Dan's Data letters #33Publication date: 3-Mar-2003.
Last modified 03-Dec-2011.
When browsing about looking for new hard drives, I noticed a lot of the spec pages list both the operating and non-operating impact they can sustain. While I've seen this before, I never thought much as to what the numbers actually meant.
Let's say, for example, the drive is a WD2000JB. It's listed as taking a non-operating shock of 200G for 2ms. What exactly does this translate to as far as dropping height goes? Since I'm pretty sure it makes a difference, it's also listed at 1.32 Pounds. Also, a handy formula for figuring it would be helpful for future reference.
The shock - or deceleration - an object suffers when it's dropped onto a surface doesn't just depend on how fast it's moving when it hits, which can trivially be worked out from the standard gravitational acceleration of 9.8 metres per second per second, in situations where air resistance is negligible, as it is when you're dropping relatively dense things relatively short distances. You also need to take into account the consistency of the surface, and of the object. The harder the surface and the object are, the greater the shock, for a given impact velocity. A hard drive is more likely to survive a fall onto shag pile carpet than onto concrete.
Assuming the object you're dropping is negligibly squishy (which, for hard drives being dropped in domestic circumstances, you can pretty much do), and that you drop it from two metres up (rather a long way), and that air resistance isn't a factor, then the average deceleration if the floor's squishy enough that the object punches into it one centimetre in the course of coming to a halt (and then, probably, bouncing back; elastic collisions are another subject) will be about 180 gravities. One centimetre of deceleration distance is not outrageous, for a fall onto fairly luxurious short-pile carpet. Office carpet tiles with no underlay, though, will give you a shock that's easily twice as great, and the shock from a two metre fall onto concrete will be much higher again.
The above calculation ignores any spin or unevenness of landing, though. It assumes you're dropping the drive perfectly flat onto the ground, and that one corner won't hit first, and that the drive won't be turning. In the real world, one corner will probably hit first, and the drive will be turning; it'll rattle onto the floor, giving multiple, weaker shocks.
Given all of the variables involved, there's not a lot of point digging out your elementary physics textbook and working out the impact speed for different drops; the numbers you get won't be terribly informative, because of all of the other factors you're not taking into account.
LED living room
I was wondering if you could offer some assistance. My wife and I are doing up our house a little (she's focusing on the colours and rooms etc, while I just want to spend up on electronics and network/audio wiring) :-).
Anyway, ever since I saw your comments on the Luxeon Stars, I've been thinking of embedding them in my living room ceiling. However you mention their heat output. I was wondering if you could tell me what considerations I'd need to take to light my living room with these. I frankly don't know the first thing about it, so any help you could give would be great.
To light the room to full, normal, read-a-newspaper-anywhere level would take rather a lot of Luxeon Stars. Assuming you're using them as downlights, then if you use five watt white Star/Os, with the integrated lenses that give them roughly 51mm-downlight-width beams, you'd need, I'd say, about eight of them for every 50 watt halogen downlight you'd otherwise have used.
Unremarkable halogen lamps can deliver about 18 to 20 lumens per watt (the lumen is weighted to account for visibility of light to humans, so light that's further away from our eyes' sensitivity peak in the green area of the spectrum gets a lower score); regular household incandescents are down around 16 to 18. White Luxeon Stars manage around 23 lumens per watt. Green and blue Luxeons score much higher, and coloured LEDs in general can make great accent lights, but they're no good for main lighting, unless you want your living room to look like a horror movie set or discotheque.
Luxeon Stars are more efficient than smaller LEDs, and the blue-white light of white LEDs is more visible. But they're still not streets ahead of halogens in the light-per-watt stakes, and certainly not in the light-per-dollar department, at least until you've replaced your halogens quite a few times.
That said, there'd be no major heat management issues. A five watt Star, running at full power, produces five watts minus its luminous output worth of heat. A few watts of heat is not hard to deal with; stick the LEDs to a piece of aluminium angle with metal-filled epoxy, for instance, and they'll be more than adequately heat-sunk. Halogen downlights are a much worse problem; they don't care about being hot, but their dichroic reflectors are specially designed to reflect visible light downward and let heat through; dumbly installed, downlights can start fires.
Given the rather exciting expense involved, I think you'd do better to use the LEDs for accent lighting, or scatter them around for mood lighting, and stick with incandescent or compact fluorescent as your main everyday (or everynight, as the case may be) light source.
I'm hopeful of making it in the R/C field but cant find any websites or as such information on technical aspects of those darn nifty Mini-Z Racer [as reviewed here] motors, as I need to experiment with them, and I'm on a tight budget. Do you know of any sites that might be of help to find out even the physical measurements of them?
Mini-Zs use, as standard, flat-type 130-size motors, which are available as "hobby motors" from better electronics stores. You can see quite a few 130s in the first section of Mabuchi's product list, here. Click any of the flat-type motors with "130" in their name for full specs.
Where are the magnets?
After reading your bit on rare earth magnets, I decided to open up an old Seagate ST-412 HDD I had lying around. I think this holds about 10mb. The thing is a dinosaur.
Anyhoo, I was trying to find some sort of uber-powerful magnet in it, and, for the life of me, I can't!
Is there one in ultra-old hard drives, or do they use some sort of other technology? I saw a DC motor in there to move the read/write head. Is this what they use instead of the magnets?
Yes, you've got a stepper motor drive. Stepper motors can be useful in their own right, but you're not going to be able to salvage any magnets. For magnets, you need a voice coil drive; the first voice coil drives for PCs hit the market in 1986.
Stepper motors move one tightly defined angular step for each pulse that's sent to them; they're therefore pretty good for drive head positioning, provided there's little slop in the motor-to-head drivetrain (they generally use a wrapped spring-steel belt arrangement), and the drive's used in the same orientation in which it was formatted (otherwise gravity can pull the heads far enough out of alignment to cause problems).
Voice coil drives, in contrast, have much finer positioning ability, and the voice coil can change the head position practically infinitesimally in response to feedback about how well the heads are lined up with the track. Voice coil drives are insensitive to alignment, and have fewer moving parts.
The big platters from ancient drives can still make nice wall decorations, though!
Recently I upgraded the heatsink/fan on my Athlon machine and when I turned the machine back on, all of the fans came on, the HDs powered up, the power LED came on, but nothing else happened. No video or beeps. So I did the usual routine of reseating the CPU and heatsink, which also didn't work. After that I took everything out except CPU, memory, and video. I tried like that and with a HD connected. I kept getting the same result, so finally I gave up for the night.
I work as a lab assistant in the computer department of the college I attend and I was able to borrow a PSU from the vast amount of spare parts there. I brought it home, popped it in, and everything powered up fine. Took it out, put the old one in, nothing. Put the borrowed one back in, powered right back up. I purchased a new PSU and am currently writing this mail from the same machine with the new PSU in it.
My question: Is the 'bad' PSU in any way salvageable? The system had been running fine for over a year before I powered it off to change the heatsink, and like I said, it still turns on the fans/HDs/LEDs.
There's a good chance it's rescuable, but that doesn't mean it's worth rescuing. Pretty much whatever's wrong with it (it seems to be powering up one or more of its rails, but not all of them) is likely to be repairable, but PSUs are so cheap these days that practically nobody bothers fixing them.
If you know a waste-hating person who has a multimeter and a soldering iron and knows how to use both of them, then you could perhaps get a repair done for less than the price of a new PSU.
Standard religious war #3
I'm basically known as the computer guy around my office and I was recently talking to a friend of mine who is considering purchasing a new laptop. She's very artistic and was planning on purchasing a Mac.
Seeing as I'm a red-blooded American, and not some commie pinko hippie, I politely suggested that she might want to consider a PC laptop. So, she asked me what the difference was. After relating the items of proprietary hardware, limited selection of software and higher price, I realized that none of those particularly apply to her specific situation. So the question comes down to: What makes a Mac so great when it comes to graphic and music software? And, is it so much greater as to justify getting a purple computer?
There's nothing about Mac applications, per se, to make the Mac a better platform (there are some apps you can only get for the Mac, but pretty much everything mainstream is available on both platforms and works the same on both, too). It's not so much the hardware, either; a Mac is less likely to have hardware problems than some bargain basement clone, or a tweaky machine full of bleeding-edge brand new parts, but a competently assembled PC made from good components won't give much away to any Mac in reliability, and certainly not in performance. The big deal about Macs, rather, is the operating system and its user interface.
Mac OS, generally speaking, is a more logical and consistent way to tell a computer what to do than is Windows.
Windows is riddled with grandfathered-in weirdness and things that look the same but work differently; the current Windows flavours aren't as bad as the old ones, and those of us who've used Windows for long enough can deal with it quite well. But Mac OS was and is a better system for people who just want to get some darn work done. And Mac OS X, with its new BSD-based design, is making considerable inroads in the serious geek market.
People often refer to Mac OS as being more "intuitive", but that's a bad word to use. Practically nothing that's described as having an intuitive interface actually has one, unless you widen the definition of "intuition" to cover all of the interfaces that lots of non-computer-savvy people already know how to use. There's nothing inherently intuitive about turning a steering wheel anticlockwise to turn the car left, for instance; everybody just learns that you look at the top, or more distant, edge of steering wheels and tuning knobs and various other such controls when you want to know which way to turn the thing to get a particular result.
A computer interface with a knob-thing that worked the other way from this might often be called "non-intuitive", but it'd be better to call it "inconsistent with commonly expected behaviour".
Consistency is only vaguely related to logicality, as well; there are various programs (for Windows, mainly, but for the Mac as well) which have interfaces utterly different from the standard one provided by the OS, and maybe from everything else you've ever seen, too. You can argue that they're not necessarily inconsistent with expected behaviour, but that's only because you don't know what the heck to expect from an oval window with eight unlabeled buttons and a contoured slidey thing.
Such interfaces may be deeply mystifying for new users, but can also be perfectly logical within themselves, and may even be elegant once you learn how to use them. The label-less interfaces for the Kai's Power Tools graphics plug-ins suggest themselves here.
The Mac-versus-PC holy war will never end, unless one or the other platform dies out, and it'd probably endure for a while even then; gaze into the staring zombie eyes of the Amiga, for instance.
Most of the debate in the platform-versus-platform area is of very low quality, and founded on ignorance of one platform or the other, or, often, both.
Macs are great. If you can afford a Mac that's as fast as you need (there's a thriving second hand Macintosh market, which is a good place to look if you can live without the latest and greatest), and if there are Mac OS versions of all the software you want to use, then Macintoshes are perfectly good options.
The curse of the bare LCD
I recently purchased a new Sharp LCD panel, model LM10V335 [for which there's a PDF format datasheet here]. It was a spur of the moment purchase for $10 (so if it ends up being worthless, at least it wasn't EXPENSIVELY worthless). I believe it is supposed to be used for a laptop. Now I'm wondering if I can actually use it. Is there any way to get information to this panel? Say, by RCA or S-Video, or maybe even a standard monitor connection?
It's possible to buy controller widgets that accept some common video input or other and can drive an LCD panel - see here for a selection, although I suspect none of those ones will work with your screen. These gadgets aren't cheap, and neither are they plug-and-go solutions - you're looking at at least $US200, plus power supply, plus case, plus time. If there's a cheap source of pre-made LCD driver boards, I don't know where it is.
For most people's purposes, all a bare panel like the one you bought is good for is replacing a busted laptop screen.
I live in India. I recently brought 2 R/C cars at a garage sale with the undiminished human positive hope that I could make them work (both do not have transmitters). Now I have a livid wife, not to mention a SAD 9 year old son.
Let's face it, I am no nuclear physicist, and the only electronics I know is to put the batteries the right way up. Can you help me to locate the right transmitters, or could I use a "one for all" transmitter?
Also, since the cars are Chinese made, I find it impossible to locate a circuit diagram on the Net. Any tips?
If you'd bought proper hobbyist-type R/C cars, the kind that you build from a kit (sometimes they come pre-built, but that just means someone else has built the kit for you; you can still take them apart to repair them), then you'd be fine. Buy a couple of cheap transmitters (hard to find by themselves, but under $US50 with a receiver and a couple of servos, and often available second hand), and away you go.
Unfortunately, it sounds as if you've got the cheaper non-kit kinds of cars, where all of the electronics are probably integrated into one soldered-together lump. They may be useable with a standard transmitter if you find out what frequency they use (you'd have to take them apart to read the soldered-on crystal on their receiver board to find out), but I'm afraid there's a good chance they won't be.
Fortunately, "proper" R/C cars should also be quite easy to find on the second hand market, so with any luck it won't cost you too much to make amends to your family!