Monochrome conversion via sensor swap!

[Herra Kuulapaa]

I'm warming up this old thread.

 

My background is that I have monochrome modified a lots of different cameras ranging from old ones to new (Nikon, Sony, Canon), but mostly I'm making D5100, D600, D800 Nikons due to the Exmor sensor type I really like due to its impressive performance. Now this is not advertising, just information, and my motivation is to experiment, develop the process and helping fellow experimentalists :)

 

Back to the sensors and modifying then. Issue with modern full frame Nikon/Sony Exmor sensors is that the cover glass is sitting very tight on the sensor. They have in fact used a resin that has a release temperature higher than what sensor can take – And I have cooked a good amount of sensors to say this. So the issue is how to remove the cover glass. To prevent using heat I have moved to using a CNC method for that. Process is to very slowly machine three sides of the glass off and then fixing glass to a holder and machine the last side off. Speed needs to be so slow that glass is reduced into dust rather than sharp shards flying around. :(

 

Using this method I’ve been able to successfully modify a number of D600/D610 and D800 sensors, which have been used with excellent results in astrophotography narrow band imaging and terrestrial NIR. I however got interested in UV imaging and now the plan is to further explore the capabilities of naked monochrome sensors in this field. I believe that the naked sensors paired with proper optics (which I'm unfortunately lacking) provide a clear improvement in sensitivity due to lacking glass/RGB layers. Especially in the UVB range - What do you think?

 

Ps. I emphasize I'm not writing this with marketing in my mind, but to study and contribute :)

[Andrea B.]

Herra, the UV world is small. So it is important for UVP to extend membership to photographers who are also vendors. Such folks have provided excellent information about various UV matters. As long as nothing is directly marketed within a post and there is no advertising or prices mentioned, we are happy.

 

When a vendor is offering a new product or service, I do permit one Product Announcement here on UVP as long as we can pre-approve the contents.

 

************

 

I find this very interesting that the glass can be removed from over a sensor by machining. As mentioned (somewhere) the only person I know who has removed the Bayer filter here in the US is Dan Llewellyn at MaxMax.com. He mentions on his website improved UV sensitivity once that glass is removed.

 

What means "CNC", please?

 

Is the resin sealant along the edges of the glass cover? I ask because you mention machining off "three sides" and then the "last side". I had originally envisioned machining off the "top" surface of the glass cover. But what you have described seems to indicate the edges of the glass cover.

 

As for UVB, it is hard to say just how much might be recorded unless you test with narrowband UV-pass filters. I've always been curious to try UVB, but do not have such filters. Also there is the problem of illumination. Not much UVB available assuming UVB is approximately 280-315nm. The ratio of UVB:UVA in sunlight is usually considered to be about 1:20 as per this reference:

 

The Value of the Ratio of UVA to UVB in Sunlight

2011 by Kollias, Ruvolo and Sayre

Photochemistry and Photobiology, Wiley Online Library

[Andy Perrin]

I am thinking CNC is using "computer numerical control"? It's basically a robotic milling machine. For when your hands just aren't enough.

[Herra Kuulapaa]

Yes, CNC is pactically a computer driven cutter drill, which can move at least along x,y,z coordinates - just to do steady and precise work when our hands and other tools are not enough like Andy said.

 

Resin is located between the cover glass and the sensor frame and there are at least different types of it in different sensor models. Here is a Nikon D600 sensor, which is a good example of sensor quite difficult to be modified:

http://www.kuulapaa.com/Tahtidata/D6001.jpg

 

Cover glass needs to be machined off by the edges to prevent glass shattering like it would do if just gradually thinned (shattering is still an issue, but reduced by this edge approach). Glass is tempered by its nature and I believe it's somekind of quartz, very hard and brittle. It also eats tools like bread :(

 

Here is a Nikon D5100 sensor. Light is slightly different, but you can see a different shade in the resin. Also the glass is seated in small depression. Resin makes D5100 sensor a bit easier to be modified:

http://www.kuulapaa.com/Tahtidata/D51001.jpg

One can remove the cover glass from the D5100 sensor by heating the glass with very hot soldering iron or small torch. Try to apply heat only on small surface and limited time to prevent heat originated bricking.

 

The most sensitive thing are the gold bonding wires. They can be broken even with staring them too long :(

http://www.kuulapaa.com/Tahtidata/Alaston%202.jpg

 

Also I recommend avoiding going too near the sensor surface edges. There is a myriad of micro wirings on the surface and damaging them will lead to interesting behavior of the sensor.

http://www.kuulapaa.com/Tahtidata/Chip_2jpg.jpg

 

I have one D600 sensor available for anyone who would like to experiment with monochrome sensor with only "green" pixels, this is joke of course although the sensor really is like that now and I think the total sensitivity for UV is quite good ;)

[Avalon]

Great that you found way to swap sensors with monochrome version. Although I see it's older CCD technology with microlens. Backlit CMOS have greater sensitivity especially in UV spectrum. I'm curious it this sensor with silicon nitride coating which blocks UV and blue spectrum?