Is the CMOS covered by glass?

Peter Gregg · Sep 2, 2001

I read the review at http://www.imaging-resource.com and unless I am reading it wrong there seems to be a glass cover over the CMOS sensor.

This is really a revelation to me. Does that mean when cleaning the CMOS, it is the glass that gets cleaned and the sensor itself never can get touched? To me this is a big relief if it is true.

Can anyone set me straight on this?

Thanks,
Pete
Peter Gregg - Miami Florida

about half way down the page at:
http://www.imaging-resource.com/PRODS/D30/D30A3.HTM

or for a closer view without the explaination given:
http://www.imaging-resource.com/PRODS/D30/ZBOXBLG.HTM

Mark A. Lamb · Sep 2, 2001

That is the way I read it and see it from that photo also........... Big relief here also!

Mark

Peter Gregg said:
I read the review at http://www.imaging-resource.com and unless I am
reading it wrong there seems to be a glass cover over the CMOS
sensor.

This is really a revelation to me. Does that mean when cleaning the
CMOS, it is the glass that gets cleaned and the sensor itself never
can get touched? To me this is a big relief if it is true.

Can anyone set me straight on this?

Thanks,
Pete
Peter Gregg - Miami Florida

about half way down the page at:
http://www.imaging-resource.com/PRODS/D30/D30A3.HTM

or for a closer view without the explaination given:
http://www.imaging-resource.com/PRODS/D30/ZBOXBLG.HTM

Peter Galbavy · Sep 2, 2001

Peter Gregg said:
I read the review at http://www.imaging-resource.com and unless I am
reading it wrong there seems to be a glass cover over the CMOS
sensor.

This is really a revelation to me. Does that mean when cleaning the
CMOS, it is the glass that gets cleaned and the sensor itself never
can get touched? To me this is a big relief if it is true.

Well, all the CCD sensors I know of have an IR filter mounted on the front. I am not sure if CMOS has the same IR / UV characteristics, but I expect that Canon would put something on the sensor, even if just for protection.

Cannot say for certain though

Peter

Z (is real) · Sep 2, 2001

When the mirror is locked up for cleaning, what is exposed is a piece of glass that is a noticeable distance in front of the sensor. (Quarter to half an inch. Something like that.) Presumeably this is an IR/antialiasing filter, though I do not know. It is however what most folks are cleaning when they do their own work on the camera.

This is perhaps a relief right up until you have a spot that you can't get rid of no matter what and you realize its almost certainly on the back side of the glass or on the sensor itself. In which case you'll have to send the camera back to Canon or get more adventurous in your disassembly of the camera.

-Z-

Chris Maddock · Sep 3, 2001

Z (is real) said:
When the mirror is locked up for cleaning, what is exposed is a
piece of glass that is a noticeable distance in front of the
sensor. (Quarter to half an inch. Something like that.) Presumeably
this is an IR/antialiasing filter, though I do not know. It is
however what most folks are cleaning when they do their own work on
the camera.

This is perhaps a relief right up until you have a spot that you
can't get rid of no matter what and you realize its almost
certainly on the back side of the glass or on the sensor itself. In
which case you'll have to send the camera back to Canon or get more
adventurous in your disassembly of the camera.

If the muck is on the sensor itself, then it will leave a sharp blemish on an image at all apertures, whereas if it's on the filter/glass/whatever, it is relatively hard to spot on an image unless the lens is stopped well down, say f11 or smaller.

I would assume that something on the back of the filter would acuse blemishes on the image falling somewhere between these two.

KRs
Chris

Z (is real) · Sep 3, 2001

Chris Maddock said:
If the muck is on the sensor itself, then it will leave a sharp
blemish on an image at all apertures, whereas if it's on the
filter/glass/whatever, it is relatively hard to spot on an image
unless the lens is stopped well down, say f11 or smaller.
I would assume that something on the back of the filter would acuse
blemishes on the image falling somewhere between these two.

In my case, I'm fairly certain its on the back of the glass plate. The first cleaning attempt didn't get rid of the spot and generally made things worse. The second one made left a few small artificats but was overall better. It still didn't touch the damn spot though. Its definitely in the f8 to f11 range. (The glass is of course quite thin. I doubt it affects the visibilty vs. aperture much.) By the way, I judge "cleanliness" by shooting a white wall and then running auto levels.

I'm also guessing that this spot has been there since I bought the camera. I first noticed it then and thought it was a lens problem. So I cleaned the one lens I owned then and it went away. Well, no it didn't. Its just that the test shots I took were at fast aperture... I've since noticed it off and on. You'll get the gist here that I'm not driving myself batty over dust spots. Occasionally I have to retouch one out because it is visually detracting, but generally I don't have problems. (I take pains to keep my optics clean of course.)

I also need the focus screen cleaned and the auto power off fix done. So Canon can deal with it...

-Z-

Peter Gregg · Sep 3, 2001

Why didn't they seal that area up?

Maybe that wasn't possible, but it seems like if they were going to seal anything that would be a great spot to seal, then we could be left to deal with the glass plate and the sensor would be safely tucked into a sealed area.

Pete

Al Pacheco · Sep 3, 2001

Yes, Pete, the actual sensor is sealed behind a piece of optical glass. However, that glass is cemented to the CMAS array and may as well be considered a part of it. It's unlikely that any dirt will get behind it.

The other glass filter is placeddirectly behind the shutter curtains and is what has to be cleaned now and then.

Here's a photo that shows one of the first CMOS sensors of the type Canon uses. It was developed about ten years ago for NASA/JPL. I work for the company that developed the technology of placing analog circuitry next to each sensor on the array. Canon incorporates quire a few of the methods that were developed by us during the last decade.

Peter Gregg said:
Why didn't they seal that area up?

Maybe that wasn't possible, but it seems like if they were going to
seal anything that would be a great spot to seal, then we could be
left to deal with the glass plate and the sensor would be safely
tucked into a sealed area.

Pete

Z (is real) · Sep 3, 2001

Peter Gregg said:
Why didn't they seal that area up?

I don't know. The glass will slide back and forth a little bit and does not strike me as being sealed in any airtight fashion. This is one reason compressed air might not be a good idea as it probably increases the chance of blowing dust past the glass filter into the sensor cavity. That said, I have no evidence that its easy to get dust behind it. For all I know, the spot I have happened during manufacturing. (Not to dis Canon. I don't knwo how it got there and they're going to fix it for free. I've been using the camera for 9 or 10 months without being totally annoyed or anything...)

I guess my overall attitude is I wish it were easier to do all the cleaning myself, but dust is a fact of life and I can deal with it...

-Z-

Phil Askey · Sep 3, 2001

Yes you're right, and to be honest although I'd thought about this I'd never really considered its significance.

What is significant is the THICKNESS of the glass and its distance from the sensor (and thus the lesser effect of dirt on it appearing in images).

See this diagram:

http://www.dpreview.com/reviews/CanonD30/Images/seethroughdiagram.jpg

You'll note that the piece of glass in question (cyan coloured on the diagram) does indeed appear to be at least a quarter of an inch away from the sensor / low-pass filter and by the metal surround the implication is that it completely seals the sensor from the outside world.

This would go a long way to explaining why the D30 has far less of a dust problem than other D-SLRs.

The big question is why haven't Nikon sealed their sensor from the outside world?

Peter Gregg said:
I read the review at http://www.imaging-resource.com and unless I am
reading it wrong there seems to be a glass cover over the CMOS
sensor.

This is really a revelation to me. Does that mean when cleaning the
CMOS, it is the glass that gets cleaned and the sensor itself never
can get touched? To me this is a big relief if it is true.

Can anyone set me straight on this?

Thanks,
Pete
Peter Gregg - Miami Florida

about half way down the page at:
http://www.imaging-resource.com/PRODS/D30/D30A3.HTM

or for a closer view without the explaination given:
http://www.imaging-resource.com/PRODS/D30/ZBOXBLG.HTM

Pavel Chernev58938 · Sep 3, 2001

I also think that there are two glass elements - the window of the CMOS sensor chip and the filter. I did some "experiments" with several D30 bodies and found out that dust contamination can be two types - on the filter and on the sensor window. Filter dust is the more likely case (changing lenses!), produces a rather de-focused image and is cleanable (with low pressure filtered air in clean environment). Sensor port dust results in much sharper particle image and is not cleanable (without disassembling the camera).

Phil Askey said:
What is significant is the THICKNESS of the glass and its distance
from the sensor (and thus the lesser effect of dirt on it appearing
in images).

See this diagram:

http://www.dpreview.com/reviews/CanonD30/Images/seethroughdiagram.jpg

You'll note that the piece of glass in question (cyan coloured on
the diagram) does indeed appear to be at least a quarter of an inch
away from the sensor / low-pass filter and by the metal surround
the implication is that it completely seals the sensor from the
outside world.

This would go a long way to explaining why the D30 has far less of
a dust problem than other D-SLRs.

The big question is why haven't Nikon sealed their sensor from the
outside world?

Peter Gregg said:

I read the review at http://www.imaging-resource.com and unless I am
reading it wrong there seems to be a glass cover over the CMOS
sensor.

This is really a revelation to me. Does that mean when cleaning the
CMOS, it is the glass that gets cleaned and the sensor itself never
can get touched? To me this is a big relief if it is true.

Can anyone set me straight on this?

Thanks,
Pete
Peter Gregg - Miami Florida

about half way down the page at:
http://www.imaging-resource.com/PRODS/D30/D30A3.HTM

or for a closer view without the explaination given:
http://www.imaging-resource.com/PRODS/D30/ZBOXBLG.HTM

Click to expand...

Al Pacheco · Sep 3, 2001

But you shouldn't have the shutter open when changing the lens so how does any dust that perhaps gets into the mirror box migrate to the filter's surface?

Is the action of the mirror slap and curtain opening creating some sort of mini-turbulance?
Or is the dust in place at the point of assembly?

Pavel said:
Phil Askey said:

What is significant is the THICKNESS of the glass and its distance
from the sensor (and thus the lesser effect of dirt on it appearing
in images).

See this diagram:

http://www.dpreview.com/reviews/CanonD30/Images/seethroughdiagram.jpg

You'll note that the piece of glass in question (cyan coloured on
the diagram) does indeed appear to be at least a quarter of an inch
away from the sensor / low-pass filter and by the metal surround
the implication is that it completely seals the sensor from the
outside world.

This would go a long way to explaining why the D30 has far less of
a dust problem than other D-SLRs.

The big question is why haven't Nikon sealed their sensor from the
outside world?

Peter Gregg said:

I read the review at http://www.imaging-resource.com and unless I am
reading it wrong there seems to be a glass cover over the CMOS
sensor.

This is really a revelation to me. Does that mean when cleaning the
CMOS, it is the glass that gets cleaned and the sensor itself never
can get touched? To me this is a big relief if it is true.

Can anyone set me straight on this?

Thanks,
Pete
Peter Gregg - Miami Florida

about half way down the page at:
http://www.imaging-resource.com/PRODS/D30/D30A3.HTM

or for a closer view without the explaination given:
http://www.imaging-resource.com/PRODS/D30/ZBOXBLG.HTM

Click to expand...

Click to expand...

Peter Gregg · Sep 3, 2001

After I posted the question as to why not just seal the CMOS sensor that is behind the glass, I remember from way in the back of my head writings that refered quite a few times to the sensors not liking heat. I wonder if that is not one of the reasons they just didn't seal it up.

If heat really is the reason, then I think in the future that can be solved somehow with a heat sink on the backside or some kind of a heat sink system.

If it isn't the reason, then oh well, I hope some engineer thinks of sealing that sucker up sooner than later

Pete

Martyn38590 · Sep 4, 2001

I guess you've had your fears allayed now, but here's more soothing words ;o)

As far as I recall from my semiconductor physics course, any semiconductor device must be sealed away from oxygen rich atmospheres, otherwise the oxide tracks fur up and the whole thing goes out of spec and ultimately fails. Most chips are sealed in plastic, or with a hermetically sealed metal lid.

With CCD and CMOS imagers the metal lid is generally replaced with an optical quality glass lid - often silica glass IIRC. I believe you can seal using a (relatively) thick passivating layer of a glassy substance directly onto the chip, but this is more expensive than plopping a sheet of glass over the carrier well and makes it harder to attach the leads to the chip.

Basically until someone perfects a flip chip sensor where the incoming light goes through the substrate of the chip rather than in through the top, our sensors will be covered with a handy sheet of glass. I doubt flip chip devices will be introduced into anything where a user is likely to prod about with anything dangerous like a blower brush or canned air. Itll likely stay in the field of astronomy where the last little photon counts.

That's probably more info than anyone needed, but I thought I'd explain why things are as they are, just for interests sake

Enjoy
Martyn

Mirar · Sep 4, 2001

I find it very interesting. Thank you.

Martyn said:
I guess you've had your fears allayed now, but here's more soothing
words ;o)

As far as I recall from my semiconductor physics course, any
semiconductor device must be sealed away from oxygen rich
atmospheres, otherwise the oxide tracks fur up and the whole thing
goes out of spec and ultimately fails. Most chips are sealed in
plastic, or with a hermetically sealed metal lid.

With CCD and CMOS imagers the metal lid is generally replaced with
an optical quality glass lid - often silica glass IIRC. I believe
you can seal using a (relatively) thick passivating layer of a
glassy substance directly onto the chip, but this is more expensive
than plopping a sheet of glass over the carrier well and makes it
harder to attach the leads to the chip.

Basically until someone perfects a flip chip sensor where the
incoming light goes through the substrate of the chip rather than
in through the top, our sensors will be covered with a handy sheet
of glass. I doubt flip chip devices will be introduced into
anything where a user is likely to prod about with anything
dangerous like a blower brush or canned air. Itll likely stay in
the field of astronomy where the last little photon counts.

That's probably more info than anyone needed, but I thought I'd
explain why things are as they are, just for interests sake

Enjoy
Martyn

Pavel Chernev58938 · Sep 4, 2001

My guess is that dust enters the mirror box when changing lenses and then migrates to the surface of the sensor. The flapping mirror probably creates enough air mass movement. I managed to get the sensor of my D30 dirty only once (after about 15 lens changes in a very dusty environment on a trip). There were 2 specs of dust on the filter surface, while the mirror chamber was really dirty.

Allen Pacheco said:
Is the action of the mirror slap and curtain opening creating some
sort of mini-turbulance?
Or is the dust in place at the point of assembly?

Pavel said:

Phil Askey said:

What is significant is the THICKNESS of the glass and its distance
from the sensor (and thus the lesser effect of dirt on it appearing
in images).

See this diagram:

http://www.dpreview.com/reviews/CanonD30/Images/seethroughdiagram.jpg

You'll note that the piece of glass in question (cyan coloured on
the diagram) does indeed appear to be at least a quarter of an inch
away from the sensor / low-pass filter and by the metal surround
the implication is that it completely seals the sensor from the
outside world.

This would go a long way to explaining why the D30 has far less of
a dust problem than other D-SLRs.

The big question is why haven't Nikon sealed their sensor from the
outside world?

Peter Gregg said:

I read the review at http://www.imaging-resource.com and unless I am
reading it wrong there seems to be a glass cover over the CMOS
sensor.

This is really a revelation to me. Does that mean when cleaning the
CMOS, it is the glass that gets cleaned and the sensor itself never
can get touched? To me this is a big relief if it is true.

Can anyone set me straight on this?

Thanks,
Pete
Peter Gregg - Miami Florida

about half way down the page at:
http://www.imaging-resource.com/PRODS/D30/D30A3.HTM

or for a closer view without the explaination given:
http://www.imaging-resource.com/PRODS/D30/ZBOXBLG.HTM

Click to expand...

Click to expand...

Click to expand...

Search

Is the CMOS covered by glass?

Peter Gregg

Veteran Member

Mark A. Lamb

Active member

Peter Galbavy

Senior Member

Z (is real)

Veteran Member

Chris Maddock

Senior Member

Z (is real)

Veteran Member

Peter Gregg

Veteran Member

Al Pacheco

Guest

Z (is real)

Veteran Member

Phil Askey

Guest

Pavel Chernev58938

Member

Al Pacheco

Guest

Peter Gregg

Veteran Member

Martyn38590

Forum Enthusiast

Mirar

Well-known member

Pavel Chernev58938

Member

Similar threads

Keyboard shortcuts