Good Old Days (Page 2): How do you go about scraping a sensor?

JimH123 Loc: Morgan Hill, CA

Oknoder wrote:

Odd thing is when I use a DSLR I now use my old D5100 instead of my D800e. I like the swivel screen and I removed the hot mirror to increase Ha sensitivity. I bought a spare sensor and have stripped off half of the Bayer layer to make it mono. I checked it last weekend to make sure it still worked, before I continue taking off of the rest. It sucks losing the microlenses but it does seem more sensitive in the scraped areas.

Matthew

How do you go about scraping a sensor?

CraigFair Loc: Santa Maria, CA.

Oknoder wrote:

I love your Lovejoy image the best, I too on occasion go back and play with my older images, mostly just to see how my processing skills have improved.

Glad to hear you're on the mend
Matthew

Thank you Matthew, Lovejoy was a really fun project.
And I'm surprised I got such a good shot at that point in my learning curve.
I'm still only 1/2 way to being adequate.
Good luck with the D5100 project, that was my first DSLR and I too move on to the D800E.
Craig

Oknoder Loc: Western North Dakota

JimH123 wrote:

How do you go about scraping a sensor?

I used a chemical, have to check the name to remove the microlenses, when I get home I'll post the name. I do know it's not something gotten from Walmart. The hardest part was removing the glass that protects the actual sensor. All the experts say you can heat the epoxy holding it on the sensor PCB assembly. I used the heat gun on my reflow/soldering station, it barely made a dent. I ended up using a grinding wheel for my stereo, and just cutting the glass, much easier this way. I was going to epoxy to protect the little super fine wires that connect the sensor's semi-conductor to the PCB, but since the glass kind of does that for me, why bother. Ultimately I used a sharpened wooden dowel to rub the actual sensor and after a bit you can see where the layer has been removed. Where the filter remains looks similar to the multi colors of an oil sheen, underneath looks like a dull gold.

I have burned through a few sensors already, mostly due to those pesky gold wires. Thankfully they are pretty cheap, $20-30 on eBay. I just keep putting it in my camera and shooting a cpl images to make sure it still works and I'm not wasting time on a sensor that I already broke. My advice do not under any circumstances use your main camera, take your time and test often, be gentle with the ribbon cables, hard to find just a single cable, find a model that doesn't cost much and search the web and see if there is a step by step teardown of your chosen model. I can now strip out my sensor in a matter of minutes, and reassembly is a snap.

HTH
Matthew

CraigFair Loc: Santa Maria, CA.

Oknoder wrote:

I used a chemical, have to check the name to remov... (show quote)

For those of us not so Micro inclined there is always the experts like.
http://www.spencerscamera.com/store/store_product_detail.cfm?Product_ID=20&Category_ID=1&Sub_Category_ID=2
That's who I had do my D600 and they did a great job.
Just an alternative. I had a friend do my Canon T2i conversion.
He got lucky and only broke the one piece that doesn't go back on.
Craig

Oknoder Loc: Western North Dakota

CraigFair wrote:

For those of us not so Micro inclined there is always the experts like.
http://www.spencerscamera.com/store/store_product_detail.cfm?Product_ID=20&Category_ID=1&Sub_Category_ID=2
That's who I had do my D600 and they did a great job.
Just an alternative. I had a friend do my Canon T2i conversion.
He got lucky and only broke the one piece that doesn't go back on.
Craig

Spencer's removes the IR/AA or what is called the hot filter stack, which I might add is not glued in on the D5100, as it is in the D600 or D800 series, and is only 0.8mm thick, so removing without replacing it with a like thickness piece of glass was not necessary to achieve focus to infinity with any of my lenses, only had to desolder the vibration dust control from the PCB. The piece of glass I was referring to was the piece that is epoxied directly over the sensor, it is optically clear and is the last protective layer for the sensor. I have a scrap sensor laying around I can try taking a picture of both the glass on the sensor and hot filter/AA filter stack. The only company I know of that offers the service I am attempting to DIY is called MaxMax, which I have no experience with, but have heard nothing but good things, aside from their prices, which tend to be high. IMHO

I wouldn't even consider doing this mod to a D600, as the sensor probably costs upwards of a couple hundred dollars, whereas the D5100/D7000 sensor can be had for $20-60, so any mistake is minor not painful, like it would be with my D800e. I thought about removing the hot filter from my D800e but its filter stack is a little over 3mm, considering it has a vibrating dust cover filter, AA filter, IR filter, and a AA canceling filter. Not sure why they did not just omit the AA filter instead of adding another filter to cancel out the first in the stack, like they did with the D810. Maybe it had something to do with the filters being glued together in a sandwiched stack. The only real problem/hazard I can see is that when the viewfinder mirror flips up, and the shutter opens the sensor is exposed to the elements if not connected to some form of lens or some other form of glass.

CraigFair Loc: Santa Maria, CA.

Oknoder wrote:

Spencer's removes the IR/AA or what is called the ... (show quote)

Thank you Matthew I did not know the Crop Frame and the Full Frame would have different construction techniques but it makes sense.
Craig

Oknoder Loc: Western North Dakota

Most of the cameras are very similar, even going back to the D70s and older, the filter stacks in the higher models seem to be different. It might be that since the filters have to be bigger for the bigger sensor in the FF DSLRs, they have to be cemented in for some reason. In all honesty I did not see much difference between the D600, which is the instructional video I used to teardown my D800e to repair it, other than control components and of course the sensor itself.

I also video tape the whole sequence, just in case I find myself lost while rebuilding it, which has happened from time to time, I hate having left over screws.
Matthew

Oknoder Loc: Western North Dakota

JimH123 wrote:

How do you go about scraping a sensor?

Sry for the late reply. I thought I sent one but apparently my phone decided you didn't need to read, or that I needed to rewrite the whole thing.

Once the sensor assembly was removed, which holds an aluminum frame, PCB with a couple of ribbon connectors and the actual sensor itself. I removed the PCB from the frame and as per the guides I found, attempted to use a heat gun, from my rework solder station, to loosen and remove the glass covering the sensor. It did not work very well and in the end, I used a diamond grinding wheel on my dremel to cut the glass. This had to be done with extreme care, there are dozens of fine hairlike wire connectors, which are used to read the data coming off each photoreceptors. After I was able to cut away all the glass, without damaging any of the wires, I wont say how many sensors I destroyed but I wasn't successful with the first couple, LOL.

With the glass removed I tried a couple of different solvents to help remove the microlenses and bayer filter. The one I found that worked the best was Methylene Chloride or by common name Dichloromethane. This is a very caustic solvent so extreme caution should be used. I used a small amount and filled a pyrex container, to the point that the sensor would be covered by the solution. I then placed the pyrex petri dish in an air tight bowl, as it evaporates very quickly and gives off noxious fumes. It must be added that I conducted this experiment in the garage, and when the sensor was in the petri dish I left it to soak for 36hrs, within a fire cabinet I have in my garage.

After pulling the dish out it was apparent that my air tight bowl was not really air tight as the majority of the solution had evaporated, the microlenses were removed easily and looked like a coating of snot covering the sensor. Using RO/DI water and Qtips to wipe the sensor clean of the microlenses. The actual bayer filter is printed directly on top of the semiconductor used as the sensor's photo-receptor. Before attempting anything else I had to add a couple of dabs of epoxy, due to the solvent dissolving whatever Nikon used to hold the sensor to the PCB. When this dried I decided to test the sensor before going any further, in case it did not work, so as to not waste any more time on a dead sensor. The image was of poor quality but I wasn't looking for clarity, since the sensor wasn't cleaned very well, just testing that it was working and the camera could get a full readout of the sensor. To remove the bayer filter itself, I used a sharpened wooden dowel with a rounded tip to rub the sensor until I could see the bare sensor beneath. Before I did this, I had to protect those fine wires, that can break if you breath hard on them. I used epoxy and slowly let it fill the trough that surrounds the sensor, in which all the wires are located. Once this dried I went to work rubbing the filter off. The bayer filter looks to the naked eye like the sheen oil produces when on water, a rainbow effect. With the filter removed it has a dull gold color appearance. From what I have read there is an edge that must be avoided or risk killing the sensor. Not sure why, but at first I have concentrated on the center, with mild to medium pressure the filter is removed. I have taken caution not to subject any particular area to cleaning more than any other, as digging too deep into the semiconductor will leave those pixels with a scratch that is easily visible or even kill a full row of pixels. From what I have read though to kill the whole row you have to dig pretty deep and the semiconductor is extremely hard and durable. All the scratches and variations will be handled with the use of Flat frames to eliminate any exposure differences.

I haven't yet completed the debayering process of the sensor, as it seems like there isn't enough time in a day with being fairly busy at work. But since I won't be doing any DSO imaging anytime soon I should have a bit of free time to play with it. I believe this pretty much covers what I have done thus far. I cannot say this mod will be worth the effort put into it, but when bored I like to tinker.

Sorry for hijacking your thread Craig,
Matthew

CraigFair Loc: Santa Maria, CA.

Oknoder wrote:

Sry for the late reply. I thought I sent one but apparently my phone decided you didn't need to read, or that I needed to rewrite the whole thing.

Once the sensor assembly was removed, which holds an aluminum frame, PCB with a couple of ribbon connectors and the actual sensor itself. I removed the PCB from the frame and as per the guides I found, attempted to use a heat gun, from my rework solder station, to loosen and remove the glass covering the sensor. It did not work very well and in the end, I used a diamond grinding wheel on my dremel to cut the glass. This had to be done with extreme care, there are dozens of fine hairlike wire connectors, which are used to read the data coming off each photoreceptors. After I was able to cut away all the glass, without damaging any of the wires, I wont say how many sensors I destroyed but I wasn't successful with the first couple, LOL.

With the glass removed I tried a couple of different solvents to help remove the microlenses and bayer filter. The one I found that worked the best was Methylene Chloride or by common name Dichloromethane. This is a very caustic solvent so extreme caution should be used. I used a small amount and filled a pyrex container, to the point that the sensor would be covered by the solution. I then placed the pyrex petri dish in an air tight bowl, as it evaporates very quickly and gives off noxious fumes. It must be added that I conducted this experiment in the garage, and when the sensor was in the petri dish I left it to soak for 36hrs, within a fire cabinet I have in my garage.

After pulling the dish out it was apparent that my air tight bowl was not really air tight as the majority of the solution had evaporated, the microlenses were removed easily and looked like a coating of snot covering the sensor. Using RO/DI water and Qtips to wipe the sensor clean of the microlenses. The actual bayer filter is printed directly on top of the semiconductor used as the sensor's photo-receptor. Before attempting anything else I had to add a couple of dabs of epoxy, due to the solvent dissolving whatever Nikon used to hold the sensor to the PCB. When this dried I decided to test the sensor before going any further, in case it did not work, so as to not waste any more time on a dead sensor. The image was of poor quality but I wasn't looking for clarity, since the sensor wasn't cleaned very well, just testing that it was working and the camera could get a full readout of the sensor. To remove the bayer filter itself, I used a sharpened wooden dowel with a rounded tip to rub the sensor until I could see the bare sensor beneath. Before I did this, I had to protect those fine wires, that can break if you breath hard on them. I used epoxy and slowly let it fill the trough that surrounds the sensor, in which all the wires are located. Once this dried I went to work rubbing the filter off. The bayer filter looks to the naked eye like the sheen oil produces when on water, a rainbow effect. With the filter removed it has a dull gold color appearance. From what I have read there is an edge that must be avoided or risk killing the sensor. Not sure why, but at first I have concentrated on the center, with mild to medium pressure the filter is removed. I have taken caution not to subject any particular area to cleaning more than any other, as digging too deep into the semiconductor will leave those pixels with a scratch that is easily visible or even kill a full row of pixels. From what I have read though to kill the whole row you have to dig pretty deep and the semiconductor is extremely hard and durable. All the scratches and variations will be handled with the use of Flat frames to eliminate any exposure differences.

I haven't yet completed the debayering process of the sensor, as it seems like there isn't enough time in a day with being fairly busy at work. But since I won't be doing any DSO imaging anytime soon I should have a bit of free time to play with it. I believe this pretty much covers what I have done thus far. I cannot say this mod will be worth the effort put into it, but when bored I like to tinker.

Sorry for hijacking your thread Craig,
Matthew

Sry for the late reply. I thought I sent one but a... (show quote)

No problem Matthew it's all about Astronomy.
Craig

SonnyE Loc: Communist California, USA

CraigFair wrote:

No problem Matthew it's all about Astronomy.
Craig

...And cameras and equipment. ;)

JimH123 Loc: Morgan Hill, CA

Oknoder wrote:

Sry for the late reply. I thought I sent one but apparently my phone decided you didn't need to read, or that I needed to rewrite the whole thing.

Once the sensor assembly was removed, which holds an aluminum frame, PCB with a couple of ribbon connectors and the actual sensor itself. I removed the PCB from the frame and as per the guides I found, attempted to use a heat gun, from my rework solder station, to loosen and remove the glass covering the sensor. It did not work very well and in the end, I used a diamond grinding wheel on my dremel to cut the glass. This had to be done with extreme care, there are dozens of fine hairlike wire connectors, which are used to read the data coming off each photoreceptors. After I was able to cut away all the glass, without damaging any of the wires, I wont say how many sensors I destroyed but I wasn't successful with the first couple, LOL.

With the glass removed I tried a couple of different solvents to help remove the microlenses and bayer filter. The one I found that worked the best was Methylene Chloride or by common name Dichloromethane. This is a very caustic solvent so extreme caution should be used. I used a small amount and filled a pyrex container, to the point that the sensor would be covered by the solution. I then placed the pyrex petri dish in an air tight bowl, as it evaporates very quickly and gives off noxious fumes. It must be added that I conducted this experiment in the garage, and when the sensor was in the petri dish I left it to soak for 36hrs, within a fire cabinet I have in my garage.

After pulling the dish out it was apparent that my air tight bowl was not really air tight as the majority of the solution had evaporated, the microlenses were removed easily and looked like a coating of snot covering the sensor. Using RO/DI water and Qtips to wipe the sensor clean of the microlenses. The actual bayer filter is printed directly on top of the semiconductor used as the sensor's photo-receptor. Before attempting anything else I had to add a couple of dabs of epoxy, due to the solvent dissolving whatever Nikon used to hold the sensor to the PCB. When this dried I decided to test the sensor before going any further, in case it did not work, so as to not waste any more time on a dead sensor. The image was of poor quality but I wasn't looking for clarity, since the sensor wasn't cleaned very well, just testing that it was working and the camera could get a full readout of the sensor. To remove the bayer filter itself, I used a sharpened wooden dowel with a rounded tip to rub the sensor until I could see the bare sensor beneath. Before I did this, I had to protect those fine wires, that can break if you breath hard on them. I used epoxy and slowly let it fill the trough that surrounds the sensor, in which all the wires are located. Once this dried I went to work rubbing the filter off. The bayer filter looks to the naked eye like the sheen oil produces when on water, a rainbow effect. With the filter removed it has a dull gold color appearance. From what I have read there is an edge that must be avoided or risk killing the sensor. Not sure why, but at first I have concentrated on the center, with mild to medium pressure the filter is removed. I have taken caution not to subject any particular area to cleaning more than any other, as digging too deep into the semiconductor will leave those pixels with a scratch that is easily visible or even kill a full row of pixels. From what I have read though to kill the whole row you have to dig pretty deep and the semiconductor is extremely hard and durable. All the scratches and variations will be handled with the use of Flat frames to eliminate any exposure differences.

I haven't yet completed the debayering process of the sensor, as it seems like there isn't enough time in a day with being fairly busy at work. But since I won't be doing any DSO imaging anytime soon I should have a bit of free time to play with it. I believe this pretty much covers what I have done thus far. I cannot say this mod will be worth the effort put into it, but when bored I like to tinker.

Sorry for hijacking your thread Craig,
Matthew

Sry for the late reply. I thought I sent one but a... (show quote)

Wow! I don't think I will be trying this anytime soon. This some extreme effort.

CraigFair Loc: Santa Maria, CA.

JimH123 wrote:

Wow! I don't think I will be trying this anytime soon. This some extreme effort.

Thank you Jim H for not HiJacking my thread too. :lol: :lol:
Craig

SonnyE Loc: Communist California, USA

:hunf: