Leica KAF-18500 CCD used in the M9 and M Monochrom is discontinued, Sensor replacement no longer possible

The S8612 is a "Class 3 Glass" which corrodes in the short term when not protected. The Data Sheet states that it should be used in a sandwich with Class 1 glass, such as BG38, to protect it from the environment. BG55 is a class 2 glass- " Long-term changes in the polished surface are possible under some circumstances."

NOW for real confusion: the individual Jan 2014 Data Sheet (STATUS: 01.12.2014) for S8612 lists it as a Class 2 glass, and it is clearly stated as a Class 3 glass in all other data sheets and the Schott catalogs. I Emailed Schott a long time ago about this. A revised Data Sheet shows it back as a Class 3.

The 1/2014 Data Sheet that disagrees with all the other data sheets.
It even refers you to the 2013 catalog, which clearly indicates it is Class 3. I've learned to download data sheets, look for revised data sheets, and keep all of them.

This is from the Schott optical catalog:
Group 1
No substantial surface change occurs in most of the optical filter glass types.
These types are not identified specially in the “Properties” brochure.
A change in the surface is only possible under extreme conditions, if subjected
to a continuous spray of sea water, or if used in rain or water.
Group 2 Symbol:
For the optical filter glass types BG18, BG40, BG50, BG55 and all KG glass types,
there is virtually no long-term change when used and stored in moderate climates
or in closed work and store rooms (constant temperature below 35 °C, relative
humidity
less than 80 %). A desiccant should be used if the possibility of wetting
exists. For use and storage in open air and tropical climates, it is advisable to apply
a protective coating which SCHOTT can provide upon request.
Group 3 Symbol:
For the optical filter glass types BG42, UG5, UG11, BG39, S8612, S8022 and S8023
a change in the glass surface is possible after a few months of normal storage.
For this reason, applying a protective coating or lamination is recommended for
durable optical filter glass from Group 1 (SCHOTT can provide both).

SO- The difference between class 2 and class 3 is substantial. ONSEMI/SCHOTT/LEICA probably believe the coatings used on the BG55 filter are good enough to protect a class 2 glass, even if it did not work on a class 3 glass. The latter- my speculation. AND if they are wrong, and my BG55 filters ends up corroding, I'll send to you to make it full-spectrum.

If anyone wants this catalog-

 
Last edited:
The S8612 is a "Class 3 Glass" which corrodes in the short term when not protected. The Data Sheet states that it should be used in a sandwich with Class 1 glass, such as BG38, to protect it from the environment. BG55 is a class 2 glass- " Long-term changes in the polished surface are possible under some circumstances."

NOW for real confusion: the 2014 Data Sheet for S8612 lists it as a Class 2 glass.

The 2014 Data Sheet-

I Emailed Schott a long time ago about this. A revised Data Sheet shows it back as a Class 3.

This is from the Schott optical catalog:
Group 1
No substantial surface change occurs in most of the optical filter glass types.
These types are not identified specially in the “Properties” brochure.
A change in the surface is only possible under extreme conditions, if subjected
to a continuous spray of sea water, or if used in rain or water.
Group 2 Symbol:
For the optical filter glass types BG18, BG40, BG50, BG55 and all KG glass types,
there is virtually no long-term change when used and stored in moderate climates
or in closed work and store rooms (constant temperature below 35 °C, relative
humidity
less than 80 %). A desiccant should be used if the possibility of wetting
exists. For use and storage in open air and tropical climates, it is advisable to apply
a protective coating which SCHOTT can provide upon request.
Group 3 Symbol:
For the optical filter glass types BG42, UG5, UG11, BG39, S8612, S8022 and S8023
a change in the glass surface is possible after a few months of normal storage.
For this reason, applying a protective coating or lamination is recommended for
durable optical filter glass from Group 1 (SCHOTT can provide both).

SO- The difference between class 2 and class 3 is substantial. ONSEMI/SCHOTT/LEICA probably believe the coatings used on the BG55 filter are good enough to protect a class 2 glass, even if it did not work on a class 3 glass. The latter- my speculation. AND if they are wrong, and my BG55 filters ends up corroding, I'll send to you to make it full-spectrum.
I see, this Is awfully confusing isn’t it? I’ll bet whoever made the decision to use s8612 is probably kicking himself right about now! i did a few test shots and the BG40 seems like it improves just slightly when using a coated UV filter (B+W UV MRC) I should also note that I’m using a 35mm asph summicron. So it would seem that my only option to preserve most of the look and the focus will be the BG40 glass... thoe new sensors should be here in a week or so, I’ll covert them and post my findings. I suspect that the color shift may be coming from the sensor overheating as I accidentally removed some of the thermal epoxy while removing the original filter. I’ll keep you put to date on what I find
 
I'll bet you are right about the color shift and the sensor overheating!

I really appreciate this conversation. Nice to see someone that loves their work.

The project manager that oversaw the glass selection should have caught this. An engineer wants to get the best possible performance with components selected, S8612 is optically the best performing glass from Schott for the job of cutting IR and preserving UV. The latter is important to some people, especially in the technical field. In 2009, Kodak was out of the high-end camera business and the CCD's were aimed at the scientific market. I've used the Pentax 85/4.5 Ultra-Achromatic on the M Monochrom. The engineers bet that the "seal" (both sides coated) selected to protect the filter would stop the corrosion. They bet, and where wrong. The Project Manager should have realized this was "iffy" at best, went against the recommended procedure from Schott for a Class 3 glass, not enough life-testing was performed, and used another glass. BG55 was reintroduced in 2011, too late for the M9 and M Monochrom. Managing engineers can be like herding cats. That's what my supervisor told me when I was new to being a manager 26 years ago. I'm glad to be "Senior Staff" and back in the Lab.
 
Last edited:
So a quick questions for any one with an M9. As I’ve said before I’ve only worked on film bodies before this M9 sensor debacle so I have a slim understanding of how the processing works in these, when the camera is in auto white balance is it determining the WB off the sensor itself, the light meter, or what looks to be a small sensor above the VF? I’ve been doing some more tests with the new sensor I put together now that the smokes clear and it’s very inconsistent In auto white balance but when setting it “manually” (taking a shot at a white surface or the sky then it sets the temp for you) the colors come out perfect. I’ve been contradicting myself quite a bit with the BG40 glass after each test but I understand now that it’s the white balance that’s the issue. Since playing around with the manual white balance I’ve been able to get very similar results compared to the original filter. I’m definitely settled on what glass I’ll be using for these repairs but It seems like white balance would have to be manually set unless it’s just my particular camera that’s having an issue.
 
Also I did a few more tests with heating and cooling the sensor and it seems that heat will throw off the colors quite a bit giving an almost Full spectrum look, in extreme cases it also leads to the failure of on side of the sensor mostly the left it seems. I have one more test I want to run with the new sensors but I think I already know the answer, I just want to be sure. If all goes well I’ll be open to taking in repairs around the 1st week of October if anyone is interested. I’m still working on getting a business license but this project has cost me over $2k on top of my other bills so once I can get some revenue coming in I’ll be able to make thing official.
I like the “Sensor Surgeon”idea but the M9 sensor repair is really the only sensor related service I’d offer. I like the sound of “Leica surgeon” but I’m pretty sure I can’t use their name
 
>As I’ve said before I’ve only worked on film bodies before this M9 sensor debacle so I have a slim understanding of how the processing works in
>these, when the camera is in auto white balance is it determining the WB off the sensor itself, the light meter, or what looks to be a small sensor
>above the VF?

The little white window is the color balance sensor. It handles the auto-white-balance. The AWB is set expecting the S8612 transmission curve. The new BG55 sensor required both a change in the color Dye of the Bayer filter and to the firmware. You cannot change either of those. SO- you are left to doing this once the DNG file is out of the camera. Setting the white balance manually- small price to pay for a sensor without corrosion. Some sort of post-processing software to adjust the stored red-green-blue values stored in the DNG file might also work. Basically, apply "fudge factors" to bring the values close to where they would be for the original cover glass.

Cannot use Cover Glass Doctor- too close to Glass Doctor.
I like Advanced Sensor Rescue, "ASR". Also a rescue ship. You are the Chief Surgeon.
Have to administer CGR. Cover Glass Repair.
 
Last edited:
>As I’ve said before I’ve only worked on film bodies before this M9 sensor debacle so I have a slim understanding of how the processing works in
>these, when the camera is in auto white balance is it determining the WB off the sensor itself, the light meter, or what looks to be a small sensor
>above the VF?

The little white window is the color balance sensor. It handles the auto-white-balance. The AWB is set expecting the S8612 transmission curve. The new BG55 sensor required both a change in the color Dye of the Bayer filter and to the firmware. You cannot change either of those. SO- you are left to doing this once the DNG file is out of the camera. Setting the white balance manually- small price to pay for a sensor without corrosion. Some sort of post-processing software to adjust the stored red-green-blue values stored in the DNG file might also work. Basically, apply "fudge factors" to bring the values close to where they would be for the original cover glass.

Cannot use Cover Glass Doctor- too close to Glass Doctor.
I like Advanced Sensor Rescue, "ASR". Also a rescue ship. You are the Chief Surgeon.
Have to administer CGR. Cover Glass Repair.
Yeah I know the curve is a lot different on the BG55 but the bg 40 is close with the color dye but the transmittance is higher across the board, the thing that confuses me is that sometimes it will work just fine and other times it won’t. When it does work the colors are great, the shadows are little bit lighter because of the higher transmittance of the glass. Using the manual white balance setting gives perfect results every time so that leads me to believe I might be using the the camera wrong. (Hurts a little to say that being that I repair the things but I’ve never owned a digital up until now)

Advanced sensor rescue would be way too close to the name of the shop I currently work for, they’re already aren’t going to that I’m planning to leave and start my own Leica specialty service so I won’t to try and ruffle the least amount of feathers as possible haha
I definitely want something universal since the M9s will eventually either not be worth repairing or (optimistically speaking) there won’t be any more sensors to repair. Maybe M Series surgeon? I’ll offer Coverglassectomies, framelineotomies, and oil transfusions...
 
M Electronic Series Surgeon- Let us handle this MESS.

Hard to say why using the manual white balance yields perfect color but the AWB does not! Just does not make sense.

The BG55 would do no better on the original sensor than the BG40 you are using- because ONSEMI reformulated the color dye of the Bayer filter at the same time they changed cover glass.

I would be interested in seeing some images posted here, out walking around with the camera.
 
Here’s some test shots, nothing pretty just quick little snapshots on my day off.
You can see what I mean by “sensor failure” on the left side of the image. This was caused by me removing the thermal epoxy my first time around. I’ve since tried replacing it with super X which helped a little bit it’s still lacking the heat absortion that thermal epoxy has.
Anyway here’s what it looks like with auto white balance, as you can see the colors aren’t terrible but not exactly great
4DA94DC8-E24B-4C47-9530-04EE514E5546.jpeg

here’s a shot after “manually” setting the white balance to the clouds or in this case my white garage just out of the frame. (I say “manually” because in my eyes this is still very much automated, id say its more “selective” than truly manual but I didn’t make the rules) the colors in this image are much better, it almost reminds me of more Portra than the Kodachrome that these sensors were said to be emulating.
FFDFEC7D-3499-45DA-8727-AB2D79F869C2.jpeg
 
The KAF-18500 is really two CCD chips joined together with multiple readouts on each side of the chip.
But that color shift is bizarre.

The colors of the good half look close enough to fix in post for anyone wanting Kodachrome. You might also try shooting with an IR cut filter (Hot Mirror) over the lens to see if IR leakage is causing a color shift. The shade of green looks a bit like an M8 without a Hot Mirror filter over the lens.

The thermal epoxy is extremely important. The noise level for the CCD goes up "A LOT" with temperature.

Dark Signal Doubling Temperature = 5.3 °C

OKAY! What's the point of having all these data sheets and just saying "A LOT". The Sensor Noise doubles with every 5 degrees centigrade. Another reason why I don't ever shoot continuous mode and don't believe in Liveview on cameras that will not publish noise vs temperature.
 
I can see the similarity to the M8 in the greens, I’ll take try to take some more shots today assuming the sensor will allow it, since it’s heats up so fast I get maybe 3 or 4 good shots before cooling down and at least one of those has to be wasted for setting the white balance. I think there’s some hot mirror filters at work so try that out tomorrow.
The spare sensors are showing up today, I’ll install an original one and take some more controlled test shots so we can have a more accurate before and after. I feel like I’m getting really close to having this all dialed in, granted it won’t be EXACTLY the same as the original but it does have a slight advantage in the shadows. Plus I’ll be offering the the service for cheaper than what kolari is charging.
 
That sounds great. I'm keeping watch for an M9, M-E, or M Monochrom at a good price.

Getting the thermal epoxy is probably most critical at this point- especially before experimenting with another CCD.

Somewhat related- do you also have a suitable Clear coverglass to make the CCD full-spectrum? Would need to by 0.8mm as well.
I've always wanted an M9ir. If I find an M9 with corroded cover glass, I'm grabbing it. I like to use Magenta filters with my full-spectrum color camera to block Green and leave the green channel as IR only.
 
That sounds great. I'm keeping watch for an M9, M-E, or M Monochrom at a good price.

Getting the thermal epoxy is probably most critical at this point- especially before experimenting with another CCD.

Somewhat related- do you also have a suitable Clear coverglass to make the CCD full-spectrum? Would need to by 0.8mm as well.
I've always wanted an M9ir. If I find an M9 with corroded cover glass, I'm grabbing it. I like to use Magenta filters with my full-spectrum color camera to block Green and leave the green channel as IR only.
Yeah the thermal epoxy is very important, so important that it’s best to leave it in tact! My new CCDs just came in the mail, gotta sus it’s interesting to see the different patterns the corrosion takes on.
as for full spectrum, I can definitely source clear coverglass. I’ll try for .8mm but I’m the event I can only find 1mm I can install a .1mm shim. The original layer of epoxy was .1mm thick so adding that onto the 1mm filter I’d have to bring it back .1mm to cancel it out. I Shot a 90 summicron and noticed it slightly back focusing. At .1mm it didn’t cause any side effects with wide angles, it’s a such a small difference that it there is, dare I say, tolerance to compensate for any discrepancies in manufacturing.

But a short answer to your question is yes I can do a full spectrum conversion.
 
So just for fun I installed an original sensor in my M9, I shot it around for a bit and the white balance is still terrible so it’s either my particular camera or the M9s which from what I’ve been reading I think it might just be M9s in general. Anyway I went outside and shot a shrub in front yard with the original sensor, then I shot it with the “new” (bg40) sensor and the results are pretty shocking... I’ll let you all judge it for yourselves! Also I didn’t record what shutter speed I shot these at, but these were shot at 5.6 160iso on a 35 cron asph.
This is the original filter
75418B07-834E-4A90-A3FB-FFE50CF118DA.jpeg
14CF2DA1-0A5E-472C-8F4E-599198FA0BE4.jpeg
1F5693B3-0D54-4B71-BAF3-42798E494376.jpeg

9C87D40D-0AFA-4405-B939-E6045CF567A3.jpeg

And the BG40 sensor
324F5769-DD5D-4503-97FA-569BC1A3E4C3.jpeg
D42E27C1-F13E-49DB-BA3B-AE85ECD29CD0.jpeg
8325FEE5-2329-4493-B117-6AB9866D4808.jpeg
335B0571-69F3-4E81-B2BC-E8C18171E341.jpeg
 
How bad was the glass on the original Cover Glass?

The BG40 images look really good.

If the thermal issue is solved- I'd declare victory and you would be in business.
 
How bad was the glass on the original Cover Glass?

The BG40 images look really good.

If the thermal issue is solved- I'd declare victory and you would be in business.
The original is pretty bad, the glass is almost pitted through

and the overheating wont be a problem anymore so long as I leave the ceramic sensor housing in tact. the only reason It ever was is because I removed part of the sensor I should have left alone.

Comparing the two side by side I like the contrast of the original but I prefer the dynamic range of the BG40, that being said contrast can always be added. But what a Relief! After almost 2 long months of staring at data sheets, disassembling and reassembling, and mindlessly testing i feel like I’ve reached a conclusion!

I’ll shoot for taking in repairs around the first or second week of October, I’ll keep everyone up to date on when this will happen since I have a few things to figure out on the administrative end (insurance, name, and invoicing stuff)

Thank you Brian and everyone else in this forum for the feedback and support!
 
Back
Top