Dan's Data letters #177Publication date: 2 November 2006.
Last modified 03-Dec-2011.
The console wars cometh, and I have a simple question to ask you.
Or are we seeing some rather fanciful marketing?
If the "Sixaxis" controller starts out with much more battery life than most people need then it's quite possible that it'll still be OK after four years. Maybe some significant number of controllers will still be OK after six. LiI technology is slowly improving, and if the charger's gentle, that'll help too.
But I really wouldn't bet on it.
If Sony are, as has been reported, going to do no-questions-asked replacements for controllers with dead batteries, this could end up costing them quite a bit of money. Unless, of course, they just insist that any controller that's less than six years old must have been killed in some other way, and is thus ineligible for replacement.
(The fact that the Sony spokesdroid made reference to "memory effect", though, doesn't bode well. Rechargeable lithium batteries don't have any "memory effect" at all; they just wear out, and I'd be quite surprised if any lithium-anything batteries being manufactured today still held much of a charge after six years.)
Do you know where I can order a battery for my Combat DigiQ? My tank won't charge and "- -" and "CH" flash on the LED display (instead of the constant "CH" display).
I don't know who sells them, but they're not totally obscure. The pack's two "1/4 AAA" size cells, like these ones. Call around your local electronics stores and ask if they know of such a thing.
If you can cram some similarly specified two cell NiMH (or even lower-capacity NiCd) pack in there, it ought to work fine. "Mini RC" gadgets that charge from their transmitter don't put anything like a full charge into the pack, so capacity isn't that important, but they charge their tiny batteries very hard for the brief period they're connected to the charger. So it's possible that NiCd cells with half the capacity will actually work better, since the NiCd chemistry is less bothered by rough charging, all other things being equal.
Just finished reading "Upgrade your UPS" where you used two 12V batteries, etc. I'd like to pick your brain, so to speak.
I'd like to replace a 110VAC to 24VDC adapter with a 24VDC source/charger. The unit powers a modem and draws about 1.5 amps. Not sure if I should find a 24V battery and charger, two 12V batteries and a 24V charger, or two 12V batteries and two 12V chargers. Spoke with several technicians and became confused in the process. If I use a standard 600 watt 110VAC UPS, the modem might get juice for a few hours. I'm looking to maintain power for several days. What might you recommend and why?
Two 12V batteries in series and a 24V charger with enough output that it can run the modem without freaking out should work nicely.
If you want seven days of non-stop 1.5 amp operation, though, you're going to need 252 amp-hours. You'll definitely be using lead acid for that kind of capacity at a vaguely reasonable price, and even deep cycle lead acid batteries don't like being run flat, so let's add some headroom and say 300Ah.
300Ah 12V batteries cost at least a few hundred US dollars each and are rather heavy, but a couple of people can move them around. Around this capacity people start thinking seriously about using more batteries with fewer cells, like four six-volters instead of two 12s, or even single cells (2V each), mainly so that you can swap out whatever part of the battery has a problem, but also because you don't have to shift so much weight at one time. For real "industrial" durability, there are also nickel-iron batteries, which have even worse energy per kilogram than lead acid and are more expensive, but are tough as old boots.
The cheapskate option is to build your giant battery with a series-parallel rig of however many cheap(er) automotive or marine batteries you need to get the capacity you want. Put two 150Ah 12V batteries in parallel and they're a 300Ah 12V battery; put two of those pairs in series and you get 300Ah at 24V. The more cells you end up with in your battery bank, though, the more of a hassle it'll be to make sure all of them are healthy.
The gripping hand here is whether you actually need to deliver that much power, for that long. Are you sure the modem really draws 1.5A, or is that just what the power supply can deliver? Even if "24V 1.5A" is printed on the sticker on the bottom of the modem, that could be a start-up surge current or something, not the amount the modem draws most of the time.
A normal phone line modem can be expected to draw little more than 10 watts. 1.5 amps at 24 volts is of course 36 watts. I'm sure there are plenty of lightweight computer modems these days that draw only a few watts. I'm also sure there are plenty of big industrial special purpose modems that draw a lot more, though.
If your device actually draws a mere 15 watts, and you decide that three days of run time is acceptable, then you'll be able to get away with something like a pair of 50Ah 12V batteries, which is a much less alarming prospect. You can get man-portable deep-cycle and "RV" batteries with better ratings than that from umpteen places.
Why, your whole battery bank is now likely to weigh less than 50 kilos!
Actually, a couple of big automotive starter batteries could cut it if all you need is around 50Ah. Their 25-amp-load "Reserve Capacity" rating will probably be only be 30 to 35 amp-hours, but they may manage 45 amp-hours into the far lower load of a single modem.
For the charger, get something like this (that's an Australian-dollar price). I haven't tried it, but I bet that charger won't mind at all if something's draining an amp or so while it charges. Dumb chargers certainly don't mind; smart chargers usually judge battery capacity based on terminal voltage, and don't worry about the amount of time the charge cycle takes.
Also, importantly, make sure you include a fuse in one of the supply wires. Or a circuit breaker, which can be reset very easily and won't cost much.
The fuse/breaker is essential because your battery bank, even if it's just a couple of car batteries, will be capable of making a five pound pipe wrench glow a cheerful orange. Shorted output wires with no overcurrent protection aren't so much a fire risk as a fire guarantee.
Of course, you could also use a big battery bank to power a standard UPS, like I did only more so. This is likely to be a neater solution, with no concerns like the possibility that a charger will allow its batteries to discharge through it if it loses power (that's a thing you'll have to check, though I think pretty much all smart lead acid chargers these days don't have that problem). If you don't have anything else you'd like to plug into the UPS, though, it's not as attractive as it would be in most applications.
[Hank got back to me. His modem did indeed turn to draw less than he thought - only 13.2 watts, according to the sticker on the bottom.]
Magnesium-powder flash preferred
Time stopped immediately after World War II for my Dad. Sort of a "Slaughterhouse Five" situation I guess. He's still using a fixed-focus Argus model-A 35mm Leica copy-cat - with a retro Bakelite and chrome body. Likewise, his favorite radio is AM-only with two knobs: On/Off-Volume and tuning.
Now he wants a digital camera. I am searching for a simple yum cha <2MP camera with fixed internal storage, no picture card needed.
He wants an optical rangefinder, no color LCD screen - and here's a new paradox - because the image disconcertingly moves in the opposite direction of the camera pan or tilt.
And here's the rub: It must be a driverless USB mass storage device, so he can plug it into the local FujiFilm kiosk directly.
The VuPoint DC-ST100G-VP $US40 1.3MP 16MB is the only model I can find.
Are there any other AM vacuum tube digital cameras out there for my old soldier?
I shall presume you have already waved your arms at him about the fact that the view through the viewfinder moves in the same "disconcerting" way as the view on a screen.
Sorry, but I don't really know if anything's turned up in the toy-cam market segment that takes particularly good pictures. The world now teems with suspiciously cheap "SIX MEGAPIXEL!!!!1!!" integrated-storage cameras, some of which are screenless and all of which of course interpolate up some lower resolution and have a lens that can't even use as many pixels as the camera really has. But I don't know whether there's one of them that's actually worth using. They all, generally, take really disappointing pictures, though as I've said in my old reviews of toy-cams, you can get pretty decent results out of some of them (generally not the ones that're pretending to be a proper camera) if your expectations are low enough. But they're all likely to compare quite badly with the old film camera he's already got.
Honestly, I'm thinking of suggesting a camera with a screen, over which he can put some tape if it bothers him too much.
If you were made of money, I'd suggest you get him an Epson RD-1, that weird Leica-lens rangefinder camera which debuted a couple of years ago for $US3000 and has not gotten any cheaper since (!!). It's got a screen, though, as has the recent, considerably more expensive Leica M8. Darn.
Or you could stick with the Leica idea, and pay a bit less...
Besides its jokey styling, the little Leica-shaped Aiptek is actually not a bad toy-cam at all.
Some camera that has a screen but doesn't have on-screen preview could perhaps work - all DSLRs are like that, but they of course fall down on the other requirements.
Your caselight piece has given me some major pointers on what to do and not do, like don't go storming in and buy the stuff.
I have one simple query. I'm planning on putting my name on my case using LEDs poked through the mesh in the side panel, and was looking into serial/chain method. There are approx 60 LEDs - what would your recommendation be? One chain or two chains? Maybe parallel with a small amount of wire to the LEDs?
And most importantly, as I know next to nothing on electricals, what resistor and power setup would you recommend? I've got approx 56W spare in my 460W PSU.
I'd rather not use an LED board, as the system greatly relies on the air coming through that mesh. Any suggestion on holder type? The holes are 6mm.
Provided your name isn't Spencer Q. Steinhausen-Pumpernickel, this should be easy enough to do.
(It's also easy enough to create a hideous wiring mess, of course. Board-less "rat's nest" wiring was the norm before the invention of the printed circuit board, and even professionally laid out version of it tend to be fairly chaotic, as you can see if you have a look inside an old radio.)
Even if your LEDs only need two volts, 60 LEDs in series would need 120V. Break them into two chains and you still need 60V, which your PSU can't deliver. You need to make the chains short enough that they can run from 12V, which is the highest you can get from a computer PSU.
This neat-o calculator lets you figure this stuff out fast.
If you elect, for instance, to use yellow LEDs (which'll want about 2.2V) and run them from 12V at a nice safe 10mA (for long LED life and low heat), you'll need 12 strings of five LEDs each, each with a 100 ohm series resistor, which'll provide very good regulation (a string of LEDs with little or no series resistance is likely to be prone to thermal runaway).
You'll only need a watt or three of power for the whole 60-LED array no matter what 5mm (or 3mm) LEDs you use, so it'll run fine from the 12V rail. The PSU won't even notice it. If a 5mm LED is drawing more than a tenth of a watt (100 milliwatts, mW), it's probably running too hot; 3.6V white or blue LEDs running at their rated 20mA draw only 72mW each, and 2.2V LEDs at 10mA are only 22mW each.
Regarding holders - if you're poking LEDs through existing mesh then just clipping out a bit of mesh and using dabs of hot-melt glue (or silicone, or solventorific "shoe goo", or whatever) to retain the LEDs should work quite nicely. You can also get cheap two-part clip-together plastic LED bezels which are easy to use and can be had in versions that fit all standard 5mm and 3mm LEDs, but they're only useful if you're mounting LEDs in a pretty solid panel - they probably wouldn't help significantly if you're mounting to mesh, and could be unsightly.
Rods to the hogshead
There is a question which I have been wondering about for a while, and thought that perhaps you might answer it.
An electric car that runs on batteries charged from grid power has a very efficient electric motor. But how efficient is the power plant, say a coal or gas or nuclear plant, when compared to a standard petrol engine? If a petrol engine is about as efficient as a power plant, I se no reason to use an electric car.
Or are there other things which make it worthwhile?
The thermal efficiency of a petrol or diesel engine (diesel's better than petrol, but not a whole lot better) is frequently quoted as being in the thirty-something-per-cent range. That, however, is the ceiling efficiency, when running at optimum RPM. A lot of the time, the efficiency of a normal car engine is a great deal worse - when accelerating up a hill, for instance. When you're sitting at the traffic lights with your engine running, your efficiency is zero. When an electric car (or a hybrid) isn't moving, virtually no power is being used. Hybrid cars therefore have a big fuel economy advantage over ordinary small economical cars if, and only if, they both spend a lot of time in stop-start city traffic.
Now, fuel-burning power plants are quite horrifyingly inefficient too. Even a really modern coal-fired plant still manages less than 40% thermal efficiency. But it manages that efficiency pretty much all the time, no matter what.
(A correspondent's pointed out to me that combined cycle plants can do a great deal better, which can make them cost-competitive despite the fact that they run on more expensive gas or fuel oil. Australia has a few combined cycle plants in the works, at least, but running combined cycle on coal is less than trivial.)
But then again, the reason why major power plants run all the time is that they're generally used to supply "base load" power - the amount of power that always has to be delivered, rain or shine, day or night. Most big plants are somewhat adjustable in their output, but not adjustable by a large enough amount, or quickly enough, to deal with peak loads that happen when, for instance, everyone comes home from work in the summer and turns on their home air conditioners. Peak load plants are usually quite a lot less efficient than base load plants; most of them are monster gas turbines that can be turned up and down very easily, but pay for it with a horrible amount of energy wasted as heat.
And then there's yet another caveat - that waste heat can be re-used if the peak plant's somewhere near some humans who need heat for some purpose (keeping houses warm, industrial processes), whereupon you can pipe the heat to them, and greatly improve the aggregate efficiency of the plant. This is called "cogeneration", and it fuzzes up the calculations further.
Getting back to the electric cars - charging batteries is not anything like 100% efficient, and you lose a few more percentage points in the motor, and regenerative braking doesn't actually recover anything like as much energy as you might hope. The whole coal-to-driving efficiency chain for electric cars powered by modern but non-cogenerating coal-fired plants, though, still works out somewhat more thermally efficient than it does for petrol-powered vehicles, and it also works out considerably cheaper. Coal remains plentiful, oil does not.
The location of the pollution matters, too. Power plants are generally in relatively unpopulated areas and have tall smokestacks, so their output is well mixed with air before many people breathe it. Modern coal plants also have fly ash collectors that allow the nastiest of their pollution to be collected and buried.
Cars, in contrast, pollute at ground level everywhere they go.
Nuclear plants are a thorny issue. Their thermal efficiency is lousy, too - all that steam coming out of the cooling towers represents wasted energy. But their pollution output is, arguably, close to zero per kilowatt-hour. Nuclear waste management seems to me to be a very simple issue that's only made complex by politics. But that still means it's complex, so you can't accurately estimate how much of a problem it actually is.
In a world run by rational people, you'd just leave the waste from 40 years of operation of a large nuke plant to cool inside the decommissioned plant for another 40 years - or longer, if you like - then finally truck it away in, oh, maybe ten standard shipping containers (each of which has had its volume halved by armour - that's really all the space it would take), drop it in a big geologically stable hole somewhere, and post ten guards to stop nuts from trying to steal it.
The waste from 40 years of equivalent coal plant operation would fill a fair-sized harbour. And it'd be radioactive, too - not radioactive enough that you should worry about it if you're not breathing it in (that's why fly ash collectors are such a big deal), but certainly radioactive enough that it'd be categorised as low-level waste and treated like Instant Death if it had come from a nuclear plant instead.
There's that politics thing, again.
OK, cameras have too many megapixels, but... how many are good?
I thought 300dpi is a good resolution, so for a 14 x 17 inch blowup I need 14 x 17 x 300 x 300 = 21MP. Of course, for 8 x 10 I only need 7.2MP. These numbers seem higher than what you are implying.
Am I way off thinking 300dpi is necessary for good images?
For close inspection, yes, 300dpi is the benchmark for photo prints. It's a general rule of thumb that you need 300dpi for a print that's going to be inspected from 30cm by someone with 20/20 vision.
Bigger prints, of course, are often viewed from further away. Look at a poster close up and you'll see that it's pretty fuzzy. I wrote more about this in this column.
If you want a big print with perfect photo detail then yes, you need the kind of sensor resolution that still costs outrageous amounts of money. But you just can't take a picture with that many pixels in it with the vast majority of lenses, no matter how much resolution your sensor or film has.
For ordinary poster prints, 100dpi is generally adequate. To print a full A3 sheet at 100dpi, you only need 1169 by 1654 pixels. A full A1-sized poster still only needs 2339 by 3311 pixels.
Lately, I've been hearing about the wonders of filling my car's tires with pure nitrogen. Namely, in one of Jay Leno's Popular Mechanics articles, and what seems to be every auto shop.
Air is already ~80% nitrogen, can you use your hullabaloo debunking skills to tell whether that extra 20% really matters?
No, it doesn't. Jay is what we professionals refer to as "wrong".
For normal driving purposes - including heavily-driven vehicles like taxis and big rigs - you will never notice any difference at all from nitrogen-filled tyres, except of course what it costs to have them filled that way, and whatever delusions you manage to develop, on account of how you're a human being.
If very high run-to-run consistency is important to you then nitrogen is the way to go, which is why it's used in tyres for various exotic applications, including racing cars. But tyre pressure will still change drastically as the tyre temperature changes; Leno's claim that a 32psi cold tyre will stay at 32psi after a burnout as long as it's filled with nitrogen is refuted by high school physics.
Leno also says a nitrogen fill reduces corrosion on the insides (I presume, since that's the portion covered by the tyre) of aluminium wheels, which is not the kind of thing I get too upset about, but, y'know, whatever.
All a dry nitrogen fill actually, demonstrably gives you is a little less change in pressure with tyre temperature, mainly because there's no water in there. A dry air fill would work almost as well for that, but air from a normal compressor isn't dry, as anybody who's absent-mindedly used a compressor to blow dust off something and managed instead to blow the rusty water from the bottom of the tank all over the thing they're trying to clean can tell you.
(There are various systems for draining water buildup from the bottom of an air-compressor tank, and for trapping it before it gets to the end of the hose. A bit of water in the air supply doesn't matter to most pneumatic tools, but for certain applications - like spray-painting, for instance - it's very undesirable.)
Nitrogen may also help prevent very hot tyres from combusting, in extremely-high-performance applications like dragsters and aircraft landing gear.
The idea that nitrogen protects tyres from slowly rotting away because of the oxygen inside (presented in another Pop Mech piece) is also defective, though. Tyres have plenty of oxygen outside them to do that, but all that actually happens is that they wear out and/or lose enough elasticity for non-internal-oxidation-related reasons that you have to replace them anyway, long before they actually rot.
Perhaps, maybe, barely conceivably, a nitrogen fill might make some primitive and poorly-vulcanised tyre rot slower, but I doubt it.
For the usually quoted purposes - better fuel economy, slower leakage - nitrogen is useless.
Jay Leno does seem to have considerably more clue than the average Rich Dude With 80 Cars. I can't help but think that the ten thousand dollars he spent on his nitrogen generator could perhaps have been better used, though.