In the old days a consumer product might have dozens of moving parts. It was easy to think of ways that the equipment might fail. Any of the moving parts might break or otherwise contribute to failure. But nowad
ays a piece of electronic equipment might have very few moving parts or maybe no moving parts. How can such things fail? What is a typical failure mode these days? I have recently come to appreciate that nowadays, for some consumer electronic devices maybe the most common failure mode is the only moving part, namely a cooling fan.
Shown in the photograph is a video card from a computer in our office. It has no moving parts, other than the cooling fan (in the green circle). In recent weeks, the computer had been a bit flaky. Sometimes the screen display would act up. Not right away when the computer is turned on. Only after a while would it act up. Eventually we figured out the explanation. The cooling fan had seized up. The bearings were shot and the fan did not rotate any more. So the card would heat up and eventually it would stop working. Later when the computer cooled down it would work normally again for a while. The other day, we replaced this card. It was not easy to find a video card that did not have a cooling fan, but eventually we found one.
Which is only the latest in a series of consumer electronic devices that have had their cooling fans fail on me. As a couple of examples:
A DirecTV digital video recorder. The cooling fan was getting noisier and noisier. The noise was a warning signal of imminent bearing failure, of course. Not only that, the DVR also occasionally flashed a warning message on the screen, saying that the CPU was getting hotter than usual. By some good luck I happened to have a spare DVR with a working fan in it, and I swapped it with the noisy fan. Now hopefully the replacement fan will last a while.
A power inverter. This is one of those things that plugs into the cigarette lighter socket in the car, and it receives 12 volts DC from the car, and it gives you 120 volts AC. And just the other day this power inverter started acting up. When it was powered up, if it was oriented in a particular way the cooling fan would be very noisy. Again a warning signal of imminent bearing failure. Time to get a new power inverter.
One of our firm’s main RAID servers (see “Picking a hard drive for your RAID system“) has a feature that monitors the rotation speed of the cooling fan. If ever the fan were to slow down, we would receive warning email messages from the server. Hopefully this would provide early warning and would give us some time to think about what to do.
Unfortunately it would not be a realistic goal to try to keep enough spare cooling fans around to be prepared for all failures. The cooling fans are of varying sizes and shapes. The mounting holes might be different from one to the next. Some of them (like the one in the photograph) are integrated with a heat sink. I think one simply has to hope that one may be lucky and that most cooling fans will have a fairly good service life.
Which then leads to the endless debate among tech geeks. Is it better to leave things going 24 hours a day? Or is it better to turn them off every night and turn them back on every morning? Some people figure that if the device is left on all the time, it will wear out sooner. Other people (me included) figure that one of the evils to guard against is expansion and contraction as things get hot or cool. Every time a device is turned off for the night, the result is that all of the parts contract. Next morning they all expand.
What do you think? Leave it on all the time? Turn it off every night? Time to post a comment.
Carl, I was thrown off a bit by the title. I was hoping the article related to EFS-Web (RAM), especially with all of the problems of late. With regards to the post, I tend to turn off all gadgets when not in use (exception: phone).
Please keep the posts coming… good content and always learning something new. Thanks!
As a guy who studied this problem as an engineer in the aerospace industry, cycling hot/cold should be worse (and was worse). But the computer industry now says that the lifecycle data on current generation computer hardware indicates a tossup. I leave em on.
since most of your failures are fans which you haven’t had much luck replacing, I’d turn em off at night until something else started failing first.
A maintenance tip – Regularly clean the fans on laptops (and other electronics having mechanical fans) by using compressed air (cans) to prolong the life of the pc. Too much dust = less air flow = higher temperatures. High temps from diminished airflow kill a lot of laptops. I’ve rescued more than one laptop that was shutting down unexpectedly by simply blowing out the dust.
Bob
Just in case it’s of any use to anyone: my son’s laptop had problems–very noisy fan and laptop got very hot–was probably on the verge of seizing up too. He opened it up and it turned out it was simply a big ball of dust caught up in the fan, which he was able to remove, and it worked fine afterwards. So I suggest checking for this if you suspect any problem with the fan.
It was my understanding, many years ago, that electronic devices were left on all the time because the tubes had filaments. So on and off was a big deal for the filaments, and many radio operators went to great pains to avoid having the power go off. Not sure that is true of transistors, but for sure the cpu gets hot.
If one of the main causes of failure in a cooling fan is that the bearings wear out, shouldn’t it be reasonable to turn the thing off when it is not being used? My main concern, however, is power surges during severe thunderstorms, so I turn most things off (unfortunately) frequently, so the on-off question is moot.
It is surprising to me how much dust accumulates on all fan blades and ventilation ducts in a house that otherwise seems clean.