3

If a recent model Pi Camera faces the sky for months or years, is there anything known about damage to the sensor from long term exposure sunlight?

Of course I know this might not be advised, but I'm asking about knowledge, experience or data for this camera, either with the standard lens, or a fisheye.

Assume the standard color camera (e.g. camera module V2) is viewing through some glass or plastic window for environmental protection. Since exposures can be long (no fast movement needs to be recorded) I can add a neutral density filter to cut down on the intensity, but I'd like to know if there is any existing information on this so I don't have to burn up a bunch of cameras first.

Consider the application of a dispersed group of sky cameras for tracking meteors as a way to estimate where the meteorite or fragments may have landed by analizing images from several locations. The cameras may be left facing skyward to operate autonomously for extended periods of time in weatherproof housings such as shown in Instructables.

Somewhat related: Are there industry standards or specs for image sensor resistance to damage from intense light?

Images from the Instructables project Wireless All Sky Camera:

enter image description here

enter image description here

Improve this question

3 Answers 3

Reset to default
1

I’ve used a lens from quality sunglasses over the lens of an outdoor installation, and after 3-4 years it has suffered no noticeable loss in picture quality (detail, contrast, colours). So, as the above answer suggests, a rudimentary filtering out of the useless UV would be helpful. Have a look at outdoor CCTV cameras for industry use (factory perimeter e.g.). The polarized glass is one thing, but these cameras stay online for a decade in all weather.

It’s just a 640x480 streaming ethernet (I think I used wget+curl to grab jpgs, and perhaps vlc to grab the stream) camera with manual focus, so installing the overlay (the sunglass lens) was trivial. It faces direct sunlight max 5 hours per day. There has been no noticeable degradation in the images. I have not made a baseline with an unfiltered lens. Google UV lenses for CCTV cameras; your wide lens needs something bigger.

Trivia: the PV / solar panel crowd have done a lot of interesting research into what diminishes the effect of their panels... I would test a filter over a 5V cell, and see the current with a multimeter across a low-ohm power resistor. Anything that gives high transparency, but very low current - compared to the unfiltered baseline - would probably be a good filter. Glass with a metal component alone diminishes effect a lot.

Also know that professional installations often utilize a small heating element (a power resistor would do, or some nichrome wound around a ceramic element) to ‘push’ out dampness and condensation. It would be prudent to use a thermocoupler (MAX6675 on I2C is ok, and low-power) to turn that heating element on/off w. a MOSFET (IRF3202 is very good, but perhaps overkill, needs 10V to gate for full on, NTD4906N is logic level and a better choice), or a DS18B20 (onewire?) A DHT22 (I2C) would let your system know both temperature and humidity, and then trip heating element based on ambient humidity. Essentially, humidity is dissipated if internal temperature exceeds external temperature. I heard this from an electrician who installs camera systems. To further protect electronics, they can be coated in nail polish (Great scott made a youtube video about this), or embedded in epoxy.

Brought to you by the Awful Hacks dept.

Improve this answer
4
  • 1
    Yes, for max ~5 hours.
    – user2497
    Jan 25, 2018 at 12:28
  • Thanks for your answer; it's great to hear about some actual experience with this!
    – uhoh
    Jan 25, 2018 at 12:30
  • 1
    @uhoh Added some gratuitous info. Please post some pictures of your finished project (with schematic), I am interested.
    – user2497
    Jan 25, 2018 at 12:47
  • OK I will, if it doesn't look too awful!
    – uhoh
    Jan 25, 2018 at 12:50
4

As the author of this Instructables, I thought I would contribute to this thread. After more than 2 years outside in the harsh conditions of the Yukon winters and endless late-spring/early-fall daylight, the sensor doesn't appear to be suffering from a lack of sensitivity (videos here).

The camera used in this project is an ASI224MC. It is equipped with an piece of glass to protect the sensor from dust and humidity. According to the transmittance chart, it blocks UV light below 350nm and somewhat lets less energetic UV light through (350 to 400nm).

On top of that the Arecont lens is made of optical glass which has the same properties as the protection glass mentioned earlier (blocks UV below 350nm).

Then there is the Acrylic dome which blocks the same portion of the UV light as optical glass.

To sum it up, in theory the sensor in this project is protected from UVB and receives a small amount of UVA close to blue light wavelength.

In summer however, the camera is covered because there is no darkness from June to August so I don't have any data for this extreme sun exposure.

By the way, meteor detection is a the top of my TODO list of software updates.

Improve this answer
2
  • Thank you for your reply! It seems that there are no deleterious effects caused by heating of the image sensor by the focused spot of visible sunlight slowly moving across the sensor's surface either?
    – uhoh
    Jan 25, 2018 at 21:33
  • 1
    I haven't noticed any effects on the sensor. I have a dark subtraction feature in that project and darks are uniform across the field of view even when stretching the levels. I have about 3 dead pixels after 2 years but that's quite normal on a camera after 2 years of use. A DSLR shows the same issue after years of use. And they're not even on the path of the sun.
    – Thomas Jacquin
    Jan 27, 2018 at 0:34
2

Just like any other CCD device, the camera module is subject to radiation stress. A good rule of thumb is to filter out radiation which is not intended to be captured, and to limit all exposure to intense radiation of any kind.

Further Reading:

Radiation Hardness of CCD Sensors

Improve this answer
5
  • Thanks for the link. I'm pretty sure the types of low to medium energy radiation discussed in the link (that would be related to the Sun) is stopped by the Earth's atmosphere. Even in space, much of it would be stopped by the lens. I'm really interested in damage caused by the Suns light concentrated on to a few pixels by the lens. This is likely to cause heating, especially in the color filters which absorb roughly 2/3 of the light hitting them and convert it to heat, as well as the small amount of near UV hurting all polymers, e.g. microlense arrays above the color filters and IR filter.
    – uhoh
    Dec 8, 2017 at 23:25
  • @uhoh Given that sunlight damages plastic, even through dual layer glass windows, don’t arrive so quickly at this conclusion. Ask the vendor or manufacturer.
    – user2497
    Jan 25, 2018 at 9:55
  • @user2497 Radiation hardening of the CCD sensor will not mitigate softening/melting the microlens array or color filters on top of it. These are two different things.
    – uhoh
    Jan 25, 2018 at 10:18
  • 1
    @uhoh But it IS radiation that causes it. Anyway, I’ve used a lens from quality sunglasses over the lens of an outdoor installation, and after 3-4 years it has suffered no noticeable loss of picture quality. So, as the above answer suggests, a rudimentary filtering out of the useless UV would be helpful. Addition: Have a look at outdoor CCTV cameras for industry use (factory perimeter e.g.). The polarized glass is one thing, but these cameras stay online for a decade in all weather.
    – user2497
    Jan 25, 2018 at 10:40
  • 1
    @uhoh It’s just a 640x480 streaming box with manual focus, so installing the overlay was trivial. It faces direct sunlight max 5 hours per day. There has been no noticeable loss of quality in the images. I have not made a baseline with an unfiltered lens. Google UV lenses for CCTV cameras; your wide lens needs something bigger.
    – user2497
    Jan 25, 2018 at 10:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.