Dan's Data letters #194Publication date: 29-Dec-2007.
Last modified 03-Dec-2011.
First off, I think your old page about NiCds is very well done. I didn't know about M-size NiCds, so I have learned something today.
I have also stumbled upon a good source of AA rechargeable nickel cadmium batteries, if one is willing to take a chance and visit the local rubbish tip to look for the solar-charged LED lights which push into the ground.
Inside these solar lights resides either one or two AA-size NiCds. I have found about fifty of these units over the past years. Most people never give a second thought about what might be wrong before throwing them away.
There is only one fault with them; the spring contact to the negative side of the cell becomes dirty. Then it won't recharge, the owner looks outside one night, sees the light "stopped working", and throws it away.
On board these things is a charger, solar cell and the batteries. I will be making a rather large solar charger one day, but for what it is worth, there are the "free" NiCds.
OK Dan - what do you think of this battery supply?
Solar garden lights are a wonderful resource. Why, they seem to grow naturally in suburban neighborhoods. Just go for a walk in the very early morning and you can pick yourself several dozen!
(Given the low value of garden lights even if you buy them brand new, stealing them is probably not actually worth the risk. I am, however, firmly in favour of the idea of transplanting several streets worth of garden lights into the garden of one unsuspecting house.)
The actual quality of the components in cheap garden lights (as opposed to the $250 type) is, of course, not very high. I don't know the exact efficiency of the solar cells, but you can bet on them being pretty lousy; likewise, the NiCds are low-capacity, and probably haven't exactly benefited from sitting in a hot little plastic box in the sun for a year or two.
That said, the NiCds definitely won't have been charged hard, and may not have over-discharged either. Running a single NiCd cell flat is fine, but running a whole series-connected pack of them flat will cause the stronger ones to "reverse" the weaker ones, which is very bad for them.
Garden lights need a certain voltage to work at all, though; once the LEDs don't get enough volts across 'em to light any more, the discharge rate probably drops to nearly zero and the pack doesn't get to completely flatten itself.
(The circuits in garden lights are rather elegant, by the way. Here's a page that covers a couple of them.)
Cheap garden light LEDs may be fine, too - just as good as the ones you get in cheap LED flashlights. Chinese knockoff LEDs are pretty decent these days and, once again, garden-light LEDs are unlikely to have been pushed very hard, even if the light is years old.
And, of course, you're doing the environment a favour by policing up old NiCds from rubbish heaps.
NiCd disposal here in Australia (Jim is in Queensland) is still, I think, not a solved problem. The Australian Recycling Near You site is a much better resource than existed back in 2002 when I wrote this, but it's just informed me that Australia still "does not currently have a national recycling scheme for primary and secondary/rechargeable batteries", and the best option seems to be something like this business-targetted postage-paid box service.
Those four-point-eight-litre boxes cost $AU55, though, which is rather more than it's likely to cost to deposit a big old trailer full of unsorted (but, you swear, completely non-toxic!) waste at a quite restrictive local dump. And the recycling service says batteries it receives are then "shipped overseas to a licensed recycling facility where the cadmium is recovered", which given the generally laughable "licensing" standards for "green" industry (even right here in Australia...), suggests to me some hell-hole in China.
Junked garden lights probably aren't a really major pollution source, but cadmium leaking out of rubbish tips is still not a good thing.
You may also find it easier than you'd think to make solar panels out of garden light cells. There's at least one quite exciting super-cheap low-efficiency solar panel technology that's supposed to make it to market quite soon, but in the meantime even basic solar panels are still expensive enough that a few afternoons spent soldering together scavenged cells could be well worthwhile.
Packaging and mounting your panels, so that they're pointing the right way and won't drown in the rain or blow away in the middle of the night, is more challenging. Speaking with the confidence of someone who's never had to do it, I suspect aluminium window framing supplies would be helpful. A search just turned up this, too, but I've no idea if it's actually worth buying.
I enjoyed your page on computers in space, and thought you might be able to help me find some related info.
(Note: This is remarkably unimportant, but possibly fun.)
As HAL did all the stuff the International Space Station's life support system does, plus a bunch of other control stuff, plus a bunch of AI stuff, it seemed reasonable to guess that it would be at least as big as the computer systems that run the ISS.
I've looked around for info on computers used in spacecraft in general and the ISS computers in particular and found some interesting info on architecture, interconnects, etc, but precious little about the physical sizes of things.
There are too many variables for me to guess.
Since HAL is a computer of a type that we still have absolutely no idea how to build (even though he "became operational" in 1997 according to the original book!), 2001's wall of crystals is as good a guess as anyone's. True intelligent computers seem to be about as far away today as they were in 1968, the imaginings of "singularity"-envisaging techno-mystics aside.
Clarke's vision of HAL as a large machine was of course informed by the size of computers of the time. If he were writing 2001 today, he'd probably make HAL a lot smaller. Today's TV and movie sci-fi megacomputers are often pretty huge as well, though - I think mainly so that the actors can have impressive drawers full of glowing crystals and giant circuit boards to interact with.
(It'd be pretty funny if, for once, our heroes had to hack the alien mainframe to stop the mothership from blowing up Earth but couldn't do so because, thanks to the vastly superior alien technology, the whole alien command and control system was the size of a pinhead and completely wireless, and could be anywhere inside the Death-Star-sized ship.)
The core ISS systems, on the other hand, are based on very un-sci-fi technology - they're still based on 80386 variants. And they'd probably all fit in one big server case.
I don't know how they're actually physically laid out, though. Since there's triple redundancy for both the command and navigation computer systems, it'd be idiotic to put them all in the same place.
If you needed to, though, I'd be surprised if you couldn't fit the ISS's whole computer system into a "bar fridge" server case, at most.
Can you, with all your knowledge on rechargeable batteries, shed some light on the life expectancy of the battery used by Toyota in their Prius?
On their site they mention that they have lab data showing the equivalent of 150,000 miles with no deterioration.
On another site, I found they say it lasts 150,000 to 200,000 miles or ten years, but they also state a 80% efficiency after just 10,000 recharging cycles (whatever that means in a product that is continually partly charged and discharged).
The Prius uses a NiMH battery, and I always understood that NiMH doesn't last forever. Has Toyota developed a new type of battery, or is their "lab data" the car salesmen term for "we made it up"?
There's nothing technologically remarkable about the Prius batteries, but the car treats its battery very gently. The battery's never allowed to get very close to empty, and it's never allowed to completely charge, either; the car's control systems try to keep the battery between a 40% and 60% charge state.
This technique means you're carrying around (much) more battery weight than you need for a given capacity, and it also goes a long way toward explaining why the Prius EV mode isn't very useful. But it also considerably extends the battery's life. As you've noticed, Toyota say that the battery will last for the life of the car, but they're only defining that as seven to ten years.
The Prius is officially more than ten years old now (the very earliest Japan-only models were on sale in 1997), and it does indeed seem that these claims are quite reasonable. Even old Priuses seem virtually certain to still be electrically fine up to around the 150,000 mile mark, and apparently Prius taxis in Canada have cruised past 200,000 miles with their original battery.
I'm sure some people will still be driving 1997 Priuses in 2025, by which time they'll have gone through a few batteries. But seven to ten years - even in a heavily-used car - is a pretty good lifespan. It's not like those silly "electric supercars" that'll probably need fifty thousand dollars worth of new batteries every three years.
I saw your (old?) note on hatin' on lithium ion. Are you still hatin'? I'd like your opinion.
I'm working on a PHEV (plug-in hybrid electric vehicle) project and looking at the battery choices.
Yes, LiI batteries are denser, but they are also more expensive. And then there's the blowing up thing. The A123 batteries are safer, but they lose most of the energy density advantage, so I say, why bother? Getting a practical PHEV cheap enough to make it to market is a major hassle, so tossing extra cash away on the batteries does not seem to make sense.
Anyway, I don't know if it's the lithium proponents that have been drinking the magic Kool-Aid, or it's me (in that I seem to be the only person, other than you perhaps, that still likes NiMH batteries). It's possible the PHEV people are looking to lithiums because of the Cobasys patent thing, but that's not a long-term reason to pick a poor technology.
So, if you are still wary of lithium ion, I'd like to hear more.
LiI technology's made considerable headway in the last few years (my hatin' piece is from 2004). There's always some new exotic nanotechnological 500%-capacity-increase super-cell on the horizon, but plain old 18650-size (and similar) laptop-battery cells are mainstream eBay products these days, and really don't cost very much for the capacity you get, even if you assume that in automotive use they'll only be good for a couple of years.
If the lithium battery quality improvement trend continues - which is likely, since the technology is less mature than NiMH - then you could make your vehicle sufficiently modular that the batteries you swap into it in 2010 can have a whole different cell form factor to the ones you put in it originally. Then, even if you do indeed need to change a big battery pack in a couple of years, I think LiI power could still be a winner, given its energy density advantage.
There could be further issues associated with a change in the battery specs, of course - control systems, charging - but I think there's a good chance that you'd only need a firmware update to deal with them.
It should be noted that the only way to get really dramatically better lifespan out of NiMH cells in automotive applications is by babying them, like the abovementioned Prius system that never lets the battery get close to full or empty. Without it, Prius batteries would probably be seriously degraded after four or five years, which isn't really that much better than what you're likely to get from current LiI cells.
But, of course, what the hell do I know about vehicle batteries. The biggest electric vehicle in my life was built from a Tamiya kit.
I'm about to throw my Dell Inspiron 9200 out of the second story window. All four USB ports just, of their own minds, but all together, ganged up on me and quit their jobs the other day.
I did nothing to offend; I wasn't abusive, and I did actually respect them.
I've done everything from deleting drivers, uninstalling, re-installing, expanding in Device Manager, trying them with a new OS (Linux), and they are not coming back. I will wipe my system and re-install Windows XP to see if that will coax them from hibernation.
If not, what do I do? New computer? PCMCIA card?
Welcome to Dell - please enjoy your stay!
Dell laptops are often very good value indeed, but they're not terribly durable. The Inspiron 9200 I bought for my sister is still 100% functional. I don't think it's even lost a rubber foot.
The 9300 we bought for here a bit later, though, lost its built-in sound and WiFi adapter after a few months (could have got a warranty repair, but I never got around to it).
If your computer's still in warranty then send it back and they'll fix it. If it's not, there's little chance that you can economically repair it. If the ports don't work in Linux, I'll betcha they ain't gonna work in reinstalled XP either.
It's just barely possible that the problem is something that's fixable with a soldering iron - though all ports being dead is not a good sign. On the plus side, Dell's online service guides are surprisingly excellent; they'll talk you right through taking the computer apart and putting it back together again (which long-time users will have to do, to clear dust-plugs out of the heat sinks), and you won't need any funny tools to do it.
But if a chip's breathed its last and you don't know someone who does surface-mount rework for beer, you're out of luck if you can't find a whole new 9200 mainboard on eBay for cheap.
Fortunately, you certainly can work around the problem with a cheap PCMCIA USB card, and the ports that card gives you will probably be more widely compatible than the originals were. That's what I'd do.
I am happy with the Sennheiser PX 100 headphones I own. Good reviews on the net, including yours.
I take the bus to work and fly reasonably often, though, which leads to my question - do you know of any noise cancelling box that comes without headphones? I want to keep the PX 100s but get noise cancelling, and I hope it could be cheaper than buying combo headphones.
I searched the Web, which is how I found you. But I did not find headphones.
I am also a Nigerian princess who is stuck in prison; the guards will only release me in return for a GeForce 8800 GT. Please help.
No such thing exists, I'm afraid. Active noise cancellation relies on the circuitry knowing what noise is reaching your ear-holes, which it cannot do if it doesn't have microphones in - or at least on - the earphones themselves.
It'd be technically possible to create a device with a couple of mics that you hang off your existing headphones, but it'd obviously be pretty inelegant, so nobody has. Well, not commercially, at least - I'm sure some hacker's brewed up some Borgish abomination that works quite well.
If you haven't considered canalphones, I suggest you do. They're the absolute best option for excellent portable sound and high ambient noise rejection, and there are plenty of models that don't cost very much these days, too. The only down side is comfort, or lack thereof.
Suppose I wanted to maintain a vacuum in a standard bell-jar/vacuum plate setup for around 10 millennia, give or take, without interruption or maintenance. What, other than "comical," would be my options?
These thoughts occur:
Short-term thinking would assume an electric vacuum pump, but I'm assuming that is preposterous due to the mechanical degradation of the components. Powering the electric device bumps into the same hurdles: Solar and wind are out due to mechanical degradation, right? What about radioactivity? Is there any reasonably sized un-deadly hunk of radioactive ore that could be used to provide electric power to the pump while simultaneously requiring zero maintenance?
Maybe I could even use the heat from a radioactive source to more directly drive a steam pump in a sealed engine to maintain vacuum... thoughts?
I should mention that the vacuum is only for purposes of removing contaminants from the air (most specifically sulfur), so it doesn't need to be very strong.
I presume shooting the vessel into high Earth orbit is also not possible.
You can forget about any kind of mechanical pump. It's barely possible that some jewel-like device with glass cylinders and graphite pistons and diamond bearings could do the job, but it's not as if you'd be able to demand a refund from the maker's descendants in 150 years if it wore out. I bet, eventually, it would.
Actually, I wouldn't be surprised if a pretty ordinary refrigerator compressor - easily convertible into a vacuum pump - would last a hundred years if run on a relatively short duty cycle (see also The Secret Life Of Machines, episode two). Even one thousand years is well over the odds for any mechanical pump, though.
Power sources aren't entirely impossible if you only need a little power - perhaps some kind of expanding-metal-driven ratchet spring-winder that clicks the spring a few times every noon and midnight, or a granite cistern feeding a waterwheel that's lubricated by its own water - but all mechanical vacuum pumps require close tolerances, by definition, and close tolerances invariably get less and less close over the span of years, not centuries.
There are, however, several kinds of vacuum pump - or related device - that have no moving parts at all.
The only one that I think would be feasible over really long periods, especially if the only job you want it to do is police up impurities in an evacuated vessel, is a great big metallic "getter".
It'd just sit there passively in the vessel - which could be vacuum-filled, or possibly even filled with an atmospheric-pressure noble gas if that's workable - and it soaks up the unwanted molecules. Easy.
Getters can also be helped along by an ion pump, but I don't know how you'd make one of those that'd last for a very long time. Yes, a radioisotope thermoelectric generator could power one for a long time, but the ion pump itself would degrade, and RTGs by definition have to be based on the kind of excitable isotopes which, to quote Dr Emmett Brown, are not available at every corner drugstore.
The other no-moving-parts options, like cryopumps and sorption pumps, are out of the question on account of how you'd need to set up an unusually durable religion based around supplying your vacuum bottle with liquid nitrogen.
What is your expert opinion regarding this YouTube video that's been making the rounds:
The guy goes through a ridiculously roundabout method of connecting the cell to the TV, but he's basically saying that you can power a television from a AAA cell, if only for a few seconds.
Is this even in the general vicinity of the realm of possibility? Even if the TV were somehow able to accept DC when expecting AC (which I suppose is possible if it's just a simple diode setup), as well as having the DC-DC converters needed to step up the voltage where required, wouldn't some monstrous amount of amperage be needed at the cell to perform this?
Can a AAA provide this type of power, even if only for a few seconds?
Given the recent popularity of fake instructional videos, it's not a bad idea to check out any Web sites promoted in a fishy-looking clip before you bother to expend any thinking time at all on it.
Like all of their other videos, this one is all fake.
I'm also inclined to think that it's not actually all that funny, if only because kids who don't have enough of a knowledge base to have an effective bullshit detector are likely to try these things, fail completely, and get turned off the idea of science-at-home altogether.
(I address one of the more successful perpetrators of this sort of thing here. He doesn't even do us the favour of lying all the time.)
It is, actually, possible to power large loads for a small period of time from all sorts of low-power sources, if you build an appropriate power supply (in this case, a DC-to-DC step-up converter, a big-ass capacitor and a commercial inverter would get you there most simply).
There's not much point to doing this, of course, and nothing in any of the Household Hacker videos will lead you in anything like the right direction.