B/W DSLR modification study

Generally color image sensor is arranged as the pattern of RGBG 2×2 pixels color filter array, It can make the color images through 1 shot exposure but the large amount of light were blocked by color filters on CMOS and color processing algorithm produce false colors. The optical low pass filter to prevent false color on color image sensor makes lower resolutions to output image.
In low light photography field B/W sensor with high QE and low noise is most valid device for higher resolution image.

The LRGB combination using B/W CCD camera is the major way to make the high quality astro images.
The modification of color image sensor of high quality DSLRs to B/W sensor is so meaningful , because the good price and lower noise level and wide sensor area of full frame and APS-C size sensors are so attractive benefits of DSLRs.

Since last year we have tried to change color image sensor to B/W sensor. Many image sensors were killed for various debayering tests but we haven’t got some valid fruits till 2012 November.

Below are the history of B/W sensor modification.

1. We thought that the material of micro lens is so differ from that of Color Filter Array and tried to remove using cured epoxy bonding to sensor surface but the epoxy bonding was failed to detach the micro lens layer from CFA array.

2.Using optical polishing powder tried to grind  CMOS of eos 60D, CCD of Nikon D40, Kodak 8300 color CCD. This method was clearly effective but we can’t get equal polishing rate to whole sensor surface and many scratches on image sensor were made.


3. We assumed that the material of micro lens and CFA array are the sorts of plastic and we soaked the sensors in to organic solvent for a long time.

In case of 8300 CCD , The micro lens layer and CFA array were completely removed and became B/W CCD. The micro lens and CFA of D40 CCD were melted but the bonding of CCD wafer and ceramic base was melted too and CCD sensor were separated. But those of 60D CMOS were never responded to solvent.


4.Using knife we skive the micro lens and CFA layers of CMOS surface. This is certainly effective way but the traces of knife and some deep cut remained.

5 We tried to find the way of uniform separation of whole sensor area and found the slow etching solution for glass sliming in LCD factory. In etching solution micro lens and CFA were effectively removed but the continuous corrosion were occurred at CMOS wafer and gold wiring and terminals.

6. During 7month we stopped to find new way but last month we found a very useful short movie for CFA removal at youtube by Luís Campos living in Portugal. He is very passionate amateur astro photographer and he tried to remove of his 350D camera. His tool is only a sharpened wood stick and we made his camera to perfect B/W camera.

Thanks to Luís Campos, we can try to strip the 60D CMOS of micro lens and CFA layers using bamboo chopstick. It’s needed strong grip and the speed was so slow, but surprisingly the wafer’s surface was no scratches and clean. There are some remained so small microscopic particles of CFA, but we can remove micro lens and CFA without wounds on CMOS wafer.

Through the work of great care takes about 4 hours APS-C sensor of 60D was changed B/W sensor. In my naked eye the surface quality of CMOS was 99.5% quality.

We are to test this B/W sensor sooner or later and will open the comparison data of before and after.



2 replies
  1. wide-field
    wide-field says:

    That is very interesting Mr Lee. My Central DS Astro 60D does a fantastic job at capturing Narrow Band data. (See my latest Hydrogen Alpha image of Horsehead Nebula (sent to the Gallery today).
    I would be very interested to see if your mono version of the 60D is substantially better.
    If it is, then it would be an amazing camera indeed !!
    I look forward to your test results.

Comments are closed.