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

I've been seeing some news pop up in my social media about repairing the M9s, is that you?

I have an M9-P which hasn't corroded yet but I was really disappointed when Leica announced they were going to end their sensor replacement program. I'm a lot happier knowing that it will remain fixable whenever it starts to corrode.

Oddly enough no. I believe that’s kolari vision. They made their announcement right after I made that initial post a few days ago. Really funny timing I guess we’re all on the same page haha! I reached out to leica rumors just so I can get my story out there. I’ve been considering going freelance with my repairs so I’ll hopefully have some social media stuff going pretty soon. How does Red Dot Repair sound? I’d have to do it mostly part time until it becomes clear enough that there’s a market for my services and then I’ll go full sail!
 
There is a "Red Dot Camera" that sells Leica already- might confuse your service with their shop.

You are doing the equivalent of Cataract Surgery for Leica cameras.

Please post here with results. I have both an M9 and M Monochrom with new sensors in them. If you need anyone to test your repaired cameras side-by-side with the ones repaired by Leica, I would be happy to do it.
 
There is a "Red Dot Camera" that sells Leica already- might confuse your service with their shop.

You are doing the equivalent of Cataract Surgery for Leica cameras.

Please post here with results. I have both an M9 and M Monochrom with new sensors in them. If you need anyone to test your repaired cameras side-by-side with the ones repaired by Leica, I would be happy to do it.
Oh yeah that’s true. I knew a name wouldn’t come by that easy. And I would certainly be open for testers once I get one going. Unfortunately there was a slight complication with the sensor while I was prepping it for the replacement. I kept it covered but I noticed some dust had settled on it, I used a swab to clean as I would any sensor (including raw sensors) and the solution used had left marks on the micro lenses and damaged a part of the coating. This sensor is like nothing me or my colleagues have seen before. With it being essentially ruined I started experimenting with ways to clean the sensor without damage and it seems rodico does the trick. The sensor I have works just fine but the damage in the lower section of the image isn’t very desirable. I’m now trying to source another corroded sensor to repair. im hoping I can get some directly from Leica but that’s a long shot.
 
I wonder what Leica did with the thousands of sensors that were returned to them when they were doing replacements.
I wonder how many are sitting on a shelf, or if they were returned to ONSEMI to get new cover glass.

Leica replaced the entire board when replacing the CCD. The CCD is not cheap, the board is probably a fraction of the cost. But I doubt they were just tossed out. Even at $1600 to replace the CCD that Leica charged in the end, that does not cover the price of buying a new sensor and board. Anybody else think they were refurbished with new glass?
 
I wonder what Leica did with the thousands of sensors that were returned to them when they were doing replacements.
I wonder how many are sitting on a shelf, or if they were returned to ONSEMI to get new cover glass.

Leica replaced the entire board when replacing the CCD. The CCD is not cheap, the board is probably a fraction of the cost. But I doubt they were just tossed out. Even at $1600 to replace the CCD that Leica charged in the end, that does not cover the price of buying a new sensor and board. Anybody else think they were refurbished with new glass?
That’s something I’ve been trying to figure out, I reached out to leica’s parts department to see if they recycled them or if they were just thrown away. With how difficult it is to remove the original filter (They are literally CAKED with epoxy!) I highly doubt they were putting in the time to replace the filter themselves. As for the price I think that’s mostly coming from the sensor itself, the M9 is a very modular build. No disrespect to Leica at all, I think the design is genius but If you can disassemble a hasselblad you can just about disassemble the M9 in my opinion. Wouldn’t recommend it but I definitely thought I’d put that out there.
 
m8_cover.jpg
Join to see EXIF info for this image (if available)


"Only Somewhat related".... I had a difficult time finding the cover glass used for the Leica M8, it is not specifically given in the KAF-10500 data sheet. This glass does not corrode, it is a "Class 1 Glass", very stable. Leica used a 0.5mm thickness, and the glass passes 5% in the IR.

Would not be worth it now- but this could have been replaced with BG55 or another glass cover to eliminate the worst of the IR problem.

I'm posting this in the discussion mostly so I can find it again in the future.
Keywords: KAF-10500 cover glass Kyocera B-7.
 
Exciting news! The new filters just got delivered today, it was cut from an 80X80mm square so the seller made me 3 filters for the price of one! Even though my current sensor is slightly damaged after some experiments with different methods of cleaning I’ll still go through with the repair. I should hopefully have some spare sensors on the way as well so I can have a small stock of them. Unfortunately Leica won’t let go of any old sensors Being that this cameras is pushing 10 years old this might be their way of saying it’s time to upgrade. Anyway I’ll get the sensor back together when I get some free time this week and post my findings!
649D8F0C-D8B3-4062-8E86-64B0EAB1DDB2.jpeg
 
I wish you luck and all the best for this venture! A lot of people with affected cameras are now having some hope for keeping and using them.

I've noted that Kolari suggests using a 0.6mm stack- this means having to reposition the sensor, which would be difficult for an RF camera. The position has to be perfect for the special "Offset Microlens array". Moving the sensor closer to the rear of the lens would slightly change the geometry. The Offset Microlens array- the lenses just do not point straight up. Lenses are offset to point towards the center, to help with the short flange distance of the M Mount.

Someone on the Leica Users Forum stated that MaxMax was using BG38 for a replacement filter.

I found this at the MaxMax site:


The filter that MaxMax proposes to use is the XniteCC1, and could possibly be BG38. It has about the same transmission in IR as the filter used on the M8. The BG38 is very stable, but has the same problem for IR leakage as the Kyocera B-7 that caused the problems with the Leica M8. The CCD sensor has much more, more than 10x, the sensitivity to IR than does a CMOS sensor. A few years ago I helped out on a project to build a new low-cost IR sensor using COTS detectors- and did a lot of comparisons of available sensors. The answer was a CCD for Near IR. At work I've spent $25K (over 15 years ago) on a Sensors Unlimited sensor with ~320x240 resolution, but it used InGaAS for IR sensitivity.

Somehow Leica+OnSemi decided to use a 0.8mm BG55 glass. This glass was not available in 2008/2009 when the M9 was being developed and introduced. BG55 was reintroduced in 2011. This solved the IR problem, and the repositioning problem. Same Refractive Index, more stable- a Class 2 Schott glass. S8612 is class 3.
 
Last edited:
I wish you luck and all the best for this venture! A lot of people with affected cameras are now having some hope for keeping and using them.

I've noted that Kolari suggests using a 0.6mm stack- this means having to reposition the sensor, which would be difficult for an RF camera. The position has to be perfect for the special "Offset Microlens array". Moving the sensor closer to the rear of the lens would slightly change the geometry. The Offset Microlens array- the lenses just do not point straight up. Lenses are offset to point towards the center, to help with the short flange distance of the M Mount.

Someone on the Leica Users Forum stated that MaxMax was using BG38 for a replacement filter.

I found this at the MaxMax site:


The filter that MaxMax proposes to use is the XniteCC1, and could possibly be BG38. It has about the same transmission in IR as the filter used on the M8. The BG38 is very stable, but has the same problem for IR leakage as the Kyocera B-7 that caused the problems with the Leica M8. The CCD sensor has much more, more than 10x, the sensitivity to IR than does a CMOS sensor. A few years ago I helped out on a project to build a new low-cost IR sensor using COTS detectors- and did a lot of comparisons of available sensors. The answer was a CCD for Near IR. At work I've spent $25K (over 15 years ago) on a Sensors Unlimited sensor with ~320x240 resolution, but it used InGaAS for IR sensitivity.

Somehow Leica+OnSemi decided to use a 0.8mm BG55 glass. This glass was not available in 2008/2009 when the M9 was being developed and introduced. BG55 was reintroduced in 2011. This solved the IR problem, and the repositioning problem. Same Refractive Index, more stable- a Class 2 Schott glass. S8612 is class 3.
Thank you! And I found someone with 5 spare sensors 2 belonging to the M8 and also a “new style” sensor, just working on prices right now. As for the filter I read that Leica used the Bg38 but ran into issues with the refractive index leading to them reflashing the firmware. Everyone I asked that make filters said the BG40 is near identical to the s8612 except it resistant to the corrosion. I’m aware of the micro lens array which is another reason why I went with the 40. At .9mm the same as the original filter it has the same light transmission so it should be a flawless conversion with little to no adjustment needed. The whole point of my repair is to preserve the sensor, kolari seems like they’re making a completely new one
 
The original glass of the filter in the M9 is spec'd at 0.8mm- as per the data sheet. Did the actual filter measure differently? The Data sheet states that the filter has MAR coating and is sealed. The seal could account for some difference.

For a name-
"Sensor Surgeon",
Seeing spots? Worried about "Cataracts" on your M9,
We make it "Leica New".
 
Last edited:
As for the filter I read that Leica used the Bg38 but ran into issues with the refractive index leading to them reflashing the firmware.

BG38 is close to what was used in the Leica M8, is a class 1 glass that is stable over time. It has a lot of IR leakage.

You could make a sandwich using thin layers of BG38 for the inner and outer layers to protect S8612 or another glass that is not as stable. I would be interested in what you read about Leica using BG38. Leica had to revise firmware for the new sensors for exposure- the new filter transmits less light. The color dye in the Bayer filter was also changed.

 
Last edited:
I just want to say I am enjoying this conversation very much and I'm happy that ingenuity has solved the problem of "unfixable" M9 cameras.

I'd been feeling quite annoyed because just earlier this year I was speaking with Leica about upgrading my sensor even though it hadn't shown signs of corrosion yet. I wanted to pre-empt it and get it fixed once and for all. But they said their sensor upgrade program is only for sensors that have started to show signs of corrosion only, therefore my M9P is not eligible to be upgraded. Then when the announcement came that they're stopping the program altogether, well I was very annoyed.
 
I'd been feeling quite annoyed because just earlier this year I was speaking with Leica about upgrading my sensor even though it hadn't shown signs of corrosion yet. I wanted to pre-empt it and get it fixed once and for all. But they said their sensor upgrade program is only for sensors that have started to show signs of corrosion only, therefore my M9P is not eligible to be upgraded. Then when the announcement came that they're stopping the program altogether, well I was very annoyed.
I feel for you. My M Monochrom started showing corrosion within the 5 year period and Leica repaired it for free. It had a couple of black spots on it, was not too bad. No questions asked. My M9 had the tiniest "black Spot" after more than 7 years, I Emailed the picture to Leica New Jersey and got the camera to them ASAP. I did this when the price increase to $1500 was first announced, and they did the repair for me for $950, as the Email request was received before the price increase went into effect. They had the camera back to me within 10 days, Post-Office to my door, no questions asked. No "it was a dust spot" or "it's not corrosion". No "Do you want us to proceed, here's what we found"- just fixed it and sent it back to me. The service included a full CLA and 1-year warranty on the entire camera.

With Jadon and other repair services looking at this problem as an opportunity, I would not count the M9P out for the count. It will take some experimentation, and getting a process down. If I can find another M9 "for cheap" like the one that sold at KEH, I'll pick one up. Better yet- an M Monochrom to do a full-spectrum conversion, If Kodak had made the M9ir that I wanted in 2009, this never would have happened to it. Would have bought 6 of them for work, I had the people lined up to buy them.
 
Last edited:
So yesterday I attached the new filter onto the sensor, I was able to get a few test shots out. Some far my findings are that the new filter has the same light transmission and the focus is spot on (the two things I was mainly concerned about) the only caveat is it seems the color temp is off by about 1200k so with the conversion it looks like you’ll have to give up autowhite balance. Unfortunately the sensor failed after about 10 shots so I’ll be reopening it today to investigate. I have 2 more sensors on the way that I’m planning to convert in the even I run into an M9 with a compromised sensor that can’t be saved. I’ll post some of the test shot later today
 
Bummer about the sensor failing. Part of getting the process down.

If you have the spectral transmission of the new glass versus the old glass I would think that it would be possible to correct the DNG files in software by applying some corrections to the RGB values. I wrote my own DNG processor to read M9 (and M8) DNG files to convert to linear-DNG. I'd be interested in getting a couple of your DNG files.

I have a lot of code to write for work- in the middle of a huge project, but I'd like to take a look.


I could probably spin the above code to process M9 DNG and apply some corrections to the individual RGB files.

It would combine some of the features of this one:
which applies a Gamma curve to the M Monochrom files. I recently ported the Gamma curve code to run under Windows 10.
 
Last edited:
Bummer about the sensor failing. Part of getting the process down.

If you have the spectral transmission of the new glass versus the old glass I would think that it would be possible to correct the DNG files in software by applying some corrections to the RGB values. I wrote my own DNG processor to read M9 (and M8) DNG files to convert to linear-DNG. I'd be interested in getting a couple of your DNG files.

I have a lot of code to write for work- in the middle of a huge project, but I'd like to take a look.


I could probably spin the above code to process M9 DNG and apply some corrections to the individual RGB files.

It would combine some of the features of this one:
which applies a Gamma curve to the M Monochrom files. I recently ported the Gamma curve code to run under Windows 10.
So you’re talking about rewriting the firmware? That would be amazing! As for the sensor I’m not sure if it’s a failure or possibly a dirty connection, in all the excitement I forgot to clean the contacts before reattaching the cable so I’ll the mainboard and give everything a quick cleaning and see if that helps. How do you want me to send over the test shots?
 
Back
Top