labs.
layer
ato
size,
element.
their
be
Hmmm. I do not see any connection between case dimension flexure and = head positioning. Flying height is off into the weird end of gas vs head dynamics. Head tracking / servo is mostly in control of the lateral positioning. So there is only minimal connection between case dimensions and head tracking performance.
YMMV
?-) =20
Easy. The base plate that supports both the motor and voice coil head positioner has to be very rigid. If there's a bend, the head will not be parallel to the platter and fly at a slight angle. If there's expansion or contraction, the head will move a corresponding distance across the tracks. Mounting the head on a spring allows some angular movement compensation. The initial boot calibration routine, finds the location of the inside and outside tracks. It then divides by the number of tracks to get the approximate track position. Imbedded servo data in the track, or straddling the track, provide precision alignment.
At 30,000 tracks per inch, that's 33 micro-inches or 0.83 microns per track.
First, let's look at temperature. From a handy 3.5" disk drive, the distance between the motor and head positioner spindles is about 50mm. Coef of linear expansion for pure aluminum is about 0.0238 mm/m/K. For a one degree change between spindles, that's: 0.0238 mm/m/K * 50mm / 1000mm/m = 0.0012 mm = +/- 1.2 microns dimensional change or: 1.2 microns / 0.83 microns = 14 tracks for every degree C change.
I don't know how to calculate the HDA case deflection due to pressure changes. With the pressure equal inside and outside the HDA on a conventional drive, there's no deflection. However, with a sealed HDA under pressure the deflection would be approximately the equivalent of putting a heavy weight on the drive base. Assuming about 0.2 atmospheres differential pressure: 0.2 * 14.7 psi = 3 psi and about 18.6 in^2 (120cm^2) surface area for a 3.5" drive, that's" 3 lbs/in^2 * 18.6 in^2 = 55.8 lbs. The 55.8 lbs does not act on one point, so you can't just plop a 55.8 lb weight on the drive and watch the case cave in. It has to be equally distributed over the baseplate. That could best be simulated by dividing the 55.8lbs into smaller weights. For example, 18 weights, each 1 in^2 base area, and 3.1 lbs each, should suffice. I have no idea what deflection might be achieved, but I predict that it will cause the drive to recalibrate and possibly crash the head.
At 0.6 atmospheres, it would be much like standing on the drive.
Now do you see why sealed drives are a problem?
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
The performance gain is in the float height being closer. That is why the single gas is used. The only thing case flexure impacts is terminal internal case pressure. It cannot vary much IF that variance causes failure modes. So any design they arrive at will consider such things.
They also have to be perpendicular recording mode. That is the only way they will get those lineal write densities.
head
dimensions
That is an assumption that may not be valid.
Not so. The springiness of the support arm and the fluid dynamics of the head - gas - disk will provide proper parallel flight.
Maybe that is the case, maybe not.
There are other methods as well, such as servo between blocks.
On a fresh corpse of a drive i measured about 80 mm.
Make that maybe 1.4 tracks?
Which means nothing in presence of modern servo techniques that main = track far better than 0.83 microns.
And it shows. Start with Young's modulus to get a size change with applied pressure. Second the delta pressure will be minimal to reduce = the helium reserve needed. Of course that reserve tends to limit drive life.
Not at all. Gas pressure acts uniformly per unit area, thus geometry matters.
In what direction? On what surface?
Not at all. Sit down and do the physics, starting with defining your assumptions and getting your units and force directions straight.
Your assumptions and math are way off.
I see issues, but none related to your presumptions.
?-)
Do you know if the 2.5" drives were also affected? I ask, because Agilent was giving away free 2.5" replacement drives pre-loaded with the OS and scope hardware. They were IBM Travelstars, IIRC.
IBM were the leaders in MR head tech.... then. They were the guys who achieved the highest lineal write densities.
I do not know if Hitachi continued to lead in that area.
Who made these 7TB drives?
Oh, and most are going with SAS and 1.5 inch 10k rpm and 15k rpm drives now. Most of those end up placed into the 2.5" 'laptop drive' form factor. I have seen a few that were other form factors, and a couple strange ones that were double height 2.5 inch. Must have been more platters in there, but I thought they were moving away from that paradigm. Oh well. The double height 2.5" SAS drives are common now. Weird choices the industry makes. The reason must be because "blades" and the like can handle more depth, and "why waste it?" mentality prevailed.
The 'blade' realm is business oriented, so you will see their demands and desires driving the whole industry.
The 3PAR rack from HP is a pretty freakin' nice storage realm. All fiber interconnected, and way beyond RAID management hardware. Screamin' gear.
I think Travelstars had no problems--I have a whole bunch of them, and none has died to date.
I standardized on the Thinkpad T42P/T43 a few years ago--I have probably
10 of them in various states of repair. I don't watch movies, so there's no point in losing 2 inches of screen height for nothing. They almost all have Travelstars.Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
I am not a diver, I just like cool watches. But here goes: its called "sat= uration diving" where divers spend long periods of time underwater breathin= g a mixture of helium and oxygen. By long periods I mean days, living in a= helium/oxygen environment. The high pressure helium gets inside the watch= slowly over a long period of time but when the diver surfaces (quite rapid= in comparison to the time spent under) the internal watch pressure is much= greater than the external atmospheric pressure. I think a watch without a= purge valve can actually burst, again I am not actually a diver. I do jok= e about how I got tired of my watches exploding.
--Dan
The 2.5" laptop Travlestar drives did not have problems. I saw some failures, but they were normal age related failures after 5-8 years of operation. I still have some older laptops around with the original Travelstar drives.
The 3.5" IBM Deskstar was another story. Between about 2000 and 2005, I sent about 100 of these drives to the recyclers. I wish I had taken photos. The best I can do is this photo of my working drive collection:
If you look carefully, you will not see exactly one Deskstar drive (on the left, below the yellow note pad). However, the 2.5" pile in front is about half Travelstars. Incidentally, the drive collection is now aobut 4 times as large.
Drivel: Here's a drive with a built in death wish:
True to its name, it died young.
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
Assumption, the mother of all screwups. I think it's valid. If the drive HDA was made from rubber instead of aluminum, and it could be inflated like a balloon, the relative motion of the platters and heads could be observed. My contention is that when the supporting structure bends, things will move around until the drive no longer functions or there is a head crash. The only questions are how much will it move and can the drive compensate by recalibration?
Maybe, but I don't think so. With a flying height of 5 nanometers, and a head width of about 2mm, it's not going to take much angular tilt to make the head hit the platter. The springs will be a big help, but are limited to a spring tension not to exceed the lifting force of the head. Were the spring tension greater than this force, the head would hit the platter.
Some interesting hard disk drivel that Google found:
In the beginning, there was a seperate platter surface for the servo track. Later, the servo tracks were located between the data tracks. About 20 years ago, the servo data was with data on the same tracks.
That's the current standard.
Oops. I didn't have a drive handy so I scaled it from a JPG on the screen. Remind me not to do that again.
Oops 2.0. I thought 14 tracks/degree was a bit large. Still, it's a clue as to how sensitive the drive positioning is to temperature and possibly HDA pressure.
I don't really know how well an embedded servo tracks. In theory, it could probably compensate for almost any head to track positioning error. Reality might be different. I'll do some Googling and see what turns up.
Thanks. The helium pressure will also be low because it affects the head flying height. Too much helium pressure and the head could be flying at the same height as in air, thus reducing data density and drive capacity.
Oops 3.0. I calculated for a single side instead of the entire surface area.
I goofed. Assuming minimal drive height, the entire inside surface should be about twice what I miscalculated. Therefore, the pressure should be half or about 28 lbs.
I'll crawl back in my hole, conjur some numbers, and see if I can salvage my guesswork and miscalculations. I just hate it when I screwup this badly.
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
That is better. IIRC recalibration is to track 0 near the outside of the disc.
the
Perhaps. The physics is already below proper Bernouli effect, which = draws the heads a bit closer but severely limit how close. Head tilt would be torqued back towards parallel. Well, i expect so depending on geometry.
Ah yes embedded servo.
track
Be sure to share.
the
life.
Joining the club today i see. I like to claim my ancient ancestors for founding members. Many decades member myself.
?-)
There are ways around that; I've seen lasesr tubes with a secondary chamber of He gas, that leaks into the primary volume about as fast as the primary volume loses gas through its envelope.
And, some Ne lamps (actually a Ne-He mixture is used) can retain enough helium for decades of life. Four decades, so far, for the lamp in my electric frypan.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.