Magnets on the work bench
in [Design]
|
Prev: nexgen T-3 Fibs ProTrader Torrent - www.top-tradesoft.com
Next: I need a reflow oven for my production lab immediately!! HELP!!!
From: life imitates life on 11 Feb 2010 21:16 On Thu, 11 Feb 2010 16:36:15 +0000, Dirk Bruere at NeoPax <dirk.bruere(a)gmail.com> wrote: >Tim Williams wrote: >> "Chris" <christofire(a)talktalk.net> wrote in message >> news:9f794283-f986-4e41-93af-5600bd2dbe34(a)z26g2000yqm.googlegroups.com... >>> Those are all about 1 tesla according to their table at >>> http://www.supermagnete.nl/eng/data_table.php, which seems typical for >>> Neodymium (-Iron-Boron) magnets, and those are 'the strongest magnets >>> in the world' according to that web site! >>> >>> Tesla is the unit of flux density. The number around 40 is an 'energy >>> product' apparently: the square of B, in Tesla, divided by mu0 for >>> air, although the unit MegaGaussOersted must involve a strange >>> conversion. >> >> The type 50M shows a particularly high remenance of 1.40-1.46T, from a >> magnetization of 860-995kA/m (which means, for a 1 cm thick magnet in a >> steel fixture, you need to pulse about 10kAt around it -- 100A * 100 turns >> let's say!). >> >> Assuming the B-H curve is exactly that square, the "energy product" is: >> 1.4527M T*A/m = Wb*A/m^3 = V*s*A/m^3 = J/m^3. >> >> (Webers are volt-seconds, the amount of flux applied. Volts get into it >> because integrating EMF over time gives flux, hence, V*s. Flux actually has >> nothing to do with amps, which sounds funny for magnetism which is all about >> amps, but it's like how current and voltage are dependent on resistance, >> there's a connection, it's just not a direct connection.) >> >>> I wonder if it's significant that I ordered an even number?! Perhaps >>> if you order three, they send you a fourth one free of charge! >> >> But then you'd have to magnetize it yourself! <rimshot> >> >> Tim >> > >I've known people stick floppy discs to filing cabinets with magnets. >No harm done apparently I have passed magnets over and near floppies, no problem. I would NOT place a standing field next to one at that proximity, and leave it there, however. That is just asking for errors. Could even re-orient some of the hard sector flags. A big magnet washes over the whole disk, and everything pretty much snaps back when the field is removed. A small magnet, holding it against a steel medium behind it bothers me just thinking about it. :-)
From: life imitates life on 11 Feb 2010 21:28 On Thu, 11 Feb 2010 11:24:56 -0700, D Yuniskis <not.going.to.be(a)seen.com> wrote: >Who said anything about power loss? If the OS contains a bug >in the driver -- or, in my case, the disk had some funky >characteristics in the SCSI interface -- then the VTOC can get >trashed in one write. > >I never investigated the actual cause. It was seemingly the SCSI controller and that drive pair up. I would not blame the OS or driver first. I look at standards implementations made by the controller maker, etc. Brand names matter in certain realms too. Maybe someone thought "full compliance" meant "fool compliance". In the early days of SCSI match ups which would not work were common. I thought all that was worked out by now. Configuration can also cause problems, but that is usually with connecting to the drive at all. Once connected, and successfully formatted, which it obviously was, it should function fine. I suppose a virus could worm its way into the driver and cause the catastrophic failure you describe. Normally the DRIVE performs the operations from within the hardware local to the drive, so the interface sends and receives cluster write commands, etc. Should still be no way to "spray bits all over the drive" or the VTOC, for that matter.
From: life imitates life on 11 Feb 2010 21:53 On Thu, 11 Feb 2010 11:43:05 -0700, D Yuniskis <not.going.to.be(a)seen.com> wrote: >life imitates life wrote: >> On Wed, 10 Feb 2010 13:49:55 -0700, D Yuniskis <not.going.to.be(a)seen.com> >> wrote: >> >>> <grin> We don't allow magnets in the house (way too much magnetic >>> media!) >> >> I have put hard drives, VHS tapes, Floppies, and all other manner of >> magnetic storage medium on my speakers and have never, ever had a >> problem. I have even taken magnets and tried to erase floppies and hard >> drives etc, with closely proximal magnets, and have never been able to >> erase data from them. > >The magnets on most speakers are sufficiently far from the edges >of most cases. Field drops as inverse square of distance so >there's little threat from a speaker (to *media* -- you might >note that unshielded speakers near a CRT will cause problems, >especially for a very high resolution CRT!) > I started out with old reed contact pinball machines and worked into upright video games, so I have been around a lot of CRTs. In the case of nearly horizontally placed CRTs in such games, the orientation of the game when demagnetized and set-up and the orientation of where one places it in the arcade would determine whether it looked right or not. Turned one way, it was perfect. Turn it 90 degrees from the, however, and the colors get all funkified. I used to be able to demag a CRT WITH a magnet. A huge speaker magnet, I would 'wash' all around the face and edges of it from about two feet away, and then flip the magnet over and do it again. Working closer, and closer, and remembering to always flip and equalize, one can remove any anomalous magnetizations that occur over time, just like the coils do automatically these days, and the big tech owned coils did in the way back when. >"Refrigerator magnets" are also no threat (to media -- usually) >since most refrigerator magnets can barely hold a piece of >paper on the refrigerator (nowadays). The orientations flip in the presence of the field, make no mistake. They simply flip back when the field is removed. The concentration of the small magnet's field when it is only about 0.040" away from the disc face, makes me think it might beat the perm figure of the disc media and flip some of the opposite oriented domains and them not return when the field is removed. I too have yet to have seen floppies lose data, but I have seen old, wide head 360kB disc drives with discs that were written by the newer 1.2MB thinner head drives, that would not read on the older drive due to the reduced magnetic energy of the thinner track of data being read by a head expecting a much 'fuller' 'bit flip' 'signal' coming off the data as it passes by. Too many false highs and false lows. > >Now, where do you draw the line? Do you inspect each magnet >that comes into the house and characterize the strength >of its field? I have magnetron magnets in the other room, that sat out here in the living room for well over a year. > What criteria do you use to determine >"this one is no threat; this one IS a threat"? Threat to whom or what? Me? Not worried. My magnetic recordings? Not worried. Of course, we will not be seeing me wheeling any monsters in here on a hand truck any time soon either. But I do like magnets. I would not pass one up if I saw it at a swap meet or garage sale, etc, and it was pennies on the dollar. I could quick turn it on ebay if it was too week to please me, and keep it if it is good! Or not. >> I did own a mechanical, magnet only cassette eraser that worked >> perfectly upon completion of a dual pass thru with the cassette tape. >> >> That little sucker had an array of opposed field magnets setup inside >> it, and it was just barely a free fit on the cassette form factor. >> >> For the most part though, written data is fairly strongly held on the >> medium. A nearby field morphs it, but usually only temporarily, and >> rarely does it flip even one bit. > >When I use my demagnetizer, I move to the far side of the house. >It's field is strong enough that it can support a 7" reel of tape. >Unfortunately, the only control is on/off so there is no way to >slowly ramp the field down in intesity. Apparently the sudden >interruption of current can cause a large spike/"magnetic event". Wow. Mine was all solid magnets, like a parquet floor. Two layers on each side. On the tape... bits be gone. :-) >> Modern hard drives with perpendicular recording are even less prone to >> loss of as written data. Just prior to the era though, the bigger drives >> of the horizontal head orientation drives were likely more susceptible to >> a few errant bit flips. >> >> Maybe they should make a high security hard drive with magnets inside >> it on a lever such that when an emergency data erasure was desired, one >> would flip the lever over, and the spinning disc would erase within >> seconds. Then, a reformat might show a forensic examiner no old data to >> examine. Just putting a bullet through the platters works too. > >You need to shred the platters if you want the data to be >unrecoverable. The bullet through the platters stops all but the folks dragging heads across the drive faces by hand. The cylinder pitch would require precision alignment, and co-linear motion to the tracks. It would be a long, arduous process to get even a little data. Sounds like a series of small, encrypted virtual volumes are in order. Then, all the bits looks random, and there is nearly no file structure to find at all. > I'm *sure* there are semi-"photographic" >techniques used to "image" data from damaged discs -- if the >motivation is high enough. Discs with magnetic data recorded onto them exhibit absolutely no physical or optical signifier of what bit they contain. It REQUIRES being read magnetically with the same head type it was written with. It would be very hard. In doubt? Shoot it three times, equally spaced. > We used to do it with *tape* >decades ago... Not on a modern hard drive. We are talking about Gigabits per lineal inch. EVERYTHING about the read head has to be perfect to 'grab' the data. Almost getting right down to the molecules here.
From: life imitates life on 11 Feb 2010 21:57 On Thu, 11 Feb 2010 13:30:43 -0700, D Yuniskis <not.going.to.be(a)seen.com> wrote: >> I trust my optical disks a lot more then the old magnetic media. > >Agreed. My point is, even *those* may not be "safe". Not me. I still have every hard drive I ever owned. That includes a 10MB Tandon original HD for the XT. Back then, it was nearly $100 per MB. Now, it isn't even $100 per TB. W O W !!!
From: life imitates life on 11 Feb 2010 21:58
On Thu, 11 Feb 2010 13:30:43 -0700, D Yuniskis <not.going.to.be(a)seen.com> wrote: > ><grin> Yeah, and Bernie Madoff had "a sure thing"! :> He sent out 12% per year checks to some of his investors. Some of them paid taxes on dividends they never actually received! |