High pitched tone from PC box

[jg.campbell.ng]

My PC (Dell Precision 370, about 14 months old) is emitting a high pitched tone; I'd say between 8 and 12 KHz but that is a guess; that guess based on recall of acoustic-coupler modems. Loud enough to be annoying; hardly loud enough do much about.

It is not the (19 inch flat analogue, not CRT) screen and I have isolated other items like USB disks.

Reluctantly, I switched the machine off and the noise disappeared. The machine has been switched off only once before; not green, I agree, but I don't like switching off critical machines.

A transformer or other power supply part?

There may be a momentary slight modulation of the sound when the screen changes, hence the initial suspicion of the screen.

Laziness, I agree, but I'm putting off dismantling everything and opening up the box; and given the frequency, it will be difficult to pinpoint anyway.

TIA,

Jon C.

[The3rd Earl Of Derby]

If your not willing to take out cards,disconnect drives, ect then you're not going to eliminate the source.

[ABC]

If it that much of a critical machine, then whatever software you are running, it should be on a dedicated server, and not a standard Dell PC.

The only way to identify and rectify the problem is to open the box and take a look.

A car mechanic can't fix a car without opening the bonnet. The same is true for IT technicians.

[Phil Anthropist]

Make a simple stethoscope from a 30 to 50cm length of garden hose or similar (with or without a small funnel attached) then listen to the various bits until you pin-point the source of the whine.

[jg.campbell.ng]

Thanks, yes I guess that's what I need. Following comments of others, I had the box open; could hear nothing.

Thanks,

Jon C.

[jg.campbell.ng]

I work mostly from home, so it is critical to /me/. I suppose it could be switched off for the eight or so hours that I sleep, but then I'm not sure that I'm going to sleep for eight hours; anyway switching computers off and on just gives me the creeps.

Yes, done; and as I suspected I could hear nothing. I think I'll have to go with the stethoscope suggestion and a concerted attack on the problem.

Or, leave the wax to build up properly in my ears; ref. my post about a year ago on some other newsgroup about this machine being dead quiet, and all that happened is that I'd located it on my left and my left ear was clogged :)

Best regards,

Jon C.

[Alex Threlfall]

Probably the PSU, i've had similar things happen to mine and it drove me potty(er) too :)

[dennis]

Some motherboards emit that sort of sound when a fan has failed. But he doesn't want to open the box so he can't see if the fans are OK.

[Dave Plowman (News)]

It's worth buying a cheap one from a medical shop. About a fiver or less.

[Sparks]

Sorted then - leave the side off :-)

[jg.campbell.ng]

Of course the fan is running, I knew that all the time. Just because I ask a simple question on a newsgroup doesn't mean I'm a total idiot.

And, yes, I did open the box, see above. As I suspected I was not able to locate the (high pitched and very faint) sound; indeed, I could not hear it at all with the lid off. Note: this is so faint that you would never hear it in an ordinary office, nor, I suspect, in a machine with a reasonably noisy fan.

Aha, though, mention of fan makes me think; I cannot hear it now; temperature in the room is now 21.5 C; was about 24 yesterday when I was provoked into making the original post.

Best regards,

Jon C.

[Bob Eager]

The fan can still be failing. Calm down. That kind of noise can come from a fan bearing on its way out, and it can come and go. Feel free to ignore this advice too.

[jg.campbell.ng]

[...]

[...]

When I change to another window or change significantly what is in the currently displayed window, there is a noticable increase in volume aabout a second or two afte the changer. I guess a graphics card (Radion 128-MB something) does draw significant current. So could point at power supply.

I'll report back as soon as I get hold of a stethoscope and do a peoper investigation.

Thanks all,

Jon C.

[Andy Hall]

Fans on graphics cards as supplied by the manufacturer are notoriously crap. Typically they have sleeve bearings and squeaking is often the precursor to complete failure. When that happens, the card is typically a write off because the chip will normally overheat. Often nowadays graphics card fan speeds are wound up and down according to the temperature and that is related to graphics activity.

Equally, the CPU fan can do the same thing. Some of these also go up and down.

What I always do when building a new PC is to replace the supplied fans with good quality ball bearing ones like Zalman etc. These cost about £20 a go. Considering the cost of replacing motherboard, memory, CPU, graphics card,.... this is a good investment.

If it isn't these, then the other likely candidate is the power supply. Again the power load can cause squeaky sounds from the electronics of the supply as the load varies.

So I would change the fans first, as a matter of course, and if that doesn't address the problem, the PSU.

[RedOnRed]

The same thing can happen to a dedicated server. The user merely has to back up their data.

[Stuart Noble]

j

A hard drive of mine used to do exactly that, corrected for a while by a sharp whack. The sound would also disappear when the case was off. Turned out to be a faulty IDE connection. Hope you're backed up. Argos are doing a nice little 40gb mp3 player for £50 (Seagate hard drive) :-)

[RedOnRed]

I've been in IT for years now and something like this could be too numerous things to mention and not really worth speculating about - certainly not without seeing it.

It really is process of elimination, starting with the easiest most obvious things. The casing, for example can vibrate, especially if something is touching it. Electrical interference. The fan, etc, etc.

I visited one user recently who had logged a call saying her mouse wasn't working. After noticing no lights or display on her PC it quickly turned out it simply needed switching on! I'm not implying for one minute you are of the same callibre, but some faults aren't always what they seem.

[jg.campbell.ng]

[...]

Thanks. And to everyone else. And apologies for earlier taking on an attitute unbecoming of someone looking for help. I agree also that there is no place for an assertion "it couldn't be X" until that has been clearly proven.

When (and if) I ever locate the problem, I'll report back. Because of the frequency, I (now) know I'm going to need a device like a stethoscope to locate the sound. The quick rummage last night with the case open revealed nothing because, as reported, I could hear nothing, so there was little point in disconnecting anything.

In addition, given the list of DIY jobs to do around this house, it may have to get a lot louder :(

I am backed up --- /and/ the viability of the backup has been verified (well worth a thought).

Best regards,

Jon C.

[Andy Dingley]

Don't waste your time with a medical stethoscope - the design of the bells is different, so as to make them useful for listening to widespread sounds from squishy things. If you're after mechanical noise, get a mechanic's stethoscope with a rigid probe connected to a diaphram.

[Dave Plowman (News)]

How much are those?

I've found a cheap medical type far better than the tube method and less dangerous and as effictive as the screwdriver trick.