How To Spot A Psychopath

I started writing a whole big thing about a Flexiglow “Series Connect” power supply, but there’s not a lot of point to that since I don’t think it’s possible to buy one new any more.

The 500W Series Connect had been sitting on my to-review pile since late 2005. The nice people at Anyware who sent the PSU to me might have been annoyed about that. But it’s now clear to me that they should instead count their blessings that I didn’t get around to looking at it until now.

This power supply turns out to have armour on all of its cables that’s so thick that the main motherboard power lead feels like a garden hose under full mains water pressure.

If your computer layout happens to match where this PSU’s ludicrous cables want to go, it’ll work - though you may find it impossible to put the side back on the computer case.

For almost any other computer, it’s likely to be physically impossible to plug this PSU in, even if you only need a few of its leads.

I managed to get the main motherboard connector to plug in as long as the PSU itself was six inches in position and ninety degrees in orientation from where it was meant to be. Any attempt to move the PSU closer to its proper mounting location threatened to wrench the motherboard socket right off the board.

I then tried just cutting the useless armour off the leads. I’ve got a pipe cutter that made short work of the outer rubber layer. Under that, though, there’s braided shielding, which of course frays all over the place and stabs your fingers and is difficult to cut without cutting the conductors under it and it’s all a horrible schemozzle.

Do PC power leads need braided shielding? Of course they don’t. PC components expect to get a bit of RF noise on their DC input. It’s possible that some marginal (or heavily overclocked) components will work slightly more reliably with slightly less noisy input, or that some cruddy sound card will be a little less noisy that way, but there’s a reason why the ATX12V PSU standard does not require shielding for DC wires.

The standard does, however, prohibit PSUs from sending more than a certain amount of noise down their DC wires, because that noise can easily out-shout - by orders of magnitude - the amount of noise the wires can possibly pick up from the air.

Shielding the wires, in that case, simply ensures that the PSUs own noise remains uncorrupted by noise from elsewhere.

I still needed a PSU to replace a dead one in a home-server box, though, so I made a shortlist of power supplies with enough plugs to support the forest of drives inside the server, then stuck a pin into the list and ordered a Corsair TX750W.

Apparently this PSU actually can deliver 750 watts of power, which is (a) way more than this server will ever need, and (b) quite unusual in the consumer PSU market. “Generic” PSUs usually underperform their stated capacity by a truly shameful margin, and you shouldn’t expect even a brand-name “750W” PSU to be able to deliver more than a constant 600W or so. Some do, but many don’t.

(The TX PSUs are made for Corsair by Channel Well Technology, who make similarly high-spec PSUs for other companies, like Thermaltake.)

This PSU also has far more connectors than the server will need - but it’s got enough drive connectors, which is all I really care about. And it wasn’t much more expensive than a much less capable PSU. And under-loaded PSUs generally live for a very long time, and are likely to be more efficient. So what the heck.

I’m already glad I bought the Corsair, because it gave me such a laugh when I opened the box.

Within the box, and within the foam anti-shock packaging, but outside the final clear-plastic-bag level of packaging, this PSU comes in a fuzzy drawstring bag.

It’s a very cheap fuzzy drawstring bag; thin, with fuzzy pseudo-suede on the outside only, and redolent of the various outgassings of the fresh electronic components that’ve lived within it since the PSU was bagged up at the factory. It’s not nearly in the same class as your traditional Crown Royal dice bag.

But it is, nonetheless, within the definition of the term, a fuzzy drawstring bag.

For a computer power supply.

So, like, if you feel the need to unscrew just the PSU from your computer and carry it around with you, you won’t have to tuck it uncomfortably under your arm or carry it by the ATX cable like a big square dead rat or something.

No, man. Not you. Not the Corsair TX owner.

You can bag that sucker up, man!

And then, if any punk on the street should allege that your rig be insufficiently pimped, you can say to him “Yo, I gots my P-to-the-S-U right here in a bag, bitch! What you got? Well? You got a motherboard hangin’ round yo’ neck that I ain’t noticed? Huh?”

(And yes, I’m pretty sure there’s a factory worker out there who can’t believe he’s making these things. Previously.)

Here in Australia, you can buy your own TX750W from m’verygoodfriends at Aus PC Market for a mere $AU211.20 including delivery anywhere in the country. And you might as well; it looks to me, and to people who bothered to actually test it, as if it’d be a perfectly good piece of hardware for the money, even if you didn’t get a fuzzy drawstring bag into the bargain.

Australian shoppers can click here to order one.

These are my Not Very Useful Dice.

The “crooked in every sense” red six-siders are oddly satisfying objects. They’re classic, if rather large, sharp-edged casino dice, except for the obvious.

I haven’t thrown them enough times to see what kind of result distribution the crooked d6s give. In the aggregate they’re probably actually quite fair, since they’re all somewhat close to cubic and they have the proper numbering scheme, with opposite sides adding to seven.

(I think they’re actually likely to throw a bit low, since the smaller sides on four of them are all ones, plus one two and one six. Frankly, I just want to try sneaking them onto a craps table some day. If you want some of your own, try searching for “crooked dice“; a set of six shouldn’t set you back more than $US15 delivered.)

The other three dice are perfectly fair. Just… not very useful.

The blue one’s a d24, a tetrakis hexahedron (one of two possible shapes for a d24 - the other is, of course, the deltoidal icositetrahedron). In gaming, you actually can use a d24 to quickly determine what hour of the day on which some random event takes place. But you can also do that in various other ways on the rare occasions when you have to - like, for instance, a d4 to determine the quarter-day and a d6 to pick the hour of that quarter.

So the d24’s appeal remains… specialised. Dungeons and Dragons used to use d24s for a few things, but it doesn’t any more. (D12s seem to have been similarly deprecated.)

The larger polyhedron is a rhombic triacontahedron, a d30. It’s the big brother of the surprisingly antiquitous, famously malicious, icosahedral d20 that’s become the very symbol of gaming nerdery.

I think the d30 has a certain… machismo.

“Oh, you roll twenties, do you? Well, I beat that a third of the time.”

It’s hard to top that, if you don’t have big brass ones.

The d30 can also be substituted for by other dice, though I don’t think there’s any terribly elegant way to do it - perhaps a rolling-pin d3 (itself substitutable by a halved d6) for tens, plus a d10 for units. This isn’t something you’re likely to need to do very often, though, since d30s are almost as unpopular as d24s. People use them now and then to represent some sort of boost (lucky artifact, you’re the son of a god, you bought the DM a pizza) for what would normally be a d20 roll. That’s about it.

The red sphere is a more commonly seen item; it is, of course, Lou Zocchi’s hundred-sided “Zocchihedron“.

Lou is probably royally sick of the sight of his d100, since he spent ages trying to make the darn thing work right, and it still doesn’t, really.

The main problem with a 100-sider is that it’s basically a golf ball, and so any sort of fair roll will take ludicrously long to settle compared with the normal “d100″, which is just a pair of d10s, one for tens and one for units.

To address the rolling-across-the-room problem, Lou made his d100 hollow and partially filled it with teardrop-shaped metal weights, which slow its roll considerably, and also make it usable as a very small maraca. The d100 is still really only a curiosity, though, and may or may not be biased in favour of the more-widely-spaced numbers nearer its equator.

Companies like Chessex, Koplow Games and Lou Zocchi’s Gamescience make a number of other impractical novelty dice. The d5, d7, d14 and d16, for instance, and even the majestic d34. Unfortunately, though, most of the weird-numbered dice that I don’t already own are of the pyramids-stuck-together trapezohedron type, which as the side-count rises makes them look more and more like a spinning top rather than a die. The d34 has a particularly severe case of this disease.

I’m still tempted to acquire them, though, so I can have a whole Crown Royal bag full of dice that nobody can use.

If you’re at all interested in the aesthetic appeal of dice, by the way, allow me to highly recommend slight-of-hand wizard Ricky Jay’s book “Dice: Deception, Fate, and Rotten Luck“, a slim volume which alternates gambling - and cheating - history with a lot of gorgeous pictures of decaying six-siders.

Phoenix is the winner of the Trossen Robotics TRC Project Contest…

…and deservedly so.

She’s a little small to really conquer the indoor environment, but the design is very scalable; I think a double-size Phoenix made with super-torque standard-sized servos or the big quarter-scale ones could scuttle up ordinary stairs quite easily.

(One could, in fact, be climbing your stairs right now. What was that noise?)

More at the forum thread.

(Via.)

More videos that you’ve probably already seen, but which are new to clueless me (via):

The punch was what really sold it for me.

Mr Flare also had a large role in Babylon 5.

An excellent guide to the practical skeptical outlook.

Including something Amazing in the sky.

A great summing-up of this recent story, albeit with some disturbing attention paid to YouTube comments.

He really needs to stop reading those comments. Set the comment threshold to “excellent (+10 or better)” and all of that troublesome text will just… go away.

More at Captain Disillusion’s YouTube channel.

The sadly neglected “Things I Won’t Work With” category of Derek Lowe’s organic chemistry blog (previously) now has another entry, as a result of an innocent inquiry regarding what chemicals will, if you dump sand on them to try to stop them burning, start cheerfully burning the sand.

It turns out that chlorine trifluoride (merely discovering that one Cl and three Fs can in fact be squished together should send shivers up the spine of anyone who was paying any attention at all in high school chemistry) is a party looking for a place to happen.

Yes, this MSI motherboard northbridge fan powered by a teeny little Stirling engine is very neat.

I hope it makes it into production, and I also hope it’s well enough made that it’ll last at least as long as the crappy electric fans you usually get on a northbridge heat sink.

Small Stirling engines like this have very little power, and they need to be manufactured to very fine tolerances if you want them to run on a small-ish temperature differential - like, on top of a CRT or even LCD monitor, or on the heat of your hand for fancier models.

I’ve got one that runs fine on a cup of tea, but it isn’t smooth enough for anything better. You need something like the above engine, with glass cylinders and graphite pistons, to get really low-temperature-differential operation.

A modern motherboard main-chip, though, will easily give enough heat to run a small Stirling fan, and it shouldn’t need much wind over a good-sized heat sink like this to keep it at an acceptable temperature (actually, the Stirling fan may pretty much be just tinsel - the normal air flow through the case may be plenty to keep the northbridge cool, with a heat sink that big).

So the goofy MSI rig actually ought to work quite well even with a relatively cheaply-made Stirling engine. And if the engine craps out after a few months, you can always bodge a normal fan in there to replace it.

(Or do so immediately, so you can take the engine out and display it on top of your coffee cup instead.)

The product this little fan most reminds me of is the Heat Wave wood stove fan, which takes advantage of the large temperature differential between the top of a combustion stove and the ambient air to run a robust, long-lived Stirling engine with enough power to circulate air quite effectively, which can considerably improve the room-inhabitant-heating effectiveness of the stove. There are similar, cheaper products based on Peltier elements and boring electric motors, but c’mon, stump up the extra for the piston motor. You know you want to.

Dammit. Now I want a glass-and-graphite low-temperature Stirling engine.

Or maybe one of the Böhm kits.

Or the Gakken version, to add to my collection.

Or their steam car (note, regrettably, that neither this kit nor the Vacuum Engine car actually come with a Wondermark-ish top-hatted figurine to ride them).

(Note that all this does not mean that PC-powered steam engines are just around the corner.)

This is just one of numerous Giger/Matrix/Lovecraft things from Lego’s upcoming series, tentatively titled “Mummy, why is sleep now all teeth and bones?”

(Actually the Black Fantasy Contest on the Classic-Space forums.)

The Daily WTF just ran a story about a company that sells listings in one of those highly questionable “business directories”, and has a Web site with hilariously poor security.

(Hint: If you want to “password protect” a Web page, don’t put the hard-coded username and password in the source code of the login page. Oh, and don’t put the URL of the otherwise wide-open “protected” page in plaintext in the source either, or people - by which I mean, “bright eight-year-olds, or unusually well-trained monkeys” - may just copy and paste that URL to their browser’s address bar.)

So far, so unremarkable.

But you need go no further than the Featured Comments below the story to find the start of a good old-fashioned Internet flameout by the owners of the site.

How dare you “hack” our site, this directory is our livelihood and we forbid you to say you think it’s overpriced, we know where you live you druggie bitches, et cetera.

If you like this sort of thing, then this is the sort of thing you will like.

LED lamps for standard low-power automotive sockets - things like interior lights, number-plate illuminators and brake lights - are now widely available and dirt cheap.

So I bought one, to see if it works any better than the standard interior light in my car.

There was nothing wrong with the standard interior light, but like a lot of low-power automotive bulbs, it’s offensively inefficient.

The bulbs used in cars for things like interior lighting and instrument panel illumination have as their two chief design goals cheapness and durability. Both of these goals push manufacturers towards very low-efficiency devices. And the standard “dome” light in the middle of the ceiling of most cars generally doesn’t even have much of a reflector behind its bulb, so something approaching half of the light just goes into warming up the light fitting.

So the dome light in my car looked like a fine candidate for LED improvement to me. Particularly now that one lamp will only cost you $AU8.98 delivered from Hong Kong.

(I got mine from this eBay seller.)

My car’s interior light uses the small 31mm size of “festoon” bulb, the kind that look like a glass fuse but with points on the metal caps on each end.

The 31mm form factor doesn’t give a lot of room for modern super-LEDs. You can now get 31mm lamps with a single allegedly-one-watt white LED in them…

…or you can go for the type I got, with no fewer than six surface-mount sub-1-watt super-LEDs.

There are also replacement bulbs that use a cluster of standard 5mm LEDs. They may be OK for things like instrument panel lighting, but you shouldn’t expect as much light as you’ll get from a single 1W LED unless there are at least a dozen 5mm LEDs in there. Even then, it’s doubtful.

If you don’t see a lot of yellow phosphor looking back at you, you’re probably not looking at a very bright lamp.

I gave the new bulb a whirl on my bench power supply to see how much power it consumed. Then I tried the same thing with the (rather old) stock bulb.

The LED lamp drew only about 55 milliamps (mA) at twelve volts, for a power rating of only about 0.66 watts. Raising the supply voltage to 13.8V - which is what you’ll get when the car’s running and the alternator’s turning - raised the current draw to about 105mA, for 1.45 watts.

The stock bulb has a nominal ten watt rating. From 12V it drew around 0.725A - that’s 8.7W. From 13.8V it drew only a little more, about 0.785A (this is because the resistance of light bulb filaments rises with their temperature), giving 10.83 watts.

I expected the LED lamp to deliver much more light per watt than the incandescent bulb, and it also gets a big effectiveness boost from only throwing light out one side, wasting none of it by shooting it uselessly into the dome light fitting. But this was still a pretty huge power difference. At 13.8V, the old bulb draws 7.5 times as much power as the LED lamp; at 12V it draws more than thirteen times as much.

It was pretty easy to install the new lamp, although it did turn out to be a bit longer than it was supposed to be, making it a bit of a tight fit and also making it impossible to install it perfectly level. It ended up tilted a bit toward the left seat, though not enough to make a huge difference to the illumination on the two sides.

To cancel out any side bias, I tested the brightness of the two lamps with my somewhat accurate light meter sitting at the base of the gearshift (and with the standard plastic diffuser in place, too).

The light meter is calibrated in lux, a unit that’s weighted to match human brightness perception. This gives the LED lamp another advantage, because the long-life low-temperature incandescent bulb gives very yellow light, while the LED lamp gives the characteristic blue-white of “white” LEDs. The blue-white has more energy around the green frequencies where human vision works best, so a given raw energy level of yellow-white light will appear dimmer, and read lower on a luxmeter, than the same energy level of blue-white.

Anyway, the stock bulb gave a reading of about six lux with the engine off (13V), and about nine lux with the engine running. Not a bad illumination level, given that it was being measured quite a bit lower than the place where you’d typically be, say, holding a map you were trying to read.

Swapping in the LED lamp gave… exactly the same readings!

My light meter isn’t terribly accurate down in the single-digit lux, so I won’t swear to you that there wasn’t actually a bit of a difference one way or the other. But there clearly isn’t a huge difference. And the new lamp, subjectively, lit up the cabin of the car just fine. Despite drawing around a tenth as much power.

This sort of thing can make a big difference in certain circumstances. If, for instance, you have a typical small car battery with about 25 amp-hour capacity before it starts getting very unhappy, a ten-watt interior light will drain it in thirty hours. Swap to a one-watt LED lamp and you’ll probably still be able to start the car even if you leave the light on for ten days.

This doesn’t matter much for normal automotive interior lighting, but if you’ve got a caravan or motor home or something that has a lot of friendly yellow incandescent bulbs in it, it could be a very good idea to swap them for the new cheap LEDs.

Remember those lame Internet filters which my faithful readers helped the smut-hungry youth of Australia to dismantle, last year?

Well, the whole taxpayer-funded content-control software handout program has now officially been declared (by Australia’s new Federal Labor government) to be a miserable failure.

Apart from the fact that the NetAlert packages were quite easy to get around, it turned out that nobody actually very much wanted them.

The Government predicted that 2.5 million households, about 31% of the whole country, would want their free copy of one or another of the packages (which they’d paid for with their taxes already, of course).

As it turns out, they got a grand total of 144,088 CD orders and downloads.

And not all of the people who got the filter software bothered to use it. The ridiculously-named government department responsible says only about 29,000 of the packages were actually installed.

That’s 1.2% of the target, for those of you keeping score at home.

The total price of the software filter scheme was 85 million Australian dollars. That’s about $US78 million, at current exchange rates.

So this software ended up costing the taxpayer about $AU2930 ($US2685) per installed unit.

A copy of Net Nanny will cost you $US27 from Amazon. That’s almost exactly one per cent of the effective price of the “free” software.

All that, to stop red-blooded Aussie kids from seeing boobies and doodles.

But have no fear - the new Federal government is much more sensible! They enthusiastically explain that their own very expensive scheme to implement “mandatory ISP-based filtering to deliver a filtered feed to all homes, schools and public internet points” will work far better. You know, just as it has in the other countries that’ve implemented secret Internet blacklists which, in effect, accuse lots of random innocent people of being child pornographers.

Never mind that, despite more than $15 million worth of advertising (including a booklet sent to every household in the country), it is now demonstrable that approximately three-fifths of bugger-all Australians have any interest in filtering their own Internet connection.

No, never mind that. We must be protected from filthy filthy porn, whether we want to be or not!

This is all more evidence that, as I’ve said before, it doesn’t matter whether censorware works. Which is good, because it generally doesn’t.

The purpose of censorware is not to Protect The Children, but to get some people elected and keep other people employed.