Dan's Data letters #122
(page 2)
Publication date: 17 August 2004.Last modified 03-Dec-2011.
Part of every well-stocked bar
I have this rather large bottle of "Isoclene" that I acquired from work, which is 99.7% pure isopropyl alcohol. I used it to clean thermal gunk off an old CPU and heat sink, and it did a great job; I also cleaned my mouse and keyboard with it. Now there's still a good load of it left, what else can I use it for (aside from the obvious flamer)?
Would it damage the anti-glare coating on monitors, for example?
I've got about 200ml left, and I'd hate to see it go to waste just sitting in the bottle like this.
RW
Answer:
99.7% pure? Not any more, it ain't.
Short-chain alcohols (isopropanol is the third-shortest alcohol molecule; methanol is first, ethanol is second) are hydrophilic; the less water there is mixed with the alcohol already, the more enthusiastically the alcohol will draw water out of the air to dilute itself. This matters for cleaning alcohol, because the water will take much longer to evaporate than the alcohol and may thus leave you with spots or streaks on the thing you're trying to clean. It matters for various other applications, as well; it's the main reason why methanol-based model engine fuel will "go bad" after a while, even if its plastic bottle is kept tightly capped.
What else can you use it for?
Well, you'd need a few other substances to start a meth lab, so that's probably out. You probably also don't do a lot of nucleic acid sequencing. About 80% concentration isopropanol is a good general purpose solvent for cleaning plastics and various other things, though - including wounds, if you're not a little wussy girl afraid of a bit of a sting. Isopropanol's poisonous, though, so it's definitely for external use only.
And yes, it probably would damage monitor coatings. As I've mentioned before, optical coatings are often not instantly and dramatically damaged by cleaning them with alcohol, but it's still not a good idea. As long as your monitor is not plagued by the greasy fingers of those whose death will be slow, yea, slow and agonising, as will that of their relatives and friends, verily, even unto the seventh generation, you shouldn't need more than breath-fog and a tissue to clean its glass, anyway.
Mission: Inadvisable
I've dabbled in photography a bit. Took (part of) a class in general photography at the local tech school, etc. Anyways, due to the high cost of a wedding photographer, my best friend has inquired about my shooting his wedding for him.
My camera is currently a Canon EOS Rebel XS. My lens is a Sigma 28-80 zoom, but something is broken. I believe the lens is bad as the aperture won't open all the way and I can't take pictures. I've taken it to a camera shop and the camera worked flawlessly with a different lens, thought he problem has worsened since then.
Biggest problem is that when you look through the lens, it's darker than it should be (due to the aperture issue I mentioned). Also, when you press the button, the camera clicks and the mirror moves out of the way, but it never actually takes the picture, no film advance or anything. Autofocus is also busted and doesn't work right.
So I'm going to need a new lens, assuming that is, indeed, the problem. I've looked around and the best deals I've found were for a Canon 28-90/f3.5-5.6 zoom or a 28-105/f4-5.6 zoom, the former was about $US100 and the latter was around $US120 or $US130, plus any taxes and shipping. I'm not adverse to other brands either.
What other equipment would I need? I'm guessing a tripod (would a basic $US20 to $US30 model do?) and probably a better flash (I know NOTHING about different flashes. Any suggestions?). Is there anything else I would need to look into getting for it? Are there any beginners books (Wedding Photography For Dummies or something) I could look into? I would like to make the pictures as close to professional as I possibly can manage.
If I can pull this off right, I might even look into doing some sort of wedding photography as a side deal to make a little extra money. Average seems to be around $US2000-3000. I'm not that good, may NEVER be that good, but I'm figuring maybe something like $US300-500 for my time plus pictures might work. At that point, of course, I would DEFINITELY have to upgrade my equipment. But what I've got will have to do for now.
Oh yeah, my budget for any of this stuff is going to be severely limited. Probably no more than around $US300-350 tops for the stuff I need to buy.
Grey
Answer:
The usual advice of professional and semi-pro photographers to people in your situation is:
DON'T DO IT!!
The reason for this advice is that weddings are a high-pressure photographic environment, where you really need at least a couple of cameras (to avoid time-consuming lens changes), and need to scope out the pre-ceremony, ceremony and reception locations beforehand (for locations and lighting, and to sound out any relevant persons of the cloth about where you can stand without being consigned to hell), and need to have your act together generally as a high-throughput, bossy commercial-quality human-herder, lest you miss one of the numerous only-happens-once scenes essential to a proper photographic record of the occasion.
Everybody who's ever been asked by a friend to shoot a wedding has been told that there's no pressure, it doesn't matter if you miss a couple of things, nobody will be angry, blah blah blah. The opinion of most professionals appears to be that these assurances are likely to be hollow. Even if the bride and groom don't get angry with you about whatever you screw up (and, if you haven't shot weddings before, you will screw up something, and possibly everything), there'll be some important relative of theirs that'll put you on their People To Kill list if you miss The Ring, or The Cake Cutting, or The Reverend Sun Myung Moon Blessing All 2500 Of Them, or whatever.
Hence: DON'T DO IT!!
The punchline: In May this year, I shot a friend's wedding. I was assured that there was no pressure, it didn't matter if I missed a couple of things, nobody would be angry, blah blah blah.
These assurances turned out to be entirely valid, and everything went fine.
Some of the pictures came out even better than this one.
It should be noted, though, that I had my Significant Other to tell people what to do for me, and was also toting a pair of cameras - my own EOS-D60, and a borrowed EOS-10D. And a bag full of lenses. For indoor shots in the chapel I used a cheap 50mm prime, because it goes to f1.8 and thus lets in 3.8 times as much light as the maximum f3.5 aperture of my everyday 24-85mm zoom. Those two lenses would have been adequate for the whole thing, but I got some use out of a couple of longer and shorter ones too.
If you're determined to try to do the wedding-photo deed with your old film (!!) camera and a crappy consumer zoom lens, the problems you mention do indeed mean that you're going to have to get, at the very very very least, another lens. The Canon 24-85mm costs the thick end of $US300, but you could get away with the 28-90mm you mention, which is often bundled with cheap Canon backs these days; the price is right, and it's not rubbish. It starts at f4, though, which means exposures about 1.3 times as long as f3.5, all other things being equal. The 28-105 you mention is affordable, too, and another not-too-bad choice, though both of these cheaper lenses can't be expected to live as long or deliver quite as nice-looking results (less sharpness, and probably a bit less contrast too) as the more expensive zoom. All of these lenses are wide enough at full wide angle to still be pretty usable when you put them on a 1.6X-focal-length-multiplier camera like a D60 or 10D.
Budget a few bucks for a basic UV filter of the appropriate size for whatever lens you get, as well. You don't necessarily need it, but UV filters are the cheapest kind and they keep the lens from getting scratched. Better to throw away a scratched $10 filter than a scratched $200 lens.
Your major problem, of course, is the lack of a digital back to put the lenses on. I shot five hundred and fifty-six photos of my friend's wedding, and that's not at all an unusual number for an event like that; if I were shooting film then I'd probably have shot fewer, but if I were shooting film with only one camera I'd be constantly worried about finishing a roll at a crucial time.
Bear in mind that it's possible to rent digital backs, and buttloads of CompactFlash cards to go with them. Seriously consider doing this. Film is a big, big limitation when you're shooting an event like this; not having to worry about how many exposures you shoot reduces the pucker factor considerably. If you're on a shoestring budget, poll everyone you know who owns a digital camera and harvest as many of them for the day as you can (and put aside an afternoon to learn how to use them). Presto; lots more storage, no purchases needed.
If you don't take the bag-o-digicams option, you must have a decent add-on flash, unless this is a noon-time beach wedding. On-camera flash that irons out everybody's face into a frying pan is not acceptable for anything but the basest of happy snaps. So, while you're renting a D60 or 10D or whatever (give yourself a few days to get familiar with the rental camera, if you do), also rent a 550EX flash (now obsolete and replaced by the slightly superior 580EX and 580EX II). Reasonable power, tilt-and-turn, and a slide-out reflector-diffuser that you can use to put a glint in people's eyes while the main flash power shoots at the ceiling to illuminate them the civilised bounce-flash way, instead of the bunny-in-headlights direct-flash way.
Can't get a 550EX? Consider a 420EX (replaced by the 430EX). Can't afford that? Consider a Vivitar 850AF, or Sigma EF 500 DG, or Promaster 5550DX, or something. Anything you can angle up to bounce off the ceiling (assuming the ceiling isn't 30 feet up and dark brown...) is better than on-camera flash, provided it integrates with your camera's auto-exposure, or you've got your exposure act very thoroughly together and can set things up manually.
Other equipment? Yes, a tripod could be helpful, and even a cheap 'n' crappy one is better than no tripod at all, especially when it's brand new (don't expect a cheap 'pod to last for more than a couple of years of occasional use). You may well be able to get away without a tripod (I didn't need mine), but for posed pictures of lots of people, a tripod is indeed a good thing.
Here's a column that may throw a scare into you. Here's plenty more on the subject.
Nice try
I would like your input on an device I thought up during a particularly boring Computer Science lecture, that may increase the efficiency of single direction rotating machines such as servos.
My problem is that I have no formal training in physics beyond year 12 level, and so I cannot prove or disprove my device. The main problem is that the device uses magnets... Yes, I know, the dreaded "M" word! I am wondering whether you might be able to tell me if I am stupid and the device will never work, or conversely, tell me I am on to a winner.
I have tried contacting people in the Physics department, however they do not answer e-mails. So I have tried finding kinetic and magnetic modeling simulator software, so I can make a model to test out the device. However they all seem to be very expensive and not friendly for the uneducated.
I have attached a diagram of the device my bored mind created:
Basically, the star magnets can only rotate in one direction. So when the free moving magnet gets close enough to the star magnet it is attracted to the star magnet (1). The star magnet, while is also being attracted to the free moving magnet, cannot rotate in that direction.
When the free moving magnet hits the star magnet the molded plastic shell causes the star magnet to rotate (2) (3). I am hoping that there is sufficient momentum combined with the repulsion from behind and attraction from in front on the free moving magnet to minimize the force needed for this rotation to occur.
Once the rotation is complete (4) the free moving magnet is repelled away from the star magnet. As with the first case, the star magnet is also repelled from the free moving magnet but since it cannot rotate in that direction it stays still.
And thus the free moving magnet is repelled with sufficient force to move to the next star magnet (5).
At the moment I am thinking of using a ratchet cog setup to allow rotation of the star magnet in one direction. However I am sure there is a better solution.
The potential machine as I envision it is also included on the diagram. The device would be attached to the rotating axle of an electric servo (or other rotating device). I didn't want to classify it as a perpetual motion device - as I understand it, magnets lose their strength, so it would eventually stop.
Nicholas
Answer:
I think that every single basic combination of rotating, ratcheted, counterbalanced, one-way-bearinged, geared, belted,
chained, greased, dusted, painted, submerged, flying, pendulous, and otherwise mechanically and magnetically coupled
magnets has already been invented.
"New" designs pop up all the time, some of them baroquely complex, but I think all of them by now can be reduced to trivial combinations and modifications of previous designs, at least when they involve simple machines and magnets.
What every single one of them actually does, which you can figure out by tedious vector mathematics or by rigging up a test unit out of NIB magnets, epoxy and Lego, is either click into immobility after the initial push is spent (for a machine like yours where the components touch), or vibrate into immobility (for machines where the parts don't touch, and only interact via the infinite "springiness" of magnetic fields).
If we ever develop a magical magnetic shielding material that just silently and passively eats every magnetic force line that enters it, spitting it out again in the place the magnet expects it to come from if the shielding isn't there but not allowing things on the other side of the shield to see the field, then various perpetual motion devices will suddenly work. But without that, none of them ever do, yours included.
(Note that a Halbach array is, I'm sorry to say, not this magical shielding material.)
I know you only bill your invention as an efficiency increasing device, but it stands to reason that if it produces a non-zero rotational impulse of its own, then it can serve as a perpetual motion machine. This, I venture, is probably why the Physics people don't answer your e-mails.
Your it'll-eventually-stop disclaimer has actually been used by a number of perpetual-motion hucksters - "It's not perpetual motion, because the magnets will lose their juice, or the bearings will wear out, or one day the sun will swallow the earth and the machine'll melt..."
These are invalid exceptions. A perpetual motion machine that wears out - as everything eventually will - is still a perpetual motion machine, as long as it generates its own motive energy from nothing.
Some perpetual motion scam artists have a cannier version of this argument. They say the machine is actually converting parts of itself (usually magnets) to energy.
A machine that actually did that could generate an incredible amount of power by consuming only a small part of its mass; e=mc2 tells us that total conversion of only one gram of matter will give about ninety trillion joules (watt-seconds; one joule is one watt for one second) of energy.
A really large nuclear power plant (which really is converting mass to energy, though horribly inefficiently) can generate about 4000 megawatts of electricity. That's enough to run a big city; San Francisco at the moment apparently only draws about 650 megawatts during the night and about 850 during the day.
Ninety trillion joules is what you'll get if you run such a monster nuke plant at full blast for six and a quarter hours.
So a magic conversion machine that outputs only a few watts of power will, conveniently, consume its mass so slowly that it'll be virtually impossible to measure the change.
Of course, a total conversion machine probably ought to emit a considerable amount of its total energy output as a sleet of hard radiation. But the scam artists kinda gloss over that.
On that subject, nine trillion joules is also equivalent to about 2.15 kilotons of TNT. So if you're ever given the chance to play with such a machine, I suggest you not drop it.
You may find The Museum of Unworkable Devices entertaining.
