HOW and why we erase a HDD ? (all ways from Sunday?)
  • Selling an old HDD we erase it.
  • Throwing it away we ease it and crush/drill it out. (or burn it up)  (FBI shreds them, they have this game RIGHT !)
  • Re-Purposing any HDD we just  erase it. (Killdisk)  good enough for the average Jane or Joe.
  • Sending the HDD back (or whole PC) back to say DELL and HP and want your content erased. (killdisk  is best.) This OEM even tells you what to do. (ask them)
Why? we do this is simply to erase all old or private content. To start fresh and new and to remove any old data and any viruses on a disk.
A HDD can hide files on the disk, in delete space or slack space or Hidden HPA space.  This graphic below shows a crude HDD format,  and where things hide too.

As you type say  a document the eof marker moves and then overwrites old data. (now you know what slack space means)  There is also, spared out sector space you can not erase with software.


Just files:
We all know know, from 1981 to now deleted files are not really deleted "yet", they are only marked for deletion. (or deleting partitions too) From Peter Norton DOS days 1982  to now.
Here is one site that shows many erase apps but fails to test them (the endless list of apps here, and misses my favorite , Killdisk as it erases HPA space areas too. (unlike DBAN)
I am not talking at all how to erase deleted files or folders, no just the WHOLE DISK HDD CLEANED like in 100% .
The next free application program does not  hurt the disk in any way , it only erases all content even to use the drive later.
There are many ways to do this, the most simple of all is KILLDISK.exe , seen here. (100% safe and free)or the KILLDISK CD booted.)
Killdisk or the like is no good for the likes of the DOD/CIA/NSA. they shred the drives to tiny pieces. (by law) They do not like those old inaccessible sectors now locked out is why.
For most folks just running Killdisk is good enough, for sure reused by you.
Some serious folks, use killdisk then crush/drill/shred/incinerate the drive (not reused sure) Your paranoia and your call what to do next.
SED drives? (if you have one of these and are very modern and can be re-purposed too, ask your HDD maker how?)
This is a complex topic and some drives are SED, self erasing drives, you give it 1 command and it kills all data in seconds (it kills the encryption keys making data useless to anyone)

 Half Joking: "Maxwell's silver hammer...." or use Thermite., or Mr Elon Musk, blast it in to the SUN with a rocket. (or at Mariana  trench deep six it?, 35k feet deep , NO ! do not pollute!))
Use a high speed  grade 1/4" (6mm) drill bit and drill 1 hole through case and platters or go wild with more holes. 

Some folks remove the platters and use 1 Tesla or higher powerful magnets,  to erase the platters then cut  the patters in half. (some future platters are glass and break easy)
How the platters are made are a trade secret, but are metal (alum alloy is my guess and metal plated)
Platters or descrete DISC some need  3Tesla EMP machine to do this ever more.  (EMP machines are very dangerous inside so stay out.)
DO not grind disk platters, they are TOXIC for sure into your  lungs. (do not do that) Use a face mask to drill or burn drives.
See this cool video one guy with huge magnet killing a hdd dead, but he wrecked the heads and that is no proof at all of data tracks erased at all.
I just drill mine, that is my way, my bag.  (drill it 1 time, and smack it 1 time with my 5 lb hammer, (PPE )

Some HDD for sure from OEM's like DELL/HP have the below custom formats that   killdisk removes easy  others do not for sure, DBAN FAILS ( on purpose they say) Dell has Tools and install setup in hidden folders.
  1. I do not like DBAN to erase,  it fails to erase OEM special hidden files by default  (see options page).  (what if a virus lurks there?)  KILLDISK does erase it all. (for a fact and free) KILLDISK.exe
  2. HPA space (Host Protected Area) – DBAN does not erase the HPA   (Dell notebooks hide Dell MediaDirect utility in HPA. IBM ThinkPad and LG notebooks hide system restore software in HPA. & Some rootkits hide in the HPA)

    DCO (Device Configuration Overlay) – DBAN does not erase the DCO either.  If really curious if you have these HIDDEN sector traps, run this for free.

Before condemning any HDD you just bought used, run killdisk on the drive,  first.
Killdisk removes it all. ( but no "APP." on earth can erase the bad sectors spared out by the S.M.A.R.T brain (firmware algorithm) inside the drive. (a.k.a. inactive sectors )
Personally I never run PCs OEM orginnal   or allow  them to have CUSTOM SPECIAL OEM partitions of any shape or kind , no risk of infections there either. (no DCO and NO HPA allowed by me)

 It will erase the drive 100% (no active sectors are skipped !) 
  1. Steps: Are to move the HDD under erase  to the drive docking bay. Seen here. (plug in the HDD to this slot) and then the Docks AC/DC power pack.
  2. Plug in the USB cable to a PC that is running Windows.
  3. Install killdisk.exe
  4. Run killdisk, do 1 pass.   (1 hour for 300GB)   (to connect any  SAS drive and test or erase it here is how,  to  see one failing here "doomed".
The free program "App." called Passmark maker,  Diskcheckup.exe tells you 2 things key, smart not failing and if HPA is present in. HPA space
I run Diskcheckup  and also see if HPA is present. (seen on tab 5)
Do not read this section this is for tech.'s that like details/ facts or wants? deeper second stage erasure. (killdisk first, then crush, hammer, drill or burn)
This section is for others that can go further than me,  that I can not do ($500 is too much for me to waste on silly magnets), post erase testing is way too expensive,  proving data is really all gone is so.

If killdisk fear is real, you can drill , crush or burn up any dicks you want ,smashing all work well. and incineration (1200f degree's. tops my list.
Permanant, magnets Neodymium N52 fail. (huge)
My 3" Diameter   900lb dead pull rated fails hard to erase any HDD in my spares kits. (boxes full of spare good HDD) rated at 1.3T but that is surface rated not at all at say 10mm further to discs inside any drive.

Electro Magnetic Erasure IS a.k.a Degaussing.  (fearing Killdisk are you?)
If your do this and wreck the head doing Degauss the data may sell be on the disks, if the Degauss erase the servo track the data can not be recovered so says top recovery services, for sure HDD made after 2006 year.
There are recovery services that use special techniques or even SPIN stands but there is no for sure way to get your data back and my point here is how effective erase method can be.
The top players here say Degaussing makes the data gone pg.73 or $1,000,000 slow recovery and very limited.  lost servo track is a death deal to data. DDD.
HOW to Degauss?
In most cases , you will not actually use a magnet of any kind to erase any DRIVE, and  a few can be (old tech) with the disc platters in hand, and passed through some 2 Tesla  , 20,000Gauss magnet.
The spec, on most HDD show 1T is needed. or more. Ask your OEM maker this, and varies by maker and age of drive. (evolution of media is not a constant)
In theory? Sure
  • Using 1T magnets fail.  No matter how constructed   some can work (if lucky) with platters removed and direct DISK to magnet surface may work. Even this fails $500 magnet .
  • EMP Machines. EMP degaussing machines are sold and are of dubious value. degauss machine works?.$2500 bucks,  or . (youtube, watch) (.<<<  (lethal do not play here)  
  • Home made electromagnets fail. (I made one that run at 1200 watts that failed) called the MOT.
  • CIA/NSA/DOD shreds their drives or even whole PC's.  (speaks loud this act)  now you know why? above fails. This is put  here to show why above fails, they know it does !

Some HDD makers say  disk platter media is above 8300 Oersted, rated or even MORE. Coercivity,! (see myths here)

 
The enemy is DISTANCE TOO (huge limits)
 
This tells you why most DIY fail for using magnets of any kind,  DISTANCE FAILS hard. (the test is with N52 magnets using a real gauss meter and then blocking objects tested. (shielding means actually rediretion of flux)
. The actual plot varies by the magnets actual H field density. ! (and form factors)


 The Mighty MOT;  2,000 VOLTS PRESENT, IS THIS CLEAR?  (do not bypass uWave ovens door interlocks or you will cook your liver to death.!)
Double lethal Insides. ( learn to work with HV and Magnetron's first then play) (no second chances ok?)  LETHAL it is. (until modified below)
Full PPE here, gloves, goggles for sure. Never risk Eyes or death from electric shock ever.  (forget macho BS)
The MOT Capacitor is not here, so can't shock you. But the old HV coil is lethal. (power applied)
HV = high voltage.
 I tested this and it fails even with ADDED BY ME, 225 turns  of #16 AWG wire on the primary, (did not erase any drive but on some wrecked the head servo and made testing impossible (so is a wrong path to proof !)


Click me to go to web site.
The right side below is  real iron core solenoid, Classic example.

If you try the magnet on the right even in C shape you will fail, to get 2Tesla Gauss, or get injured trying.

The left is air core and useless as a brick for high Tesla.
Some equations show flux deep in the core only what we need is surface gauss and last 10mm distant.  (hint I can not park my HDD inside a solid iron core)  
 
The formula I was taught to calculate the Gauss of an electromagnet was: B=MNI/2R with: B = Magnetic flux in Gauss M= Magnetic permeability of core material N = Number of turns of wire I = Current input in Amps 2R = two times radius (or diameter) of core

Source https://www.physicsforums.com/threads/calculating-gauss-of-electromagnet.672680/


Inverse Square Law,  (best case)
 
I = 1/D²,   I equals 1 divided by ,  Distance squared.    "I"= intensity   
"at 2 times the distance , the field drop by 1/4"  so 10,000 Gauss (1T) is now 2500G (1/4) and useless as tits on a boar hog.
The truth is this is only a crude equation, but does show the effect very well,  the truth is the loss of field varies by the flux density, and form factor of the actual magnet.
what matters most to me is this  rule, to let me take gauss readings accurately.
I made a flux gauss tool and learn fast the above is no lie at 10mm my 1Telsa magnet is weak as a cat.  DISTANCE KILLS The effect super easy, so only EMP  might work if at all. (my conclusions)
 

What is flux what is magnetic H lines of flux look like, ask any kid, this is even covered in BOY Scouts (BSA) even in 1955, dang.
The lines can not be seen but can be measured and with iron filings seen, using this trick, see they are real.
'Magnetic anisotropy'
Even kids in grammar school see that mag lines are not magic but real, seen here. (Iron filings test)
  Why dropping drives is bad. (very simple algebra here)  the faster it falls and the faster it stops, the worse the damage will be. (even a 3 foot fall  can be bad for sure a running laptop)
For now forget G as in GAUSS, not G-force (no not the video card)
Gravity can be an enemy.  (ever fall off a roof?)
as seen here
Try dropping  any HDD 12feet to concrete. this will kill it but maybe not if the FBI gets hold of it, it will kill dead the head servo motor for sure, wrecks heads.


Last and only for pure paranoids out  there:  (I HDD is already dead skip line 1 below)
  • Run killdisk.exe if the HDD is dead, this too will  be dead, and in fact all things will be dead accessing it, or  if only partly dead, some data can be saved,  Google that and is endless ways to recover data. (apps)
  • Remove the lid, mine has "Torx" bit size T4,  now the platters are fully exposed, rip out or unscrew the arm-heads.
  • Then the hub screws and 1 to 4 platter disks are in your hand, trash the old casing it is useless.  5minutes work to get at platters tools in hand.
  • USE a 2 Tesla grade magnet of any kind, and super expensive and MUST get close enough to actual disk hard surfaces to even work at all. 
  • EMP pro grade degausser  fails on many modern drives and no proof after when the servo arm is bent to H3LL.  (zero proof data gone, zero)
  • Next hammer the disks like a mad man,, do not grind any disk ever.  (wear PPE for sure gas mask and  leather not plastic gloves)
  • Fire up the BBQ outdoors or use the BBQ chimney sold in stores. this is a poor mans forge.  when white hot and no smoke,  drop the disks on top, see them melt or even warp.   Do not sniff fumes. Lots of fresh air outdoors only.
  • Last with old discs in hand send only them to the toxic waste center in your town, without the main chassis seen below.  Throw the chassis away next week. (keep them seperated forever)  Max Paranoia. Yah.
Do the  Thermite. way. (overkill I am just of joking here)
DIRECT METHODS, (magnet erasing, drills and fire and brimstone)
Warning #1 Do not grind platters: Platters are often made out of aluminum and coated with Ruthenium alloys, Cobalt-Nickel-Iron alloys, Cobalt-Chromium-Platinum alloys or palladium. ( is a risk not near as bad ias Ruthenium) PPE:
One more way is take just the  platters to the county, recycle center tell them its Palladium and they handle this in very special ways. (erased by you first and no hdd casings submitted)
 Rule number 2 after running KILLDISK, use EMP ways is remove the platters first  and do direct close erasure and not suffer from Gauss field weaknesses so far away .  (the farther you go the more EMP you need and is not cheap doing that)
Myth 1 and 2 debunked. "The magnets in side prove no magnet can erase HDD"  (a 1/2 truth if at all) (they forget those inside are weak, oops)
Fact is the magnets inside are very weak, I tested them  this is true I can buy magnets vastly more powerful from 5000Gauss to 15,000G. (most folks never seen one of those !)

The  motors inside have weak  magnets (motor cup is steel shielded) and folks wonder why those inside mags, do not erase  8300+ Oersted's platters, the answers is simple they are shielded and are weak . (weak to the DISKS)
The photo left if motor hub(shields) and platters removed, the photo right is brush less DC motor in hand, back side, as you can see it is fully encased, if tested for leaking gauss the readings are minuscule. Use a gauss meter and see that .

Now my witches brew tests: (A pro grade compass) LoL !
Below see my hub motor. Just 20mm diameter small.
It does leak a very very tiny  mag field,  Using my compass,  it only causes a 45degree deviation error from North.  (earth gauss is about 0.5G so it seems my motor  leaks,  at 1/2 the earths field.  ) Minuscule leakage. 0.25Gauss.
Tested with compass pointing North and motor case directly against, the compass, at  90deg. East, case to case and needle moves to 45deg.  (worst case) I have no real gauss meter.  (warning 4 sides to erase)
This is my Dell SAS 10k drive,2.5"  the hub is pure steel,   (see before photo)
Test #2 on above casing is the case magical Mu-metal, answer is......NO,  this is aluminum alloy and is not Mu_metal at all,  I put compass on one side, then a magnet on the other side and the compass goes wild.
"Mu-metal typically has relative permeability values of 80,000–100,000 compared to several thousand for ordinary steel." wiki 
But in truth there is no need for this fancy mu-metal  at all . (the disk is immune to all weak fields)
I tested the steel lid and it blocks (means redirects) the field very well.  (steel works see distance chart above)
Distance causes rapid loss of flux intensity Gauss.
Not only that fact but the inverse square law for distance, is not repealed, here.
Track pickup head seen up top on servo arm,  coil center , Neodymium magnet center bottom, and steel pole shoe clamp.

The pole shoe (clamp) of steel shunts the magnet (N/S ) holding GAUSS FLUX, in close confines here.  The field is weak, (on disk side) but can be detected it compass at 6inches and closer. (150mm or less)
The magnet can never ever create much less send 8300gauss to reach the disc platters  ever, even the distance rule on gauss, proves that easy.
 So we do know for a fact this above magnet field does not change the disk data at all. (forever)
The head servo magnets send flux up and down never to the DISK, 90degrees away,) by design. (no Seagate engineers are not stupid)
LAST and best are my fire and Brimstone tests: (SOLVED) and full PPE

This does work below, proof, but simply smashing the disks with a hammer works just as well,  hit hard and  5 to 10 times or broken in half.
 Instead.   (You have this and can melt steel or platters) 1200f is goal or hotter.
Ways and means: (all are overkill and cost but the cheap propane)
 (BBQ hits 1000f, but with chimney more , (lots of air flows into hole there and gets super hot) real charcoal burns way hotter than lump BBQ briquettes and shorter time, more oxygen reaches the real charcoal is why.
Oxy/Acetylene hits : 5800 °F to 6300 °F  (bingo best way for any torch.)
or use MAPP GAS torch (sales ended 2008) torch =  2925 °C (5300 °F) see "MAP-Pro"  3700F seen here .
Bluefire propane torch,  1450 ℃, 2640f
or watch a redneck , forge maker. (can be done safely) (RIP !!)    the coffee can forge  1800F= Bright cherry-red, wow but what is a coffee can?, today. plastic junk. (for sure melts aluminum disks. One more by hand.
The choice of forge is cost or how long it lasts. Your call.  Best for safety have regular cooling fan blow AIR from say 3 feet away, no fumes  up nose. (for sure working any unknown alloy)
The winner (price) is just a dirt cheap propane torch burn up.  it melted. No more data, gone forever. Off hand I'd say nothing works better. (platter destruction) Never do this Indoors, do it in open yard or field no dry grass.(etc)
BINGO 100% effective.    Winner of the NO DATA contest !   (some disks shatter hit, we try that first, omg tiny chunks ) I have  large hammer, and smack it 5 times and they shatter. (I do that first and mostly win)

Ruthenium if here and can be is very toxic, use lots of fresh air, do not breath fumes from this element.
Why PPE is needed? (Personal Protection Equipment ,googles gloves and gas mask too!) and for sure never inside a building play here, be out side with lots of fresh air even a fan behind you, is wise.
"Ruthenium compounds are encountered relatively rarely by most people. All ruthenium compounds should be regarded as highly toxic and as carcinogenic. Compounds of ruthenium stain the skin very strongly. It seems that ingested ruthenium is retained strongly in bones. Ruthenium oxide, RuO4, is highly toxic and volatile, and to be avoided"  Introduced in 2006 for higher density platters.

Do not do this at home for sure inside any building, do it only out doors with fan blowing on BBQ top target disks, and never point the fan to you, ever.
I did this only to prove It can be done with tiny efforts and tools cheap and can melt 1 platter easy.

One can simply remove the disks and smack them 10 times with a hammer, and no more data is there, period, none that any human can find.

To dismantle a drive is 15min work, with in hand correct "Torx" bit. and  15min with the torch.  (a drill bit 1/4:" or 6mm ) is 5min. job.( and I drill mine)

Show and tell?  14 screws to get disks in hand.  (lid off first the hub ring 6 screws , done)  This is just  proof of concept, all I do is drill disks for trash , disks too old or failing, I never sell disks.
(warning 4 sides to erase 2x2 top side and bottom sides!) This one has 4 heads. not 2.  Do no lick the platters for sure not grinding them ever.
NOTE the servo mag, has confined MAG field in the pole shoe keepers. CONTAINED GAUSS and concentrated there to the COIL seen just to the right. (this trick makes the head arm faster but no huge GAUSS leaks out.) see?
version 3  7-1-2020  (redone from scratch) to show all possible ways, so all folks reading this can decide what is best for them. Killdisk wipes one pass of data = 0 is good enough for most folks.