Jump to content

Welcome to MSFN Forum
Register now to gain access to all of our features. Once registered and logged in, you will be able to create topics, post replies to existing threads, give reputation to your fellow members, get your own private messenger, post status updates, manage your profile and so much more. This message will be removed once you have signed in.
Login to Account Create an Account



Photo

Gs and g's and internet noise

- - - - -

  • Please log in to reply
12 replies to this topic

#1
jaclaz

jaclaz

    The Finder

  • Developer
  • 14,677 posts
  • Joined 23-July 04
  • OS:none specified
  • Country: Country Flag
Videoripper has just posted here:
http://www.msfn.org/...o...807&st=2475
an actual photo of a hard disk label where it is printed in large, friendly letters:

Caution, product warranty is void if any seal or label is removed, or if the drive experiences shock in excess of 300 Gs


I got curious in what 300 Gs (which should actually read as g -you normally don't put the plural on symbols of units of measurement ) do represent in layman's terms.

There are quite a number of posts in forums around where mostly inexperienced people threw in hearsay, absurd calculations, semi-random numbers and what not.

A quick check on Wikipedia clears some aspects:
http://en.wikipedia.org/wiki/G-force
http://en.wikipedia....rive_protection
http://en.wikipedia....i/Accelerometer
http://en.wikipedia....ock_(mechanics)
http://en.wikipedia..../Jerk_(physics)
http://en.wikipedia....act_(mechanics)
http://en.wikipedia.org/wiki/Impulse
whilst documents from the actual manufacturers, like these:
http://www.dell.com/...lprotection.pdf
http://www.seagate.c...B585_gforce.pdf
once stripped off the fluff, say next to nothng useful.

Anyway I seem no to be able to find ANYTHING giving an adequate answer to this question:

Can you provide a common practical example of a situation where a hard disk will experience a shock of roughly 300 g's?
(backed up by some calculation OR test data)


This document is finally clearing something:
http://www.memsic.co...ccel Primer.pdf
though knowing that

The acceleration experienced by a laptop if it is dropped from a height of three feet onto a
concrete floor.

equals:

100-2,000g

can hardly be called "narrowing" the problem. :w00t:


Would anyone REALLY (which means NOT "from what I remember from physics at high school, if you divide space by mass and multiply by PI you get ....." kind of thing) understanding the matter provide some calculations/examples?


jaclaz


How to remove advertisement from MSFN

#2
MrJinje

MrJinje

    Toolâ„¢ Developer

  • Developer
  • 1,067 posts
  • Joined 14-October 09
  • OS:Server 2012R2
  • Country: Country Flag
in regards to this conversation I think that the lawyers on the mfg side will argue that 300g is the accepted limit to break a certain internal mechanism and if said mechanism has been broken it proves a fall of 300g has occurred and it is not covered under warranty.

If you are looking into how a fall can have different effects on an object. Imagine the difference when an item falls flat (like buttered bread), equally absorbing the shock over a large surface area, versus a drop where the HDD might landed perfectly on it's corner. (with virtually no surface area) That to me sounds like it would experience the same G-force while falling, but the effects of the landing (smaller surface area), would result in more shock being exerted and at a different angles during the impact. That could be the difference, it's not just how fast or far it falls but also how it lands.

Like how circus high divers walk away after landing nearly horizontal in only 12" of water. (think belly flop) But that same jump would break both their legs if they went feet first.

#3
jaclaz

jaclaz

    The Finder

  • Developer
  • 14,677 posts
  • Joined 23-July 04
  • OS:none specified
  • Country: Country Flag

in regards to this conversation I think that the lawyers on the mfg side will argue that 300g is the accepted limit to break a certain internal mechanism and if said mechanism has been broken it proves a fall of 300g has occurred and it is not covered under warranty.

Sure, but the question is "how much" is 300 Gs?

If you are looking into how a fall can have different effects on an object. Imagine the difference when an item falls flat (like buttered bread), equally absorbing the shock over a large surface area, versus a drop where the HDD might landed perfectly on it's corner. (with virtually no surface area) That to me sounds like it would experience the same G-force while falling, but the effects of the landing (smaller surface area), would result in more shock being exerted and at a different angles during the impact. That could be the difference, it's not just how fast or far it falls but also how it lands.

Like how circus high divers walk away after landing nearly horizontal in only 12" of water. (think belly flop) But that same jump would break both their legs if they went feet first.

Good, exactly the kind of answer I was trying NOT to get. :)

I see a drive falling (and landing) in the same manner a brick would, no belly flop and no water.

I'll try again, given that:
  • a drive won't do belly flops or whatever
  • usually in houses and offices there is NO water underneath, but a floor
  • for what it matters a brick will behave exactly the same

How many g will the drive brick experience when falling from three feet on the floor (we can have three answers/ranges, say, concrete/ceramics - wood - carpet)?

We are already supposed to know that the answer(s) is/are in the 100-2,000g range, but we have this limit of 300 that we must place somewhere.

jaclaz

#4
VideoRipper

VideoRipper

    CSI Eastwood City

  • Member
  • PipPipPip
  • 330 posts
  • Joined 11-February 10
  • OS:none specified
  • Country: Country Flag
Would this page help you perhaps? :lol:

Greetz,

Peter.
To C the BASIC things in life, use Delphi

#5
jaclaz

jaclaz

    The Finder

  • Developer
  • 14,677 posts
  • Joined 23-July 04
  • OS:none specified
  • Country: Country Flag

Would this page help you perhaps? :lol:

Greetz,

Peter.


Yep. :)

Let's assume we have an EMPTY (all 00's) hard disk, let's say a 7200.11 ST500320AS 500 Gb:
http://www.seagate.c...mp;locale=en-US
With a:

Weight (typical) 543g

The typical is because some people ask this kind of questions also:
http://wiki.answers....o_a_80_gb_space
and the 00's is because other people write these kind of things :whistle::
http://www.911cd.net...o...=21827&st=7

Also, we know that 1 g=9.80665*m/s^2
and the nice calculation you provide gives a result in Newtons, and we know that 1 N= 1 Kg*m/s^2
http://en.wikipedia....i/Newton_(unit)

Now, let's fake that we also have an EMPTY (all 00's) 7200.11 ST3320613AS 320 Gb
http://www.seagate.c...tSpecifications
that weights:

Weight (typical) 405g


And that we use EXACTLY 52g of duct tape:
http://en.wikipedia.org/wiki/Duct_tape
to keep the two together.

Now we have a mass of exactly 1000 grams=1 kg that should simplify the calculations. :yes:

Let's also say that we drop the bundle from exactly 1 m (slighlty more than the original three feet) :angel

We have that to obtain:
100g=980.665 N d must be equal to 0.01 m <- this is 10 mm
300g=2941,995 N d must be equal to 0.00333 m <- this is 3 mm
500g=4903,325 N d must be equal to 0.002 m <- this is 2 mm
2000g=19613,3 N d must be equal to 0.000499 m <- this is 0.5 mm

Can you give me typical d values for:
  • concrete/ceramics
  • wood
  • carpet

And confirm that the above calculation is correct? :unsure:

jaclaz

#6
Ponch

Ponch

    MSFN Junkie

  • Patrons
  • 3,308 posts
  • Joined 23-November 05
  • OS:none specified
  • Country: Country Flag
The acceleration depends entirely on the deformation of the disk at impact, supposing the concrete floor does not give any relief.
But you could put it that way; tie both your Seagates to a 1meter rope and make them rotate at 3255rpm, that is 300g for your HDDs. At that point, you're like holding about 3 tons with your arms ...if any left. Call that a real life situation if you want.
The formula involved here is ("centrifuge") F=-M.R.(w^2) involving force, ray, and rotational speed,
so if F=M.a, a=R*w^2 and w=sqr(300.g)=54rotations/sec. I think.

#7
jaclaz

jaclaz

    The Finder

  • Developer
  • 14,677 posts
  • Joined 23-July 04
  • OS:none specified
  • Country: Country Flag
@Ponch
Yep, but it is a bit unlikely to happen in real life.

Not that a lot of people don't normally tie their hard disks to a rope and start spinning them like mad, only that very few will stand the three ton pull. ;)

A common thng that may happen in real life is:
  • the drive is on a table or bench (around 3 Feet high)
  • you, your mate or your pet make it fall to the ground

If the data calculated is correct your drive is dead. :w00t:

So, instead of "if the drive experiences shock in excess of 300 Gs" they could well write "if the drive falls from your table", which would be, I presume, more easy to understand. :whistle:

An ever better one may be "if dropped on a hard surface from more that 30 cm".

And the moral or lesson in this story is:
the Japanese have a clear advantage over us westerners, as they normally use:
  • lower tables
  • tatami carpets under them

:P

jaclaz

Edited by jaclaz, 15 February 2010 - 03:07 AM.


#8
Zenskas

Zenskas

    Advanced Member

  • Member
  • PipPipPip
  • 484 posts
  • Joined 07-May 08
  • OS:Windows 8.1 x64
  • Country: Country Flag

And the moral or lesson in this story is:
the Japanese have a clear advantage over us westerners, as they normally use:

  • lower tables
  • tatami carpets under them

:P

jaclaz

Indeed. But they don't keep old tech as long as everyone else anyway so a broken drive wouldn't matter so much :sneaky:

[Passive Gamer] Silverstone TJ08-EW | Intel Core i7 3770S | Silverstone Heligon HE02 | ASUS Maximus IV Gene-Z/Gen3 | 8GB Kingston HyperX Fury 1866MHz | Samsung 840 EVO 250GB | ASUS GTX 750 Ti Strix OC 2GB | Antec Neo HE 550W |


#9
uid0

uid0

    Advanced Member

  • Member
  • PipPipPip
  • 357 posts
  • Joined 12-June 06
I bought two drives recently, and dropped one on to concrete from about 5 feet.
The un-dropped one worked fine, the other exhibited the expected "tick of death".
Next time I'll take a bag :angry:

#10
jaclaz

jaclaz

    The Finder

  • Developer
  • 14,677 posts
  • Joined 23-July 04
  • OS:none specified
  • Country: Country Flag

Just for the record, newish Seagate drives have now a new limit of 350 Gs.

A 16.67% completely unneeded enhancement :unsure:.

 

jaclaz



#11
submix8c

submix8c

    Inconceivable!

  • Patrons
  • 4,385 posts
  • Joined 14-September 05
  • OS:none specified
  • Country: Country Flag

Slightly O/T but relevant (I bought a new one with the flyer that said the same) -

http://www.zzounds.c...--MAC1604VLZPRO

 

Quality: Well lets put it this way, you can stand on it, you can jump on it, you can drop it (caseless!) and you can smack it when you get angry because your mix aint working and she keeps workin.

Uhhh, only because the board is a single flat circuit (no individial upright channel boards) and that is sheer nonsense. They DO fail with enough G-force (stomp on the knobs). ;)


Edited by submix8c, 13 March 2014 - 06:47 AM.

Someday the tyrants will be unthroned... Jason "Jay" Chasteen; RIP, bro!

Posted Image


#12
pointertovoid

pointertovoid

    Advanced Member

  • Member
  • PipPipPip
  • 466 posts
  • Joined 16-January 09

Hi old friends, nice to see you again!

 

The deceleration in a hard shock is nearly impossible to evaluate, because it relies completely on how hard the shock is (I made crash-tests in my last job). So 350G versus 300G mean nothing. And for the destructive effect, the speed variation can be as important as the deceleration.

 

But 300G is a soft shock. 2,000G would be a more usual one, like metal falling from 1m height on concrete.

 

In other words: disk drives shall not fall on the ground.



#13
jaclaz

jaclaz

    The Finder

  • Developer
  • 14,677 posts
  • Joined 23-July 04
  • OS:none specified
  • Country: Country Flag

Hi old friends, nice to see you again!

 

The deceleration in a hard shock is nearly impossible to evaluate, because it relies completely on how hard the shock is (I made crash-tests in my last job). So 350G versus 300G mean nothing. And for the destructive effect, the speed variation can be as important as the deceleration.

 

But 300G is a soft shock. 2,000G would be a more usual one, like metal falling from 1m height on concrete.

 

In other words: disk drives shall not fall on the ground.

Yep, we already got to this conclusion, but with some actual math behind, see here:

http://www.msfn.org/...se/#entry908569

http://www.msfn.org/...se/#entry908704

AND it was asked to provide some actual calculations, not the usual "vague" stuff, like "soft shock" or "usual shock". :whistle:

 

The whole point that originated the thread was that apart the "wrong" use of the G symbol instead of g, and using the plural for a measurement unit (i.e. like the only two mistakes an engineer/technician can make when writing a measure, that were BOTH made by the good Seagate guys ;)) the info provided is meaningless as real life examples are not provided.

I mean ;):

This thread is intended to be read on a display delivering no less than 25 LMs, or, if you prefer, no less than 2*(2*pi1/2)2 CDs :w00t:.

 

jaclaz


Edited by jaclaz, 19 March 2014 - 12:29 PM.





0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users