DBG@SLACVM.SLAC.STANFORD.EDU (02/23/91)
Hard drive bearings gradually fail while the drive is not spinning.
Vibrations from the environment polish (brinnell) the contact areas
on the ball bearing races, and this eventually causes the bearing
to vibrate during operation.
ICs usually fail during on/off thermal transients. The chip material
and package material are not perfectly matched for thermal expansion,
so temperature changes cause small relative motions that cause cracks
to grow from the rough edges of the chip into the working areas.
I estimate that about 90% of the failures I have seen in personal-
computer-like equipment have occurred during power on/off transitions.
Machines that are turned off overnight fail noticeably more often.
When you consider the environmental impact of leaving the mac powered
up, you should also consider the cost in energy of replacing failed
parts (usually a whole board is discarded and replaced). This cost is
generally proportional to the price of the board (energy is an important
part of the cost of anything if you follow the components back far
enough in their production).
You also have to consider the value of your time for recovering from
the equipment failure.
I leave my Macs on all the time unless I know a power outage is
scheduled (an unsupervised restart destroyed my Jasmine 80 when
head stiction stalled the disk rotation, and heat buildup over
several days baked the disk into oblivion). I do power off my
LaserWriter overnight, though I wonder if even that energy saving
is worth it. Yes, I do pay my own electric bills (and repair bills).
I now use an uninterruptable power supply to avoid unsupervised
restarts of my hard drives and to keep my faxmodem up at all times.
-- David B. Gustavson, Computation Research Group, SLAC, POB 4349 MS 88,
Stanford, CA 94309 tel (415)926-2863 fax (415)961-3530
-- What the world needs next is a Scalable Coherent Interface!
-- Any opinions expressed are mine and not necessarily those
of the Stanford Linear Accelerator Center, the University, or the DOE.ROCKY@cup.portal.com (John Richard Bruni) (02/26/91)
Regarding unannounced power outages, I got hit with a good one a while ago. The power cut cleanly, but when it came back on my incandescent lighting flickered at what looked like a 10hz rate, cycling from off to bright for several seconds at this guestimated rate. PG&E lost a power cable, they say. At any rate, my Mac & hard drive survived just fine (they were on when the power failed & I foolishly forgot to turn them off). The power amp in my stereo failed spectacularly however. I *always* left it on. Alas, the bridge rectifier failed and raw AC ran amok through the drive and output transistors. What a stink! The message here is that at least in one case a MAC and a hard drive survived a gross insult that a less intricate piece of gear couldn`t endure even though it had several safety circuits built in.
bruner@sp15.csrd.uiuc.edu (John Bruner) (02/26/91)
One more thing to remember when considering whether to leave your computer on all of the time: there are no air filters on the Mac. A machine running 24 hours a day will accumulate considerably more dust inside than one that is turned on only when it is used. -- John Bruner Center for Supercomputing R&D, University of Illinois bruner@csrd.uiuc.edu (217) 244-4476
gillies@m.cs.uiuc.edu (Don Gillies) (02/28/91)
Re: leaving the Mac On. As part of a UIUC campus energy audit, we recently contacted several computer manufacturers (HP, IBM, Apollo) about whether it is a good idea to leave their PC's turned on or to turn them off immediately when not in use. Every manufacturer contacted so far has said to TURN THEM OFF. HP has said that its burn-in testing process for PC's consists of cycling the power on and off every second for 24 hours. I think your "anecdotal" evidence is quite likely a crock of shit. Disk drives fail because they are poorly manufactured. You "seem" to notice that they fail when the machines are turned off -- so what? People who turn off their machines when not in use have them turned off for at least 66% of their lifetime, so it is natural that the disks would die when turned off. Disks fail with old age and hard use. On the savings side, small PC's (mac SE, SE/030) consume 60 watts of power and add to air conditioning (this facts were determined empirically with electrical measuring devices). Big PC's (Mac II, IBM PS/2 50, 70) take more than 110 watts when turned on, and add to air conditioning costs. For a Mac II, at 8c/kilowatt, 66% wasted duty cycle, that is 8c/day, $58 a year in wasted power (40-hour work week). My Mac II has been cycled 2-3 times aday for 3.5 years with a Quantum-80 disk drive, and never exhibited trouble. And I'm not counting the savings in air conditioning, etc. --
gillies@m.cs.uiuc.edu (Don Gillies) (03/01/91)
Oops, that last paragraph was a little incoherent. Let's try again. On the savings side, small PC's (mac SE, SE/030) consume 60 watts of power when turned on and add to air conditioning (power consumption at UIUC was determined empirically with electrical measuring devices). Big PC's (Mac II, IBM PS/2 50, 70) take more than 110 watts when turned on, and add proportionately more to air conditioning costs. For a Mac II, at 8c/kilowatt-hr, the cost of leaving it on is $58.60 a year in wasted electricity (assuming a 40-hour work week, i.e. a 128-hour waste week). My Mac II has been cycled an average of 2-3 times daily for 3.5 years with a Quantum-80 5" disk drive, and has never exhibited trouble. The savings are even greater if you consider the reduced air conditioning. I would think that motor failure is the most improbable failure mode for a hard disk. More likely is media failure (magnetic oxide being scraped from the disk platter), stepper-motor failure (failure of the disk arm to be able to move correctly over the surface of the disk), or electronics failure (failure of the controller chips). Also, it may well be that conservation minded people cannot afford to buy the highest quality machines and blow the power to keep them running all the time. The cheapest hard disks on the market use the lowest quality components and exceed manufacturer's specs by the smallest margins (this is a quote from a friend of mine who does reverse-engineering for the #1 hard disk maker in the world). Don Gillies | University of Illinois at Urbana-Champaign gillies@cs.uiuc.edu | Digital Computer Lab, 1304 W. Springfield, Urbana IL ---------------------+------------------------------------------------------ "UGH! WAR! ... What is it GOOD FOR? ABSOLUTELY NOTHING!" - the song "WAR" by Edwin Starr, circa 1971 --
gillies@m.cs.uiuc.edu (Don Gillies) (03/01/91)
DBG@SLACVM.SLAC.STANFORD.EDU writes: >When you consider the environmental impact of leaving the mac powered >up, you should also consider the cost in energy of replacing failed >parts (usually a whole board is discarded and replaced). This cost is ^^ ^ ^^ ^^ ^ ^^ ^ ^^ To the best of my knowledge, this is wrong in the case of Apple computer. Board swaps are done simply because it is faster to stock and install working replacements, than to keep the machine down while some yahoo plays with the circuit board. I believe apple has technicians to do repairs at a national, central facility, with advanced diagnosis tools. Repaired boards on day live again in other macintoshes. Most mainframe makers follow the same policies, because it is too wasteful not too. Don Gillies | University of Illinois at Urbana-Champaign gillies@cs.uiuc.edu | Digital Computer Lab, 1304 W. Springfield, Urbana IL ---------------------+------------------------------------------------------ "UGH! WAR! ... What is it GOOD FOR? ABSOLUTELY NOTHING!" - the song "WAR" by Edwin Starr, circa 1971 --
sl242030@silver.ucs.indiana.edu (Robert P. Humphrey) (03/01/91)
In article <1991Feb28.205440.13514@m.cs.uiuc.edu> gillies@m.cs.uiuc.edu (Don Gillies) writes: >DBG@SLACVM.SLAC.STANFORD.EDU writes: > >>When you consider the environmental impact of leaving the mac powered >>up, you should also consider the cost in energy of replacing failed >>parts (usually a whole board is discarded and replaced). This cost is > ^^ ^ ^^ ^^ ^ ^^ ^ ^^ > >To the best of my knowledge, this is wrong in the case of Apple >computer. Board swaps are done simply because it is faster to stock >and install working replacements, than to keep the machine down while >some yahoo plays with the circuit board. > >I believe apple has technicians to do repairs at a national, central >facility, with advanced diagnosis tools. Repaired boards on day live >again in other macintoshes. Most mainframe makers follow the same >policies, because it is too wasteful not too. > I know this to be true. I do some courier work for a store that deals exclusively with Macintosh. I also had to get my drive replaced there. (If you wanted to know why your mother told you to never lend anybody anything that you didn't want hurt, and you wanted to know the greatest possible application of Murphy's Law I've ever witnessed, and finally, why I think that non-computer users are inherantly stupid, let me know.) Anyway, instead of taking the drive apart and replacing the smashed drive head, I had to buy a new drive, and trade in my old one, which the shop sends back to Apple, where they replace the $3 wicket and sell it to someone else for $300. Sometimes it pays to order your own parts and fix it yourself, but I couldn't find any for the new SuperDrive. Apparently knowbody had broken theirs yet but me (or rather, my ex-friend). -- ******************************************************************************* Robert Humphrey, "It's easy to grin when your ship comes in, Gentleman Scholar And you've got the stock market beat; But the man worthwhile, Is the man who can smile, sl242030@ucs.indiana.edu When his shorts are too tight in the seat." -Ted Knight, Caddyshack *******************************************************************************
rubin@chocktaw.cis.ohio-state.edu (daniel j rubin) (03/03/91)
>As part of a UIUC campus energy audit, we recently contacted several >computer manufacturers (HP, IBM, Apollo) about whether it is a good >idea to leave their PC's turned on or to turn them off immediately >when not in use. > >Every manufacturer contacted so far has said to TURN THEM OFF. HP has >said that its burn-in testing process for PC's consists of cycling the >power on and off every second for 24 hours. > Of course they want you to leave you computers on. That way when it dies due gaps in the copper traces on the PC board from thermal stress of going from hot to cold and vica-versa you can untimatly go out and buy another one from them..... They aren't stupid you know... - Dan Rubin
hamilton@kickapoo.cs.iastate.edu (Jon Hamilton) (03/03/91)
rubin@chocktaw.cis.ohio-state.edu (daniel j rubin) writes: >>As part of a UIUC campus energy audit, we recently contacted several >>computer manufacturers (HP, IBM, Apollo) about whether it is a good >>idea to leave their PC's turned on or to turn them off immediately >>when not in use. >> >>Every manufacturer contacted so far has said to TURN THEM OFF. HP has >>said that its burn-in testing process for PC's consists of cycling the >>power on and off every second for 24 hours. >> >Of course they want you to leave you computers on. That way when it dies Um, I was kinda hoping that this thread would have died by now. Statements like the one above are one reason it hasn't -- it sure looks to me like the manufacturers told him to turn it OFF. Of course, his argument that you should turn the machine on and off because that's what they do during burn-in testing doesn't make too much sense either. Doesn't it seem that you'd want to do something pretty stressful to see whether the machine was going to die? If it was going to cause more stress to leave the machine on, it would make more sense to perform burn-in testing by leaving the machine switched on. [...] >They aren't stupid you know... > > - Dan Rubin They're also not the end-all authority. > -- % Internet : hamilton@kickapoo.cs.iastate.edu | Insert cute and/or % % America Online : JonHam | deeply meaningful % % Elsewhere : ThatGeek@his.little.corner | musical quote here %
tommyp@isy.liu.se (Tommy Pedersen) (03/04/91)
sl242030@silver.ucs.indiana.edu (Robert P. Humphrey) writes: >In article <1991Feb28.205440.13514@m.cs.uiuc.edu> gillies@m.cs.uiuc.edu (Don Gillies) writes: >>I believe apple has technicians to do repairs at a national, central >>facility, with advanced diagnosis tools. Repaired boards on day live >>again in other macintoshes. Most mainframe makers follow the same >>policies, because it is too wasteful not too. > I know this to be true. I do some courier work for a store >that deals exclusively with Macintosh. I also had to get my drive >replaced there. (If you wanted to know why your mother told you to >never lend anybody anything that you didn't want hurt, and you wanted >to know the greatest possible application of Murphy's Law I've ever >witnessed, and finally, why I think that non-computer users are >inherantly stupid, let me know.) Anyway, instead of taking the drive >apart and replacing the smashed drive head, I had to buy a new drive, >and trade in my old one, which the shop sends back to Apple, where >they replace the $3 wicket and sell it to someone else for $300. >Sometimes it pays to order your own parts and fix it yourself, but I >couldn't find any for the new SuperDrive. Apparently knowbody had >broken theirs yet but me (or rather, my ex-friend). My comment should really be in alt.security, but since we are discussing change of drives I would like to point out the security risk in letting some other company have your repaired disk drive. The company might have equipment for reading the disk and thereby exposing your data. /Tommy Pedersen ________________________________________________________________ |E-mail: tommyp@isy.liu.se || Telephone: +46 13 282369 | |S-mail: Tommy Pedersen || FAX: +46 13 289282 | | Dept. of EE ||______________________________| | Linkoping University || | | S-581 83 Linkoping || | | SWEDEN || | |________________________________||______________________________|