Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!wuarchive!brutus.cs.uiuc.edu!jarthur!polyslo!gaudi!csuf3b!jamespa From: jamespa@csuf3b.CSUFresno.EDU (James Paul) Newsgroups: comp.sys.mac.hardware Subject: Re: Leave it on? Or once a week? Message-ID: <1990Feb15.103647.4309@csuf3b.CSUFresno.EDU> Date: 15 Feb 90 10:36:47 GMT References: <77800007@m.cs.uiuc.edu> <9160@portia.Stanford.EDU> <1990Feb14.052956.4203@csuf3b.CSUFresno.EDU> <3118@pur-phy> Reply-To: jamespa@csuf3b.UUCP (James Paul) Organization: C.Sci Dept., CA. State Univ, Fresno Lines: 53 In article <3118@pur-phy> sho@maxwell.physics.purdue.edu.UUCP (Sho Kuwamoto) writes: >In article <1990Feb14.052956.4203@csuf3b.CSUFresno.EDU> jamespa@csuf3b.UUCP (James Paul) writes: >>Well and fine, but remember that hard disks have a MTBF rating (mean time >>between failure) and usually spin at about 3000 RPM. :-) If you don't mind >>eating your disk drive's time away, no problem. > >On the contrary. I'd heard that almost all the wear and tear on a hard >disk was due to power up and power down, where the heads might come >into contact with the disk, or some such. > >-Sho >-- >sho@physics.purdue.edu <<-- me, I turn mine off every night. > but I have the noise mac II fan... There are different types of "wear and tear." Electrical fatigue (my term) is the wear caused to solid-state equipment due to the energization and state changes when power is supplied or removed to a device. Generally, if you power up a solid-state device "gently" there is little problem. Mechanical "wear and tear" is another thing altogether. A disk drive controller is solid state (I hope! ;-) The drive mechanism itself is a motor, drive system, bearings, etc. all operating at a respectable mechanical speed. Obviously, these mechanical parts will not run forever, and are subject to wear. The MTBF figure, as I understand it, is largely based on the expected average reliability period of the mechanical parts. (Note that the MTBF figures do not include a duty-cycle, which would provide more info.) The figures for most drive units fall between 1 and 3 years, if continously used. The danger of a head-crash or other failure during power cycling is not unfounded, but I think it's exagerated. In a truly fixed drive, (one that is not removable and is sealed,) A head crash rarely occurs simply because the platter spins up to speed. Other variables come into it, though, like dust, etc. In general, keeping a disk spinning about 3000 RPM when it's not being used most of that time is a poor trade-off in trying to baby it. If it is so fragil that it might fry when powering-up, I wouldn't _dream_ of leaving it on all the time. :-) Of course, if the application requires the disk to be available all the time, the point is moot. (We have some machines that have to be awake 24 hours, and the drives have to be fixed about every 2 years. Usually the motor-drive gives out.) -James -- -------------------------------------------------------------------------- UUCP:jamespa@csuf3b.CSUFresno.EDU Compuserve:72767,3436 GEnie:J.PAUL Packet:N6SIW@K6RAU AppleLink:D1231 America Online:JLPaul