Dan's Data letters #143Publication date: 3 April 2005.
Last modified 03-Dec-2011.
I thought you might be interested in this, if you haven't seen it already.
Unsurprisingly, quite a few people have e-mailed me about the new Toshiba batteries, though you were the first.
(You don't win anything.)
Note that Toshiba's "prototype 600mAh cell" (which is presumably a nominal 3.6 volt unit like every other LiI cell out there) has really large contact tabs - which it'll bleeding well need, if it can actually be charged that fast.
80% of is 600mAh capacity delivered in 60 seconds, assuming 100% efficient charging (which is impossible - some of the charge power always gets wasted as heat) means you're pushing 480mAh into the thing, which is 28.8 amps, if you're only taking 60 seconds to do it.
That's a huge current for a little mobile-phone-sized cell; a very good contact between the cell and the charger becomes essential (connect the thing with skinny hookup wire and you'll melt its insulation...), and the charger itself needs to be a pretty hefty item, since 28.8 amps at 3.6 volts is more than a hundred watts, which your average lightweight phone charger is not within a hundred miles of delivering.
If you assume the charging process is only 90% efficient, (for current batteries, it's actually likely to be more like 70%, requiring about 1.4 times the energy in as you'll ever get out; a really fast charge might be more efficient, but I'm just guessing), and you'll need 32 amps for those 60 seconds, with the remaining 192 amp-seconds at a nominal 3.6 volts (something like 691.2 watt-seconds) needing to be dissipated by the cell as heat in those 60 seconds. That's a constant heat load of about eleven and a half watts.
11.5 watts for one minute can be dissipated well enough by a 6.2 by 3.5 by 0.4cm object like this cell (assuming the heat's reasonably evenly distributed over it), but if it's part of an unventilated plastic-cased pack of multiple cells, and especially if the charge efficiency is you can expect that pack to get pretty darn toasty
Of course, Toshiba isn't promising these things in mobile phones next month; they say they'll be electric vehicle products first, and for that application they look quite nifty. The fast charge, in particular, means that regenerative braking (previously mentioned here) could finally start to be good for something in machines that normal humans can buy.
You'd also get around the ludicrous-current problem if you extended the 80% charge time to a luxurious, say, ten minutes; that's more than fast enough for most applications, and would greatly reduce the need for giant silver-plated contact areas and a plugpack the size of a cantaloupe.
Toshiba don't, by the way, mention what the cell's lifespan is; lots and lots of charge cycles could be great for vehicle applications, where a reasonably-price battery pack could just be swapped out at every yearly service, but if the new batteries drop dead in a couple of years (as do some LiI batteries today; four years is the outside limit, and three years is pretty common, as mentioned here) then you're not going to get near a thousand cycles of your phone before you have to throw the battery away anyway.
I have put four hard drives in a standard computer case and power the drives from the PSU (the necessary pins were bridged to turn on the PSU without a motherboard). The IDE to USB converters are routed via a USB hub and into a computer.
It works a treat and new drives can be added as required. Cost = $AU200. USB transfer rates may be a little slower than a network, but for archival storage, who cares?
I'm a professional advertising photographer and now have instant access to all files. Funny how clients never lost transparencies, but lose CD's regularly. Archive retrieval = $AU100 per CD, so my "Clayton's" server paid for itself very quickly!
Yeah, I've thought about that - I was considering just a stack of regular external USB drive boxes, but your version's a bit cheaper and less cable-jungly.
Stringing a bunch of drives together into one shared volume, and elegantly handling expansion and backup, is still a problem, but in the interim hordes of USB drives are not a terrible solution at all, since they allow for-all-intents-and-purposes unlimited storage running from one server. Dozens of drives running from each root port will have serious bandwidth issues even on USB 2.0 (let's not even think about USB 1.1...) but, as you say, for domestic purposes this ain't that big of a deal.
Why do old TVs change channels faster than new ones? My old floor sitting wood veneer Panasonic television changes (cable) stations immediately, while my newer Sony Wega shows a black screen for half a second before the new picture displays. Why? Is there some sort of published metric for this delay? (So I could look for a low number when buying a new TV.)
Failing that, multiple tuners with the current channel, as well as the next and previous (numerically speaking) all set and ready to go, would be neat.
The culprit is digital jiggery-pokery, of two kinds.
First up, fancy TVs these days often have deinterlacers (for "100Hz" display) and various other DSP gear in the signal path. Which is a good thing (especially for people stuck with NTSC), but the processing pipeline has to be fed a stable signal for a moment before it starts emitting anything. Old TVs just take the analogue output of the tuner and blow it straight into the electron guns.
The other kind of jiggery-pokery is the low-component-count integrated circuitry on which all modern TVs, even cheap models, run. I don't know whether it's even possible to buy a new TV with a manual fine-tuning dial any more, or even those nasty little multiple separate thumbwheel things from the intermediate half-digital period; everything's been collapsed down into a few chips, which are (generally) more reliable, as well as a lot cheaper, than a bunch of separate parts.
The down side of the integrated approach is that when you change channels you're no longer switching a circuit from one physically set trim capacitor to another, giving more or less instantaneous retuning. Instead, you're telling oscillators and phase-locked loops to change their settings, and they take a moment to settle down.
Sorry, but I don't think there's any way to tell how long a particular new TV takes to display a new channel, short of trying it out. The time probably varies significantly with signal quality.
There've been multi-tuner TVs for quite a while (Picture In Picture), and multi-tuner digital video recording setups (often just PCs with multiple tuner cards) are becoming more common, too. Of course, it's not like multi-clutch automated-manual gearboxes that select the gear you're likely to want next, knowing that you're unlikely to want to go from 4th to 1st and certain not to want to go from 5th to reverse; a channel-zapper may well skip all over the place. But engaging the next gear, as it were, ought still to be easy enough.
How much current does a Motorola bag phone draw when talking?
That'd be something like a 2950, right?
Motorola have a few current-model (which is to say, introduced some time in the last 10 years...) "bag phones" that all use the same 12 volt 2.3Ah lead-acid battery, and all get the same talk and standby times from it (and all make a brick look pretty slim).
You can work out running current draw roughly by just looking at the run times; assuming you actually get 20 hours of standby or 70 minutes of talk time from a fully charged battery, then dividing the capacity (2.3Ah, 2300mAh) by those times gives you 115mA and about 1970mA current draw figures, respectively.
Mobile phones don't draw power smoothly, though; these are old-style analogue phones and so aren't as clever about it as modern digital models, but you still won't get a constant current draw while running the things.
Any power supply that can deliver, say, three amps at 12V, though, should run one of these phones fine.
I need to know how to make, or if you can explain where I can get, a magnet that is powerful enough to move coins.
It's not going to happen, unless the coins are made of a magnetic metal. Normal coins are made of alloys containing metals like copper, zinc and nickel, and aren't magnetic at all (well, not at any field strength that a sane human would want to be anywhere near, anyway). The mechanisms of various coin-operated devices use this to detect fake coins (washers, for instance); if their magnet catches it, they reject it.
There are, however, some magnetic coins out there.
(Since I put this up, it's been pointed out to me that all recently minted Canadian coins are magnetic!)
You can also buy rigged magnetic coins specifically made for use in magic tricks.
If you've got magnetic coins (or, of course, just some washers or something) then you'll easily be able to move them about on a table with a concealable neodymium-iron-boron ("rare earth") magnet (of which you can find lots on eBay). Conjurors were doing tricks like this before NIB magnets even existed, and the stronger field of NIB magnets has made things easier; many magicians refer to NIB magnets as "PK magnets", because they're useful when you want to fake "psychokinesis".
Can you look up the cost of running a big oil-filled column heater, please? Buying wood is ridiculously expensive. I know column heater isn't the most economical thing, but it is efficient and once the initial warming up is done, it can be turned down and costs less.
It's hard to say.
Big column heaters generally have a rating of around 2.5 kilowatts, so they'll consume 2.5 kilowatt-hours of power per hour, if they're on constantly.
All electric heaters cost the same to run per unit of heat produced, because all electric heaters have the same functionally 100% efficiency. Since all they do is put electricity through an element of some kind, they all work about as well as each other - things like fans and glowing elements use a trivial amount of extra power compared with the thousands of watts of plain heat output.
(Reverse cycle air conditioners are a special case - as heat pumps, they output more heat than the amount of electricity that goes into them, so they're more efficient than regular electric heaters at warming your house.)
Electric heaters with a rating of X watts and a thermostat will cost less to run than heaters with the same rating and no thermostat, provided the thermostat isn't just wound all the way up, but the magnitude of the difference depends on how often they click on, which is of course a function of how cold it is outside, how big the place they're heating is, and how well insulated it is.
Electric is, of course, more convenient and can be delivered in smaller doses; you can't warm up your bathroom for 10 minutes with a wood fire.
Escape velocity from the surface of the Earth is, like, 11 kilometers per second, more than 36000 feet per second, while rifle muzzle velocities are more like 3000fps, so my bullet isn't going to shoot off into space; it's going to come back down somewhere.
Is it going to come down as fast as it went up? Will it still be spinning? Will it come down as fast as a skydiver? Faster? Slower?
Gun-happy minds want to know!
Human terminal velocity depends on the pose you adopt - a skydiver in the spread-out "star jump" pose falls much slower than one in the head-down "lawn dart" one. But bullets that don't hit something and are fired from a rifled weapon don't tumble (they don't have time to shed their spin - air resistance has much less effect on the spin of the bullet than on its velocity) and so, even falling blunt end first, they're capable of something like 100 metres per second terminal velocity, depending of course on the type of bullet (some sources say as much as 700 feet per second, 213 metres per second).
There've apparently been some campaigns in the USA to stop enthusiastic but irresponsible supporters of the Second Amendment from blazing away at the sky, and some of the campaigners have said explicitly that bullets come down as fast as they go up. This isn't true at all, unless you live on an airless planet; it's more like a tenth of the speed, for rifle bullets in Earth's atmosphere. But explaining the details of ballistic physics to the kind of people who like to shoot into the air probably isn't going to get the message across; hence, the comes-down-just-as-fast line.
If you're wondering whether the spin of a bullet's really fast enough that the thing's still strongly gyroscopic by the time it lands the better part of a minute (!) after being shot, by the way, bear in mind that a mere 3000 foot-per-second bullet (not an amazing speed for a rifle bullet) fired from a 12" twist rifle (one full rifling revolution per foot of barrel) is spinning at 180,000 revolutions per minute when it leaves the barrel. A .38 handgun with a 1:18 twist and a mere 900fps muzzle velocity still gets 36,000 RPM; that's considerably faster than any pull-cord toy gyroscope is likely to manage, and those things spin for ages.
Plenty of the initial spin is, therefore, still left when the thing lands, and so just about any rifled weapon projectile will quite accurately hold its initial orientation through its whole trajectory. If the gun was aimed upwards at a 45 degree angle, the bullet will fall back to earth at that same angle, not nose-down or nose-straight-up.
An unremarkable .303 rifle will, with standard modern ammunition, fire an 11.7 gram bullet at something like 755 metres per second (reference), for muzzle energy in excess of 3300 joules. Assuming the bullet comes back down at a lazy 100 metres per second, it'll be carrying only about 60 joules of energy, and presenting its somewhat less dangerous blunt end (or its side, for angled shots) to any unfortunate soul underneath. It's much better to be in the way of 60 joules than it is to be in the way of 3300 (a plastic hard hat will protect you from any falling rifle projectile, but achieve basically nothing against a direct shot), but little .22-calibre pistol cartridges have this kind of muzzle energy, and plenty of people have been killed by those. A 60-joule projectile feels like a pretty solid whack with a hammer, and it can definitely kill you.