Essentially the Only One

Instead of sneaking off the library to take a nap on one of the many soft couches (and doubtless filling the library with the reveberant snores that Allan has experienced elsewhere), I decided to take the life of my digital camera in my hands and clean the CCD.

Why would I want to do such a potentially destructive thing to this sliver of vulnerable silicon buried deep in the body of my Canon Digital Rebel XT SLR?

The short answer is dust. It's impossible to keep dust out the thing (indeed many are shipped brand new with a few dust particles already sitting on the sensor). Normally, you'll never notice it. However, should you take a photograph of a light block color - e.g. the blue sky - with a very small aperture size (f/22 or smaller), you might see a lot of little black slightly smudged dots. This is dust that has settled directly on the surface of the chip.


Now, one of the advantages of working in a research laboratory is that there are all sorts of useful reagents and equipment lying around. Highly purified methanol. Sterile swabs. Air and vacuum outlets. PVC and rubber tubing.

A cursory Google search of the many photography sites revealed methanol to be a solvent of choice. It has no corrosive effect on the chip, and attracts dust like a magnet. Nonetheless, at first I was wary of applying anything wet to the CCD, so I devised a blower and vacuum system to see how that would work.

This was simple. I hooked up two pieces of PVC tubing to the AIR and VACUUM outlets in the chemical hood (making a trap out of a flask with sidearm for the vacuum to guard against blowback and attaching a syringe filter to the air jet to prevent any highspeed particulate matter scouring my CCD). Then I opened up the camera, exposed the CCD using the 'sensor cleaning' option in the camera's own firmware program, and alternately blew and sucked air in and out close to the chip.

Too close as it happens. I looked down and saw to my horror that I had touched the surface with the tubing from the vacuum and left a smudge mark.

Naturally, I considered that I must have broken it. However a few snaps revealed no damage. The only result was the appearance of a very slight (much less than with dust) darkening in the image under similar small aperture, blue sky conditions, as described above.

Nonetheless, the appearance of this smudge pained me. In addition, the blowing and vacuuming had removed some, but by no means all, of the dust particles. So I decided it was time to go for the methanol.

This was the most difficult part of the whole process. I used reagent grade methanol, so I felt pretty sure it was going to be pure enough not to leave any residue, but the application was tricky. Fortunately we had some long stem sterile swabs in the lab that are relatively easy to manipulate. With great care, I brushed the smudged area with the methanol-soaked tip. To my great satisfaction, the smudge was removed and the CCD assumed its pristine former self.

I could have given the CCD a more complete cleaning with the methanol, but as I had had enough heart stopping moments for a Friday afternoon, I decided to leave it there.
Nonetheless, as you can see, while it is not yet perfect, it is greatly improved.