1
\$\begingroup\$

I know that sun like temperatures kill electronics, but what about 55°C if it's constant?

I have a small commercial web camera that is officially rated 0°C to 70°C. It has no jelly capacitors. I intend to heat it to a very constant 55°C so that it's operating characteristics never fluctuate whatever the room temperate. Clearly the device doesn't explode at exactly 70°C, so there must be some form of curve around that threshold.

Can this adversely affect the camera? Is there some form of general life de-rating for all components, not just jelly capacitors?


Are devices damaged by high current, or high temperature? amd What exactly gets "worn out" and damaged by heat? exist but don't quite answer this one.

\$\endgroup\$
7
  • 1
    \$\begingroup\$ Sorry, by jelly capacitors do you mean aluminum electrolytics? I've just never heard the term before. \$\endgroup\$
    – Hearth
    Commented Nov 3, 2018 at 0:18
  • \$\begingroup\$ @Felthry Yeh, the ones with the wet electrolyte that dries out. \$\endgroup\$
    – Paul Uszak
    Commented Nov 3, 2018 at 0:31
  • \$\begingroup\$ You are not going to get a satisfactory answer. But in general, chemical reactions of all kinds proceed faster at higher temperatures. So running at 55C will likely speed up failure. But estimating the exact effect is likely impossible. Someone would have to run batches at both temperatures to measure the extent of the effect. In general, I would probably try to run it at a slightly lower temp like 40 or 45. But it is your choice. \$\endgroup\$
    – user57037
    Commented Nov 3, 2018 at 4:02
  • 2
    \$\begingroup\$ Interesting. I won't tell a soul. A friend of mine is into astrophotography, so he cools his sensor. There is a lot to it, some of which probably applies to you as well. For example, if you blackout the lens completely and then take a long exposure, there will be a phantom image. This is just pixel bias. But it has be subtracted from the actual image. In your case, that phantom image might make your noise less cryptographically sound, so you may need to remove it, too. \$\endgroup\$
    – user57037
    Commented Nov 3, 2018 at 4:29
  • 1
    \$\begingroup\$ If you can, open up the camera. If you see some PLANES --- VDD, GROUND ---- then you likely have good heat flows. If you can, open up the camera and let air flow thru, vertically. \$\endgroup\$ Commented Nov 3, 2018 at 5:27

4 Answers 4

2
\$\begingroup\$

The general working assumption is that life halves for every 10 degrees C rise, so you could guess the lifetime will be only 25% of the life at 25 degrees C ambient.

From your proposed operating temperature to the maximum permissible is only 15 degrees so it’s hard to actually do an accelerated test. You could try pushing the rated limits in a destructive test, but you may get an unnecessarily pessimistic number (many parts will work with relaxed specs at very high temperatures). It’s also possible to get an optimistic answer, particularly if it temporarily stops functioning, but that’s less likely.

My guess is that it would be okay for some years, but that and $2 will get you a cup of coffee. A better number could come from analyzing each component and applying reliability calculations. The LSI chip and any ceramic or tantalum capacitors are probably the weak links, and any stressed power supply components such as linear or switching regulators.

Electrolytic capacitors often have an operating life of only 2000 or 5000 hours at rated temperature (usually 105 or 85 degrees C), so they are a typical weak link, so it’s good your camera has none.

\$\endgroup\$
2
  • \$\begingroup\$ the half every 10C isn't accurate. its a misapplication of a specific statistical observation. It is better than that \$\endgroup\$
    – user16222
    Commented Nov 3, 2018 at 11:17
  • \$\begingroup\$ ill get the info later (on my work machine) \$\endgroup\$
    – user16222
    Commented Nov 3, 2018 at 11:23
1
\$\begingroup\$

This is a very complicated problem that has many caveats and qualified answers. This is just a general overview copied from my comments.

As a rule of thumb heat will always decrease the lifetime of the device, however this may still be longer than the expected (warranted) lifetime.

All electronic components will fail, eventually, the question is amount of time and method of failure (Quiet fail safe, or dramatic "heat event"). Your goal as a designer/integrator is to identify the effective lifetime and acceptable failure modes, then work to qualify the device or design workarounds.

In the worst case, if there is no suitable alternative a workaround may be active cooling.

Your best bet is to purchase a few samples and do some tests at 50C and perhaps some accelerated tests at 70C in order to determine if it meets the performance and lifetime characteristics you need.

Designing this type of test is beyond the scope of this Q&A, but there are commercial facilities that will rent you time on a thermal test chamber and expert advice, or you can try something yourself.

Make sure to place temperature probes on certain parts that tend to be hotter than ambient (e.g. power regulators and filter caps) in order to make sure you are not exceeding their rating.

Additionally Note that for ceramic capacitors, the class of dielectric matters. Class 2 (e.g X7R) which happens to be the most common type (Best Price vs Size vs Rating) , have a large temperature dependence. Class 1 (e.g. NP0 or C0G) have almost no temperature dependence. These relationships may impact the operation of the device even if it does not fail or significantly affect lifetime

\$\endgroup\$
1
\$\begingroup\$

Depends... Running electronics at temperature does shorten their life but equally will damage them if taken beyond their operating temperature.

Take a Linear regulator, typically they have a maximum junction temperature of 150C. Now a designer may take "credit for" the operating environment being that of 40C which produces a much larger temperature differential to help keep the junction cool. Initial testing may have shown no additional cooling (heatsink etc...) was needed for the expected use cases.

If you then operate at a higher ambient, an ambient outside of the stated operating range, then there may be permanent damage.

My advice, if this is what you want/need to do, open the device up and use a thermal camera to see how hot parts get. By some simple maths you would get a rough estimate as to how hot the case of some devices would get. Do NOT let anything exceed 100C

\$\endgroup\$
-2
\$\begingroup\$

I have a small commercial web camera that is officially rated 0°C to 70°C.

I pretty much guess it isn't rated at all, as this is just the specification of parts in the consumer-grade electronics range. Read as: Good at room temperature, operating in a car may kill it either way.

The problem you face is uneven heat distribution. With enough temperature headroom, this is a non-issue. A device built from consumer-grade components is unlikely to fail up to an ambient temperature of 40°C. Above that, it depends on its own heat production and mounting position.

Do you know those figures for your small commercial web camera?

You cannot heal that problem by controlling the ambient temperature, regardless if 40°C or 55°C. It's about getting the internally generated heat out of that casing quickly. If you designed the thing yourself, you would add generously sized heat sinks and blowers to assist that.

\$\endgroup\$
4
  • \$\begingroup\$ Err, you've kinda missed the point. It's not you, it's me. I specifically want to run the camera at 55 degrees. Will that damage it? \$\endgroup\$
    – Paul Uszak
    Commented Nov 3, 2018 at 1:21
  • \$\begingroup\$ No, you are missing the point. How do you control the parts in the camera are at 55°C, if you only control the ambient temperature? The camera has internal heat sources which cannot be neglected any more at higher ambient temperatures. \$\endgroup\$
    – Janka
    Commented Nov 3, 2018 at 1:22
  • \$\begingroup\$ Let's just say that I've learnt from bitter, bitter, bitter experience that expanding this type of question is counter productive. Can we just please assume that the ambient temperature immediately around the device is actively maintained at 55? Some internal component may exceed this, but the device is only 1" in principle dimension and it's 3.3V. It's unlikely to exceed 70. Is that a problem? \$\endgroup\$
    – Paul Uszak
    Commented Nov 3, 2018 at 1:39
  • \$\begingroup\$ Please just ignore my answer then. \$\endgroup\$
    – Janka
    Commented Nov 3, 2018 at 1:40

Your Answer

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

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