose that practice makes perfect so it is a good idea to practice with a broken sensor if you have one. The resin will remain workable for about 15 minutes I would say.

I have also used this:

http://www.ebay.co.uk/itm/261258107141?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649

Edited April 17, 2014 by pixueto

[libraryman]

Thank you, yes I have those, very good!

[mikeobrien]

A technique that I have seen for cleaning sensors on DSLRs is to apply frog tape ( Homebase decorating dept) directly on the sensor as it comes off without any residue and takes any debris with it. Maybe you could apply it at first so it acts like masking tape. Just a thought, I would try it on a dead sensor first to check it does work without risking any damage and really doesn't leave any residue.

Sent from my Nexus 7 using Tapatalk

[Gina]

A technique that I have seen for cleaning sensors on DSLRs is to apply frog tape ( Homebase decorating dept) directly on the sensor as it comes off without any residue and takes any debris with it. Maybe you could apply it at first so it acts like masking tape. Just a thought, I would try it on a dead sensor first to check it does work without risking any damage and really doesn't leave any residue.

Sent from my Nexus 7 using Tapatalk

That's good - thank you for the tip   I found it incredidly difficult to get the epoxy resin to go in the right place and it really loved to form a whisker that went onto the image area   I found warming it up a bit to reduce the viscosity helped.  That was with standard slow setting epoxy - I haven't tried Maplin's potting epoxy though did consider it.  I would have thought that it should be possible to take the individual parts out of the pack and put them in separate syringes.  Then squirt out equal amounts from each and mix it on a non metallic board before dripping onto the job with a toothpick.

Edited April 18, 2014 by Gina

[mikeobrien]

Mm I am interested in the delayering but for my qhy8l CCD camera! I have looked to see if the CCD chip is available but haven't been able to locate one. If anyone knows of a source I would be very grateful.

Sent from my Nexus 7 using Tapatalk

[mikeobrien]

Debayering (damn spulchuck!)

Sent from my Nexus 7 using Tapatalk

[russellhq]

Mm I am interested in the delayering but for my qhy8l CCD camera! I have looked to see if the CCD chip is available but haven't been able to locate one. If anyone knows of a source I would be very grateful.

Sent from my Nexus 7 using Tapatalk

I think the bayer matrix on the QHY8L can be removed with a solvent.  That would make things a lot easier!

[mikeobrien]

Wow wouldn't it just

Sent from my Nexus 7 using Tapatalk

[Gina]

Wish that worked with DSLR sensors!

[pixueto]

Hi Gina,

Are you going to use the rotary tool with the felt tips? That will produce a very nice and smooth removal of the CFA. It will only produce microscopic scratches; those won't show up in a flat even at F22 (I've just tested it). They will only be seen though the microscope.

What is left, however, is some rests of some sort of coating which make the sensor not completely uniform. I was thinking about using solvent now that the CFA has been removed to achieve this (this would produce a perfect, golden, uniform, shining sensor). I have used nail polish but it doesn't work. I am sure the dicloroethane would have removed it but I'd rather not work with carcigen substances. I'm thinking about using toluene. Any suggestions? The solvent can even be left overnight forming a pool over the sensor if the epoxy I've used is applied all aroun as it will create a nice fence around the sensor.

[pixueto]

That's good - thank you for the tip   I found it incredidly difficult to get the epoxy resin to go in the right place and it really loved to form a whisker that went onto the image area   I found warming it up a bit to reduce the viscosity helped.  That was with standard slow setting epoxy - I haven't tried Maplin's potting epoxy though did consider it.  I would have thought that it should be possible to take the individual parts out of the pack and put them in separate syringes.  Then squirt out equal amounts from each and mix it on a non metallic board before dripping onto the job with a toothpick.

No Gina, it's much easier to mix them in the box provided and put the mix in a syringe from which you can then apply the epoxy with a small needle. It makes applying the resin much, much easier.  The only problem, though is that you only need a tiny bit of epoxy and mixing the bags will give you loads. I suppose you can take a bit of the resin and hardener and mix them separately, taking into account the proportions given in the boxes. Using a toothpick would be an absolute nightmare! Anyway, if the resin goes into the sensor, do not worry; just remove it with isopropanol or eclipse fluid. Once you apply the dremel with the felt tips, the CFA will go away nicely.

Edited April 19, 2014 by pixueto

[pixueto]

Dear all,

I know there are a couple of people attempting this now so I just wanted to show you where I am now and what I think it has gone wrong.

First of all, the epoxy from maplin is a winner. It has protected the connections really well. I felt I could get close to the edges with the rotary tool and the sensor swabs.

However, I did manage to damage the sensor quite early on (see photos attached). First, a black line appeared in the photos. Interestingly enough, I had just started debayering and did not go very deep. What I did as to remove some of the CFA from the sides where the golden connectors aren't present.  Then I continued with the debayering process and removed all CFA. The next step was to use the rotary tool at speed 3 instead of 1 to achieve a complete uniform golden sensor and remove all the coating which was left and which was preventing the flat frames from being perfect. I ended up with a PERFECT golden and shiny sensor after cleaning it with isopropanol. However, I broke the sensor completely in this process (see photo). When I examined the sensor under the microscope, I expected to see a lot of damage. However, the sensor is very clean without any sing of damage. The pixels without the CFA are there intact. As a result, I am led to believe that the damage was caused by removing the CFA from the coloured sides in the sensor. This is something that I will test next time I have a chance of attempting this mod again (it may be a while though -holidays is nearly over). Those of you attempting this now with the dremel, I would advise you to protect the coloured fringes with the maplin epoxy; apply the rotary tool vigorously in a good area in the center until the CFA is gone and you can obtain a PERFECT flat frame in that area -the surface will appear perfect.  If you can achieve this without damaging the sensor, my hypothesis is true and the danger here are the golden connectors and the coloured CFA sides. This epoxy can be the solution to this; it took care of the golden connectors but I have been careless with the coloured sides. I will test this when I have a chance but it may be a while.

[Gina]

Thank you for all that

I'm looking into this whole business again.  Several reasons, one of which is I think my 450D debayered sensor is now dead having sustained a broken gold wire in storage   I am short of 450D parts but have plenty of 1100Ds.  Several bodies less sensors, one with cold finger and one I've recently bought as not working but I believe the problem simply to be a broken battery connector - I haven't had it apart yet.  So I'm looking again at using an 1100D for my mono DSLR experiments.  A big advantage over the 450D is the more recent image processor with a maximum ISO of 6400 rather than 1600 which is very useful for DSO finding and setting up.

We now have a seemingly useable method of removing the cover glass and you have found a satisfactory epoxy resin which makes actual debayering far less risky.  I used the polishing method before but by hand which was very tedious but now there's the epoxy solution I would be happy to use my Dremel look-alike with polishing mop.  I have a dead 1100D sensor with broken cover glass and many broken wires on which I can practice CFA removal.

Regarding killing the sensor while debayering, yes, the short edges (without wires) ARE delicate and will cause failure if touched.  This was discussed much earlier on in this very long thread.

[Gina]

I have ordered the epoxy resin   I already have several syringes of various sizes that were used for medicating the goats.  I can clean these out and use them for resin.  I got the resin directly from Maplin Electronics.  I actually got 10 times as much resin for the cost of the postage   I reckon I can find other uses for this resin besides the sensors.  Once I get the resin I can see how much space it will need and sort out some suitable size containers.  I already have some plastic containers with lids that I think might do.

[pixueto]

Thank you for all that

I'm looking into this whole business again.  Several reasons, one of which is I think my 450D debayered sensor is now dead having sustained a broken gold wire in storage   I am short of 450D parts but have plenty of 1100Ds.  Several bodies less sensors, one with cold finger and one I've recently bought as not working but I believe the problem simply to be a broken battery connector - I haven't had it apart yet.  So I'm looking again at using an 1100D for my mono DSLR experiments.  A big advantage over the 450D is the more recent image processor with a maximum ISO of 6400 rather than 1600 which is very useful for DSO finding and setting up.

We now have a seemingly useable method of removing the cover glass and you have found a satisfactory epoxy resin which makes actual debayering far less risky.  I used the polishing method before but by hand which was very tedious but now there's the epoxy solution I would be happy to use my Dremel look-alike with polishing mop.  I have a dead 1100D sensor with broken cover glass and many broken wires on which I can practice CFA removal.

Regarding killing the sensor while debayering, yes, the short edges (without wires) ARE delicate and will cause failure if touched.  This was discussed much earlier on in this very long thread.

Indeed, they seem to be very delicate. I remember reading this in this thread a while ago but I had been pretty careless with my debayered 1000D sensor then (which still works) and I decided then that the cause of all those sensor failures was the epoxy rather than tampering with the coloured sides. To be honest, I am not convinced yet that the cause of the fault in my last sensor was that. If it was indeed because of the removal of the array along those fringes, then the monochrome mod is a fact as it would be  possible to achieve a complete CFA removal, with a perfect sensor, perfect flat frames and no CFA left-overs just by using the rotary tool, felp tips and chip on board epoxy. However, if the fault was caused by the dremel going too dip on the imaging sensor, then we are back to square one. I don't find the manual method satisfactory. Those who have used it (included myself) have ended up with CFA left-overs and an uneven sensor which is reflected in flat frames. I am only interested in achieving a perfect CFA removal; that is, a nice golden sensor from left to right, top to botton. I managed to achieve that and I have a method; the trouble is that the sensor couldn't take it, ha, ha. Time will tell. I have a 350D in a box and it's a candidate to have the blowtorch treatment to remove the glass but that may take a while.

Edited April 21, 2014 by pixueto

[Gina]

I tried my Bremel blowtorch on the remains of the cover glass on the 1100D sensor and yes, they dropped off.  The sensor got rather hot but that may have been because most of the glass had already come off in bits.  I'll try a full cover glass removal tomorrow.

[Gina]

Indeed, they seem to be very delicate. I remember reading this in this thread a while ago but I had been pretty careless with my debayered 1000D sensor then (which still works) and I decided then that the cause of all those sensor failures was the epoxy rather than tampering with the coloured sides. To be honest, I am not convinced yet that the cause of the fault in my last sensor was that. If it was indeed because of the removal of the array along those fringes, then the monochrome mod is a fact as it would be  possible to achieve a complete CFA removal, with a perfect sensor, perfect flat frames and no CFA left-overs just by using the rotary tool, felp tips and chip on board epoxy. However, if the fault was caused by the dremel going too dip on the imaging sensor, then we are back to square one. I don't find the manual method satisfactory. Those who have used it (included myself) have ended up with CFA left-overs and an uneven sensor which is reflected in flat frames. I am only interested in achieving a perfect CFA removal; that is, a nice golden sensor from left to right, top to botton. I managed to achieve that and I have a method; the trouble is that the sensor couldn't take it, ha, ha. Time will tell. I have a 350D in a box and it's a candidate to have the blowtorch treatment to remove the glass but that may take a while.

Seems JTW use epoxy resin to protect the gold wires and that definitely seems to be the way to go.  But it has to be the right sort of epoxy as I found out for myself!

Do you use the felt tips dry?  I ask because I tried that with a dry tip on the dead 1100D but even after some time it just polished the surface - I think it will need the car paint polish I used with the manual method.

It's very easy to go too deep and destroy the sensor that way.  Some parts cause dead pixels, some dead rows and/or columns and some complete failure - I've seen all of these!

I've yet to try a blowtorch on a complete cover glass but I have hopes of success.

[Gina]

I need to approach this scientifically viz. arrange a test setup that I can use relatively easily.  It will need to have cooling as in one case previously I found the sensor was fine at ambient temperature but failed when cooled.  Warmed up again, it worked again.  This surprised me as I expected debayering to either work or fail.  This was one with the ordinary slow set epoxy resin so I'm hoping the potting resin won't do that (if it was the resin - no way to be sure).

I have the 1100D carcass I've used in the past for astro so I can use that.  I also have an almost complete camera where the back has been replaced with a cold finger and ambient temperature cooling.  Whereas an uncooled camera would reach 30C with long exposures, I could reduce the sensor temperature by at least 10C producing a significant noise redustion.  I could add Peltier TEC and control system to provide astro level cooling.

Misting up of the sensor won't really matter as the previous sensor failed completely at low temperature and that will be what I'll be testing for.  I need to produce a new control system as the previous one relied on a large manual control box that I had on the pier.  Most of that has been dismantled and cannibalised long since   The new control system will want to be a lot smaller and I plan to have it computer controlled.  In fact I will want to control it remotely with local netbook wi-fi linked to my main desktop indoors.  I'm planning to use the final debayered, cooled camera with built-in FW and OAG on my second pier for widefield DSO imaging.

Edited April 22, 2014 by Gina

[Gina]

I have found some more information on the PX804C potting resin.  The mix ratio is 4.8:1 by volume ie. about 5:1 as opposed to 1:1 for normal epoxy resin.  I'll be able to confirm this when I get the resin. So I guess I'll be getting around 200gm of resin and 50gm of hardener or very roughly translated into volume something like 200mL resin and 50mL hardener.

[russellhq]

I need to approach this scientifically viz. arrange a test setup that I can use relatively easily.  It will need to have cooling as in one case previously I found the sensor was fine at ambient temperature but failed when cooled.  Warmed up again, it worked again.  This surprised me as I expected debayering to either work or fail.

It could be a difference in thermal expansion of the materials. Maybe when cooling down, something's contracting at a different rate and causing something to come loose? Then when things warm up the connection becomes good again?

Also, good luck with the glass removal. I'm sure it'll work for you, I've tried it on 2 sensors and it worked flawlessly both times. Just make sure the glass falls onto something soft so you can reuse it later on.

[Gina]

It could be a difference in thermal expansion of the materials. Maybe when cooling down, something's contracting at a different rate and causing something to come loose? Then when things warm up the connection becomes good again?

Also, good luck with the glass removal. I'm sure it'll work for you, I've tried it on 2 sensors and it worked flawlessly both times. Just make sure the glass falls onto something soft so you can reuse it later on.

Yes, I wondered if the epoxy disconnected a gold wire (or two) when cold and remade the contact when warm.  OTOH it could have been something entirely different.  I need to check the sensor for cold working at every stage.

If I were to use my table vice to hold the sensor unit the glass would fall onto the carpet on the floor but I think I'll do something more sensible - hot glass falling on the carpet is not ideal!   The sensor will be fitted with a cold finger first anyway so that I can check that it works before I start so I could hold it with that.  Removing heat from the sensor is a good idea anyway - helps stop it heating up when removing the glass and failing from that.  If the cover glass cracks/breaks I have another from another sensor anyway that I can use.

Whether I reuse a cover glass depends if I can devise a way of removing moisture from the sensor either as it's replaced or afterwards.  The dew point of the air between glass and sensor needs to be below -20C (or at least -10C).  I would prefer to have the sensor sealed with dry air around it and a glass covering as that makes design of the other parts much easier. 

I have three possible ideas for dehumidifying the air around the sensor.

1. Replace the cover glass in a very dry environment (and dust free)
2. Drill tiny holes in the sensor frame to purge with dry air after replacing the cover glass
3. Use the IR filter (not the colour correction blue filter) in the filter frame and drill tiny holes in this where there's a bit more room to purge with dry air

My past experience of drilling tiny holes is not good - broken drill bits etc. so I would really prefer the first option if I can devise a way to do it

Edited April 22, 2014 by Gina

[Gina]

Tried the buffing tool on dud sensor and with a bit of the ScratchX 2.0 applied to the sensor and then buffed, the CFA came off immediately and left a pretty smooth golden result.  Without the scratch removal paste and dry the buffer tool wouldn't touch the CFA.  I think it just removed the micro lenses as the surface turned from matt to shiny but still very green

[mikeobrien]

I think if someone devises a reliable method they could make a few Bob as many people would pay to have it done but be too scared to try themselves.

Sent from my Nexus 7 using Tapatalk

[russellhq]

...

I have three possible ideas for dehumidifying the air around the sensor.

1. Replace the cover glass in a very dry environment (and dust free)
2. Drill tiny holes in the sensor frame to purge with dry air after replacing the cover glass
3. Use the IR filter (not the colour correction blue filter) in the filter frame and drill tiny holes in this where there's a bit more room to purge with dry air

My past experience of drilling tiny holes is not good - broken drill bits etc. so I would really prefer the first option if I can devise a way to do it

I've been thinking about option 1 myself. I was considering placing the sensor in a sealed metal tin (like a biscuit tin), filled with some desiccant and leave it over night to dry. Then put it in the oven to re-set the epoxy.