8

How hot can my mouse get before it stops working? Assume the plastic is removed so it doesn't melt.

I'm really trying to gauge the max operating temperature for electronics that have no moving parts (ex. would the silicon circuit board melt at 1000 F? Is there something else I'm not asking?)

EDIT: Must silicon be used?

wizlog
  • 181
  • 1
  • 1
  • 5
  • 1
    At 400F you would already see components falling off the board because the solder melted. – Matt B. Dec 13 '11 at 05:26
  • There's also the temperature at which the circuit board melts and/or scorches. Basically, the epoxy burns off and the woven glass becomes visible, and the color (of the board, not the solder mask) changes from pale white to brown to black. The solder mask can burn off as well. – Mike DeSimone Dec 13 '11 at 14:11
  • 1
    @Mike: Silicon semiconductor components are useless long before the board burns or the solder melts. – Olin Lathrop Dec 13 '11 at 19:43
  • Check out this electronics.stackexchange answer which covers many of the problems with heat and electronics http://electronics.stackexchange.com/questions/13873/why-exactly-do-chips-start-malfunctioning-once-they-overheat/13903#13903 – SimonBarker Dec 13 '11 at 13:06
  • When do silicon boards burn? Does the board need to be made of silicon? – wizlog Dec 13 '11 at 22:51
  • Boards aren't made of silicon. Most are made of glass fibers and epoxy. – Olin Lathrop Dec 13 '11 at 23:14
  • So there is no silicon anywhere? – wizlog Dec 13 '11 at 23:19
  • There is silicon in the glass (as SiO2), but it doesn't have the exact same properties as the crystalline silicon used in chips. As Olin pointed out, the chips will fail first. (However, I have seen failed diodes burn the board. It was a common failure mode for 1980's all-in-one beige Macintoshes.) – Mike DeSimone Dec 14 '11 at 00:04
  • wizlog: Keep in mind that there is a major difference between what a cheap mouse can take and what a high quality circuit specifically engineered for harsh environments can survive. The web is full of stories of poorly designed laptops and PCs failing from overheating at room temperature, particularly when overclocked etc. – JonnyBoats Dec 14 '11 at 17:16

5 Answers5

9

Virtually all manufacturers data sheets will list both the minimum and maximum storage and operating temperatures for their parts. Further many parts will be available in multiple versions designed to operate in more strenuous environments.

So for example an integrated circuit (IC)may be available in the following versions:

Commercial 0-70°C. Industrial -40 to +85°C Military -55 to +125°C.

Needless to say, the cost goes up with increased performance. There are even chips speced for use in outer space. So for example a chip which costs $2 in the "Commercial" grade might cost $2,000 for the version rated for outer space! (As you can guess, they don't sell very many of that grade.)

As to your mouse, generally they try to use the cheapest components possible, so for your mouse 70°C might be the upper limit (it might even be quite less!) which is about 158°F.

One needs to keep in mind that the temperature of the components is affected by two sources:

1) The temperature of the environment. 2) Self heating - some components generate a large amount of heat in operation, CPUs in PCs are one example. This is why they are fitted with big heat sinks and cooling fans to keep them from getting too hot.

As a general rule of thumb, the more power a chip consumes the more heat it generates and the hotter it will get. Of course a mouse uses hardly any power so this should not be a big deal for a mouse.

Hans
  • 7,238
  • 1
  • 25
  • 37
JonnyBoats
  • 4,108
  • 1
  • 17
  • 19
  • What prevents it from being getting higher (the max operating temp)? Assuming its something non heat generating... – wizlog Dec 13 '11 at 01:23
  • @wizlog -- what happens is that the component values deviate from nominal, and the circuit stops performing the way it was designed to. This can put even more heat stress on selected components which will then fail utterly. – JustJeff Dec 13 '11 at 01:28
  • @JustJeff (This will sound stupid, forgive me.) What are component values? – wizlog Dec 13 '11 at 02:18
  • @wizlog, superuser may be a better place for the answers you need. – Kortuk Dec 13 '11 at 03:13
  • @Kortuk They sent me here. – wizlog Dec 13 '11 at 03:26
  • Component values are the specifications for a component - for example a 100 ohm resistor or .01 micro-farad capacitor. (in another field it might be a 150 HP engine or a 1/2 inch pipe). The point is that things are seldom exactly what they purport to be so a 100 ohm resistor might actually be 101.5 ohms. Further values can change over time, particularly when subject to a stress like heat. – JonnyBoats Dec 13 '11 at 04:03
  • The keyword here is *reliability*. Let's say the components of the mouse are rated up to 70°C and the tolerance in the component values are no problem there: That means the mouse will still work very reliable at this temps. Most certainly increasing the temperature to 90°C will still mean that let's say 5 out of 10 mice survive this treatement for a certain time period. But usually that is not sufficient: You want a high percentage of working parts even after a long time, so you don't push the envelope and go outside the specs. – 0x6d64 Dec 13 '11 at 11:07
  • @wizlog - for an example of what I'm getting at, a resistor that's 1000 ohms at normal temperature will actually become a 1500 ohm resistor at some higher temp. The characteristics of the devices that were selected as desirable by the designer actually change. So the pieces don't behave the way they were designed to. This can lead to further heat stress and eventually something will catastrophically fail. – JustJeff Dec 13 '11 at 11:23
  • 1
    @Kortuk We closed the question on SU as it was more pertaining to electronic engineering in general, and we'd like the questions to be focused on actual problems, not hypothetical assumptions. – slhck Dec 13 '11 at 14:08
  • @slhck, At EE it would be preferable to not perform engineering support for hypothetical assumptions also. It is a common issue. – Kortuk Dec 13 '11 at 17:33
2

In addition to all the good points mentioned in other answers, two points to note are:

  • many silicon power semiconductors are now rated at 175 C operating temperature (formerly 150 C was the norm)
  • Silicon carbide semiconductors can theoretically go higher, but it looks like packaging is the current bottleneck for high-temperature operation -- Powerex had a recent news item saying they are researching modules for 200 C operation. If you want to read an interesting technical paper on the subject, here's one from Powerex/Cree.
Jason S
  • 13,950
  • 3
  • 41
  • 68
  • What happens when silicon gets hotter? (I know that the melting point of silicon is 1414 Celsius.) – wizlog Dec 13 '11 at 22:05
  • well, for one thing, leakage current goes up exponentially with temperature, and at some point any transistor is going to be either strongly on or weakly on. – Jason S Dec 13 '11 at 23:28
1

As Jonny said, for individual parts the best thing is to check their datasheets. However from a general or theoretical point of view, consider the temperature at which silicon ceases to act like a semiconductor. If I remember right from semiconductor devices class many years ago, that is around 150 degC. It would be a good idea to look this up yourself instead of taking my word for it.

Also look at some datasheets for discrete silicon parts. Often they are rated for 150C or thereabouts max die temperature. Keep in mind that is die temperature. Since there is always finite thermal resistance between ambient and the die, the max die temperature is the point at which the device can no longer dissipate any electrical power, making it useless in a practical sense. Something like a mouse doesn't handle much power, so can maybe operate near the theorectical limit.

Then there is another effect. All dies are not created equal, even off the same wafer. Manufacturers test each one and sometimes bin them differently depending on special desirable characteristics they find. The ability to run at high temperature can be one of these characteristics. The dies that can stand high temperature may be sold as extended range parts, which means the non-extended range parts probably do crap out at lower temperature because they were selected that way.

Most silicon semiconductor chips are specified over at least the 0-70 degC ambient range, often with extended temperature range versions available for a higher price. In the case of chips, this is usually the ambient temperature, not the die temperature.

So to sortof answer your question, a unknown electronic device using silicon semiconductors and not specificaly handling power can probably survive 70C ambient. Maybe it can survive 100C ambient. When they dies get to 150C, they very likely won't work. What that means in terms of ambient is somewhat of a guess.

Olin Lathrop
  • 310,974
  • 36
  • 428
  • 915
  • 1
    Remember that this is a consumer mouse, probably not too much testing went into the individual parts ;-) – JonnyBoats Dec 13 '11 at 14:57
  • @Jonny: Actually quite substantial testing probably went into the actual parts. The semiconductor manufacturer doesn't know how the parts will be used, and there are always fallouts so individual dies are usually quite thoroughly tested. Whether they were additionally binned by temperature tolerance depends on whether high temperature versions of the IC have a sufficient market, like military applications. If the IC is just a mouse controller, then probably not. – Olin Lathrop Dec 13 '11 at 17:02
  • Olin: in a cheap mouse this is probably one IC in a glob top package made in China. You can bet they did not spent 1 penny extra that they did not absolutely have to. I seriously doubt that it comes from an ISO certified facility with 6-sigma ;-) – JonnyBoats Dec 13 '11 at 17:19
1

There are a lot of different materials inside your mouse besides "silicon." The chips are made out of silicon, but then they're put inside a (probably) plastic chip package, soldered to a circuit board made of fiberglass, surrounded by other components like capacitors, resistors, connectors, etc. which are not made of silicon. Each component will have its own acceptable temperature range, outside of which it cannot be relied upon to continue acting as specified in the data sheet, for reasons ranging from increased thermal noise or other electrical issues, to just plain melting or igniting. Suffice to say that temperature is in the equations that model how materials behave (thermal expansion is a simple example), and over a broad enough range, temperature may become the dominating factor determining behavior of a material and the device which is made of it.

Matt B.
  • 3,805
  • 2
  • 24
  • 28
1

When I do reflowing i heat my PCB up to 220Degress Ceclcius.. but not longer than 2-3 minutes. Obciosly these need to be in line with the specs of the IC's and the PCB has to be SMD compatible.

In genral pcb's should not be exposed to heats more than 100Degrees Celcius for long periods of times.. peaking is ok, but not advised.

WillyWonka
  • 221
  • 1
  • 3
  • There is a big difference between short time storage or soldering temperature and sustained *operating* temperature. Note that the OP specifically asked about operating temperature. – Olin Lathrop Dec 14 '11 at 15:48