Apart from condensation why do electronic components usually have a low temperature limit? For example my laptop says something along the lines of -10 °C to 75 °C temperature while in use.

I can understand the high temperature limit, as things will probably melt!

But why is cold such a bad thing?

Apart from batteries, which components will extreme cold damage, and how?

Will using it increase the damage?

Will using the equipment offset this damage (as they warm up from use)?

Also, I am talking about extreme temperatures below -50 °C, so is condensation still a problem?

Note: I am not storing it so it is not a duplicate of another question.

Note 2: I am not talking about semiconductors, but generally speaking.


11 Answers 11


I once designed an amplifier that would oscillate at -10°C. I fixed it by changing the design to add more phase margin. In this case, the oscillation did not cause any damage, but the circuit did not work well in this condition, and it caused errors. These errors went away at higher temperatures.

Some plastics crack when they freeze. Dry ice is -78.5°C, and I have broken a lot of plastic with dry ice. For example, I destroyed a perfectly good ice chest that cracked into little pieces in the spot where I had a chunk of dry ice in it.

In surface-mount designs, the differential temperature coefficient of expansion between the parts-soldered-to-the-circuit-board and the circuit board can cause large stresses. The stress-strain-temperature relationship often barely works over the specified temperature range. When the equipment is powered up, the hot components can change shape and break the brittle plastic, much like my old ice chest.

If the equipment is below 0°C and then you take it into a nice warm, humid office, water will condense on the circuit boards and can cause problems. Presumably a similar thing can happen with frost, depending on the weather. When the frost melts, there can be problems.

When I receive equipment in the morning that has been carried as air cargo, I assume that it has recently been very cold, and I let it sit around for some hours to warm up slowly and to stay dry before opening the box in the office.

Turning on very cold gear could be interesting. Some current-limiting components, such as a PTC or PPTC, will pass a lot more current.

The lubricants in motors such as fans and disk drives could also be a problem.

  • 4
    \$\begingroup\$ "If the equipment is below 0degC and then you take it into a nice warm, humid office, water will condense on the circuit boards and can cause problems." - condensation can happen every time a surface is below ambient and humidity is high enough. I doesn't have to be below 0degC (if it were water would condense, then freeze) \$\endgroup\$
    – RJR
    Commented Dec 16, 2014 at 4:37
  • 1
    \$\begingroup\$ How about purely electronic components? I mean if I used a different substrate for my board, any reason why my calculator shouldn't run at say -100'c ? Only thing I can think of is resistors cracking as they build up heat. \$\endgroup\$
    – Sled
    Commented Dec 16, 2014 at 16:15
  • \$\begingroup\$ I had a similar thing happen to me this summer at a hotel near the beach. I took my laptop from the air conditioned hotel room to work outside in the very humid air. My screen had noticeable condensation on it. \$\endgroup\$
    – OSE
    Commented Dec 16, 2014 at 16:26
  • 4
    \$\begingroup\$ The temperature of a surface needs to be (roughly) below the dew point for condensation to form. \$\endgroup\$
    – OSE
    Commented Dec 16, 2014 at 16:28

I can give you an answer because I had been one of those who either wrote or verified the specs of semiconductor ICs.

Legally and ethically speaking, I could only sign off on the parameters within which we have verified the IC/processor would work. And then my boss, and her/his boss, and everyone else would see the evidence of the tests, and they too would sign off on those constraints.

I could not ethically or legally sign off that a batch of processors would work at -100 C, if I had not put them thro the suite of tests at -100 C.

If you chose to use your equipment at -50 C, equipped with the processor I signed off on with a low threshold of -15 C, my company would no longer have any obligation to that processor. You have broken the warranty.

Testing at -50 C is a lot more expensive to do than testing at -15 C. I would have to verify the test site is actually -50 C--. It is also very dangerous.

Besides that, special/hermetic packaging is required for ICs to operate at extreme low temperatures. As an extreme example, plastic packaging could develop cracks or structural compromises when we pour liquid nitrogen onto them.

Differential expansion between the die and the packaging could tear the die away from its site of attachment, or crack the die.

There are stress tests that include simulating temperature variations in the functioning of the IC. Say your laptop is sitting in your car in frozen temperatures of -10 C. You turn it on and within 5 minutes it reaches a temperature of 85 C. And for the whole winter, you did that every evening. What about the head unit and the computer-controller that sits in your car, which you would drive for the next 15 years subjected to such fluctuations every winter in northern Maine?

There were too many mechanical issues that my mechanical engineering colleagues had to deal with when it comes to extreme low temperature testing. So, how low a temperature would you want us to verify and how much more extra are you as the consumer willing to pay for that low temperature testing?

We cannot just test one or two units to verify the absence of mechanical issues like incompatibilities between die and packaging, unlike people who hot rod their motherboards experimenting overclocking with the mere one or two processors they bought from ebay. We have to design the acceptable statistical distribution and the sampling plan that would fall into that distribution, that would apply to a stream of ICs flowing thro the product line.

Occasionally, the legality of the constraints could be rather involved, where the US govt agency requires the OEM to have their representative present while we test those ICs/processors, which could take a few days for a batch. That representative would sign off that we indeed had performed such tests at such constraints. That is how a $100 processor would cost the US govt $2000.

Such that if the US govt agency somehow decided to operate the equipment beyond the tested and verified constraints, we would no longer be legally held responsible for any mishaps or future malfunction.

  • 6
    \$\begingroup\$ Also we have to study the soak time for low temperature test. At the end of test, the IC is ejected and the test site rises by 5 C. Then the next IC comes in and be brought down to -50 C, which is a lot longer than -15 C. We can't presoak them to the exact -50 C because that may cause condensation and then vaporization, shorting the leads. Icing of presoaked units would get stuck. Low temperatures are also very dry. As the unit is pushed into the test site, the dryness due to low temps would more easily induce static buildup and discharge to kill the unit. \$\endgroup\$
    – Cynthia
    Commented Dec 15, 2014 at 11:56
  • 1
    \$\begingroup\$ I think this is a good answer. Relevant testing costs money in general. In the end you get what you paid for! \$\endgroup\$
    – Doombot
    Commented Dec 16, 2014 at 18:42
  • \$\begingroup\$ If you want components which are tested in more extreme conditions, ask for the military specced version. \$\endgroup\$
    – Mast
    Commented Dec 17, 2014 at 8:46
  • 2
    \$\begingroup\$ OK, disclaimer here. I did not define specs. I only wrote them. The designers or whoever up there define the specs. Designers may have their high and lofty specs, but yield maximization is the king. When we maximize yield, we would adhoc write specs for those that failed the designer's ideal and we produce a bin that they could market as a new product line. Maybe, one of the four 64k quadrants failed - why throw the processor away? Why not sell it as 192K cache under a different model name? \$\endgroup\$
    – Cynthia
    Commented Dec 18, 2014 at 13:57
  • 3
    \$\begingroup\$ Another disclaimer. I have never personally tested -50 C. The lowest I have done is -25 C. In my less than perfect experience, it is not possible to test -50 C using usual equipment. It is a specially constructed test site and robotic arms, with very careful precautions against static generation and discharge by robotic arm movements and against freezing and condensation, inside a very cold room. And probably operated not by the usual $15/hour operator but by the couple of $95K/year engineers who designed the site and their $60K/year technicians for a very special customer. \$\endgroup\$
    – Cynthia
    Commented Dec 18, 2014 at 14:08

Other than maybe batteries and maybe the LCD components generally don't get damaged directly, even by extreme cold temperatures. If temperatures are changed to extremes, especially rapidly, there can be physical damage due to mismatched contraction with temperature or temperature gradients.

However, operation at cold temperatures may not be possible- components change with temperature, to the point where they may no longer operate reliably, may not start up or may quit entirely. The gain of bipolar transistors drops with temperature. Much below about 50K most bipolar parts stop working entirely because of carrier freeze-out. Electrolytic caps don't like temperatures much below freezing, and their changes (higher ESR and lower capacitance) can cause other parts to be damaged. Digital CMOS parts may function more-or-less okay, but the analog portions of a chip may go out of spec or fail to work (such as the clock oscillator or BOR or ADC in a micro).

Even more weird stuff happens as you approach absolute zero - at 4.2K (liquid helium), for example, a 1N4148 can make a relaxation oscillator. Get even colder and ordinary solder can lose all resistance, which sounds really great until you get trapped magnetic flux.

  • 5
    \$\begingroup\$ Is that trapped flux the electromagnetic kind or the chemical kind used with solder? \$\endgroup\$
    – pjc50
    Commented Dec 15, 2014 at 9:00
  • \$\begingroup\$ Magnetic flux. Somehow (!) it never occurred to me that was an ambiguous statement.. thanks for pointing that out @pjc50. \$\endgroup\$ Commented Dec 15, 2014 at 12:11
  • \$\begingroup\$ When I use my smartphone in the frost, the image on the display updates very slowly (about 1 second before the new image fades in). \$\endgroup\$
    – Vi0
    Commented Dec 15, 2014 at 12:20
  • 3
    \$\begingroup\$ Yes @Vio that's a real issue- wide temperature range LCD materials need more voltage to work.. If the application is automotive you really can't have that slow response when the car is sitting at northern winter temperatures \$\endgroup\$ Commented Dec 15, 2014 at 12:22
  • \$\begingroup\$ See blessed geek's answer for a sampling of some of the ways in which your very first sentence is wrong. The rest of this is a pretty good answer, but you need to fix that first sentence. \$\endgroup\$ Commented Dec 17, 2014 at 15:09

The basic problem is that the density of the "free" charge carriers in semiconductors is a strong function of temperature. When the temperature drops low enough, there just aren't enough available carriers to allow the transistors, etc. to function, and the effective series resistance of the bulk semiconductor rises as well. The overall gain of the circuit drops below what the design engineer has allowed for, and it can no longer meet its performance specifications.


The temperature limit associated with an actual IC has more to do with thermal expansion/retraction than things like melting.

An IC is made up of different materials. The die, the substrate, the bondwires, the bonding method, the legs and the body. As the temperature changes, these different materials expand/contract and will rip apart from other materials not changing at the same rate.

You then have quality of doping, more of a problem at the edge of the wafer. That means an actual characteristics against the data sheet (rise time, propagation delay, etc.) do not meet the stated minimum/maximum as the mobility of the electrons are different (manufacturers usually make an IC and test at military temperature. If it fails, test at industrial temperature. If that also fails, test at a commercial temperature... If it fails, they scrap it and add it to their yield numbers).

Then you have the specifics of damage... Silicon does not have a lower limit w.r.t. semiconducting. It does have an upper limit at 175 °C where it will be damaged.

LCDs will form crystals and break down at extreme temperatures and equally dielectrics in capacitors start to break down.

  • \$\begingroup\$ This seems like a good reason for storage temperature limits, but doesn't really explain why there should be an operation temperature limit more strict than the storage temperature. I'd always assumed it was due to the resistivity of materials varying with temperature. \$\endgroup\$
    – Random832
    Commented Dec 16, 2014 at 20:56

Other problems at such low temperatures are for example that LCDs are freezing and have a really slow reaction.

And the more important point for modern IC technologies is an effect which makes them slower at lower tempratures (see Dealing with multi-Vt & multi-voltage domain timing/temperature inversion challenges).

I also found this interesting article which has some other important points regarding low temperature problems in it: Design electronics for cold environments.

  • \$\begingroup\$ You linked really interesting article. Can you make some "very short version" of it, before link die? \$\endgroup\$
    – Kamil
    Commented Dec 16, 2014 at 0:29

Few reasons:

  • in many cases - you can use components below minimum temperature, just don't expect that parameters will be same as specified in datasheet
  • capacitors will shrink - distance between electrodes will change
  • electrolyte in capacitor may freeze and capacity will change
  • many electronic devices are made of diffrent materials and they may act like bimetal break when you change temperature in wide range. In many cases manufacturers do what they can to avoid this by using materials with similar coefficient of thermal expansion, but sometimes this is impossible or just not required

    enter image description here

    I guess this is why high power devices are binned in high temperature. For example - some CREE diodes are binned at 85°C (185°F).

Sometimes it is not about minimum temperature, sometimes it's about how wide is temperature range.

If your device is supposed to work in very low temperatures - you should read about tin allotropic transformation.


Silicon in particular is reliant on thermal excitation of its dopants to act as a semiconductor, making the nature of its semiconductor properties highly temperature dependant. This gives you a fundamental low operating limit, and a fairly narrow range of temperature that you can design your chip to work over. If you need electronics that works across a large temperature range, you don't use silicon. Gallium Arsnide electronics operates down to millikelvin and below, but is a lot more expensive.


Resistors are designed with a mixture of materials with different thermal properties, so that the thermal effects cancel out and give a resistance value that is approximately constant with respect to temperature, over the specified range.

Outside the specified temperature range, the resistance of a resistor can and will diverge wildly from the specified value.

As a matter of interest, precision resistors sometimes balance the remaining temperature dependence with a dimensional strain dependance: As the substrate shrinks or grows with temperature, the strain on the resistive element changes it's resistance, compansating for some of the remaining temperature dependance of the resistive material.


Another factor is digital timing at low temperatures. Digital circuits typically run faster at lower temperatures, but circuit timing might fail (e.g. internal registers might fail due to hold-time violations), so the circuit would not function properly. In a laptop the HDD would probably not work due to mechanical issues (e.g. the heads don't align correctly over the disk tracks).


In general, colder makes semiconductor junctions faster, and colder is better. -50C is pretty modest actually, the big problems happen much lower.

But a lot can go wrong. Temperature cycling during the day can lead to thermal stresses. Condensation can arise and cause real problems, particularly when a cold surface hits warm moist air.

So your question is really incomplete. If stored in a thermal chamber at -50C, your laptop would probably be quite happy indefinitely. But if moved in and out of -50C, there is a lot of room for trouble. The absolute temperature is one factor, as is the range of humidity, the range of temperatures, and the magnitude of physical shock at low temperatures.


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.