Sensor Corrosion Analysis and Fix [Merged]

[rramesh]

Advertisement (gone after registration)

The issue with the M9 is not the sensor. Rather the cover glass.

I’ve been informed by Kolari Vision that they are willing to try a replacement, but they need an M9 with a delaminated sensor. 

maybe someone in the US can work with them. Can send you the contact. 

[haikos]

Do you mean separate cover glass from sensor and install a new cover glass on the same sensor?

I was under the impression that the cover glass was causing the sensor to corrode which is a chemical change; therefore irreversible and permanent.

Haig

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[jdlaing]

7 minutes ago, haikos said:

Do you mean separate cover glass from sensor and install a new cover glass on the same sensor?

I was under the impression that the cover glass was causing the sensor to corrode which is a chemical change; therefore irreversible and permanent.

Haig

The sensor does not corrode. The coating on the cover glass corrodes.

[haikos]

Thanks for clarifying. So a cover glass replacement should do the trick?

[jdlaing]

6 minutes ago, haikos said:

Thanks for clarifying. So a cover glass replacement should do the trick?

Yes.

[rramesh]

Message from Kolari Vision follows. Note that there will be experimentation with risk to determine the best way to remove and replace the sensor cover glass as each sensor type would be different. 

“If you find someone in the US with this issue let me know! I would probably also be willing to just buy a corroded camera outright if someone wants to sell and not take on risks.”

 

Edited September 5, 2020 by rramesh

[dllewellyn]

Advertisement (gone after registration)

I have been modifying cameras for over 25 years.  Among the many cameras I own are three M9's and an S2.  One of the M9's I bought had really bad sensor corrosion, and I decided to take it apart to see what was going on.  What I found was:

- The IR Cut Filter (ICF) also functions as sensor coverglass and is epoxied to the sensor itself making removal rather difficult.  The only sensors I have seen with the ICF epoxied on are some Kodak medium format backs.  Kodak made the sensors for the M9 (and then Kodak sold to Truesense and that got sold to On Semiconductor).

- The active sensor part isn't the problem.  The problem is the ICF.  It looked to me like the glass had oxidized (not quite the same as corroded).

- Certain sorts of filter glass are susceptible to oxidization.  Pretty much all the Schott BG and UG types.

- In every camera I have taken apart (I have taken apart thousands), I have never seen an oxidized ICF because they all have vapor deposition metallic coatings on them which seals the glass.  Oxidization is related to time, temperature, humidity, ozone and even the particular glass melt.

- Next I removed the coverglass (again, rather difficult to do without damaging the sensor) and measured the ICF with a spectrometer.  What I found was that Kodak used *uncoated* BG type glass for the ICF.  Somebody really screwed up somewhere.

- After, I decided to remove the Color Filter Array (CFA) and turn it into a monochrome sensor

- Lastly, I installed a new ICF coverglass

- On Semi exited the CCD business so there will be no more new M9 sensors made.  That's why Leica can't repair M9 cameras with sensor problems anymore.

Without trying to be self promoting, if someone needs their M9 sensor fixed and/or converted to monochrome, I know how.

 

[dummilux]

This is awesome. 

[rramesh]

Please read here. 

 https://kolarivision.com/leica-m9-sensor-corrosion-repair/

[dllewellyn]

You can see an analysis of the problem, the fix and even an monochrome conversion here.

https://www.maxmax.com/maincamerapage/monochrome-cameras/leica-m9

 

[sandymc]

For that those that aren't aware, Dan runs MaxMax, and is the go-to guy for camera conversion work, e.g., "de-bayering". I've worked with a number of cameras that he's converted. There's more information on his work on the Leica M9 here: https://www.maxmax.com/maincamerapage/monochrome-cameras/leica-m9

[mmx_2]

9 hours ago, rramesh said:

Please read here. 

 https://kolarivision.com/leica-m9-sensor-corrosion-repair/

Also got an email from them yesterday that they can repair it :). Really good news!

[jaapv]

It is interesting to note that Kolari got the root cause and reason for the sensor glass trouble right, as opposed to Maxmax, who simply calls it "stupid".

[dllewellyn]

Really?  I analyzed the problem, explained exactly why it happened, it was stupid on someone's part not to coat that type of glass, fixed the problem, converted the camera to monochrome, showed pictures of the problem exactly explaining what happened, analyzed with a spectrometer, and provided example pictures including comparisons of pictures of a stock and converted camera.  Anyone else do that?

[DandA]

Just a question.  Could there have been a possible reason other than a slip-up, mistake or some cost saving reason that Kodak (or the glass manufacturer) decided to use a ICF without vapor deposition metallic coatings? Did it have any optical advantage over those ICF's with vapor deposition metallic coatings (not realizing they might face pitting/oxidation down the road?)  There must of been a reason unless someone simply made a terrible mistake.

Dave (D&A)

Edited September 12, 2020 by DandA

[Denys]

Interesting development: possibility to convert the body to a Monochrom..      ..I assume these repairs also involve replacement of the ‘leather’ cover? Maybe a choice of finish?     Maybe the option of a CLA too?   I can imagine lots of new related threads here soon!

[DandA]

I posed this question to Kolari. They mentioned their replacement M9 cover glass reduces the Leica sensor stack by 0.2mm, further improving on it’s excellent corner sharpness.  I assume when the sensor stack it changed in thickness, the camera's focusing has to be readjusted?  If that's the case and assuming its done, does that in turn mean every M lens one owns also has to be recollimated or adjusted to match the camera's newly adjusted focusing?  If that's the case, it would make more sense to simply have the replacement coverglass being the same thickness as the original and avoid needing the both the camera as well as lenses readjusted for focusing.  Simply trying to get an understanding of all this.

Dave (D&A)

[dante]

Jaap, I find it hard to be sympathetic to Leica on this point. I got to pay $1,800 for a new sensor when the cover glass on my M8 got a scratch (from God knows what - I never had contact cleaned it). Nothing like paying through the nose for a camera and then paying 6x as much for a repair because - like my iMac 5k - everything is glued together. Lack of repairability is a huge design decision, above and beyond what you use as components.

Schott publishes the characteristics of its bandpass  glass, including S8612 (which I understood this M9 filter to be), for moisture resistance. It gets two rainclouds (if you don't wear your reading glasses, they look like umbrellas), which means very sensitive. All rare earth glass "rusts" - that's the irredeemable haze that infects some 1950s Leica lenses. If you've ever seen pinpoint nicks in the front of an old 75/1.4, you've seen it more recently. S8612 is the thinnest but also the most moisture-sensitive IR blocking material, and when I bought one for an IR-capable camera, the person who cut it for me suggested coated stock to slow down the degradation. It's possible that coated stock did not exist, but if it didn't, that begs the question of a design decision to install it in a context where users might be wet cleaning it. 

Of course, the only reason M9 sensor replacement is an issue is because it was not designed to be repairable. So if the cover glass had been more readily replaced, sensor corrosion problems might have gone under the radar, because they would be cheap to fix and users would chalk it up to their own actions. Who would have thought that you would be wet-cleaning a sensor with no ultrasonic dust cleaning? 🙂

SandyMC - concur on Dan's work. Excellent.

DandA - and yes, if the overall refraction of the new cover glass does not match, focus will be off. When you convert mirrorless AF, the camera compensates for this. Not sure on a DLSR whether you'd change the AF fine tuning or the mirror position, but it's adjustable.

[dllewellyn]

The M9 has a cast frame, and then they use shims to calibrate the focus between the casting and the sensor.  Castings shrink during cooling, and each one is slightly different.  So one way that camera manufacturers compensate is to measure the flange back distance and add shims between the sensor and chassis  to compensate.  With a range finder camera, the focus is not set off the image sensor as is in a typical Sony A7 type.  With a Leica rangefinder, focus is already set in the firmware /camera hardware so you can't just put in a thinner ICF/Coverglass because focus will go out.  You would need a flange-back focus measurement setup to do it right.  And then change the shims *precisely*.  Don't screw up your camera.  You are not dealing with a mirrorless camera that adjusts on its own off the sensor (to a point).  You are best off sticking with Leica's design except for (Kodak's ?) massive error.  Sorry, but you don't use that type of glass without a protective coating.  Kodak's image sensor business was on-the-ropes at that point, but, still, somebody had an intern/other person working on the specifications for the coverglass.

[Advertisement]

Hallo,

Look here: Leica M

[jaapv]

3 hours ago, dante said:

Jaap, I find it hard to be sympathetic to Leica on this point. I got to pay $1,800 for a new sensor when the cover glass on my M8 got a scratch (from God knows what - I never had contact cleaned it). Nothing like paying through the nose for a camera and then paying 6x as much for a repair because - like my iMac 5k - everything is glued together. Lack of repairability is a huge design decision, above and beyond what you use as components.

Schott publishes the characteristics of its bandpass  glass, including S8612 (which I understood this M9 filter to be), for moisture resistance. It gets two rainclouds (if you don't wear your reading glasses, they look like umbrellas), which means very sensitive. All rare earth glass "rusts" - that's the irredeemable haze that infects some 1950s Leica lenses. If you've ever seen pinpoint nicks in the front of an old 75/1.4, you've seen it more recently. S8612 is the thinnest but also the most moisture-sensitive IR blocking material, and when I bought one for an IR-capable camera, the person who cut it for me suggested coated stock to slow down the degradation. It's possible that coated stock did not exist, but if it didn't, that begs the question of a design decision to install it in a context where users might be wet cleaning it. 

Of course, the only reason M9 sensor replacement is an issue is because it was not designed to be repairable. So if the cover glass had been more readily replaced, sensor corrosion problems might have gone under the radar, because they would be cheap to fix and users would chalk it up to their own actions. Who would have thought that you would be wet-cleaning a sensor with no ultrasonic dust cleaning? 🙂

SandyMC - concur on Dan's work. Excellent.

DandA - and yes, if the overall refraction of the new cover glass does not match, focus will be off. When you convert mirrorless AF, the camera compensates for this. Not sure on a DLSR whether you'd change the AF fine tuning or the mirror position, but it's adjustable.

Leica's problem was that the corners on the sensor suffer because of the steep incidence angle of some (legacy) lenses, thus they had to use as thin an IR filter as possible. That affected IR sensitivity (the M9 was already much better than the M8, but even the M240 and M10 have similar (mild) IR contamination. )

For that reason they could not use the normal thick, even dual-layer IR filter glass available. They did opt for a  0.5 mm anti-corrosion coated glass that was the only thing available to Kodak at the time. Unfortunately it turned out that the coating did indeed exhibit micro-porosities (which you speculated on). It was not a matter of stupidity, but engineering at the limit of the technical possibilities by necessity, which ran into technical difficulties after a few years of use.