Dan's Data letters #206
Publication date: 18-Dec-2008.Last modified 28-Apr-2012.
In case you have no spare hard drives
I thought you might be interested in another source that I found for a big handful of rare-earth magnets for under $20 - http://www.thinkgeek.com/geektoys/science/770f/. And I also found the strength information provided at Magcraft to be extremely useful, or at least interesting. Not sure that I've figured out what strengths I actually need just yet...
Sonya
Answer:
You don't get that big a handful from ThinkGeek; even the $US29.99 "larger piece
set" only weighs 160 grams. The packaging for these and the Magcraft-branded ones is
nice, though, with the magnets in a small tube inside a big tube to keep down the external
field strength. Magnet dealers usually have to ship their wares in steel-lined boxes,
to prevent them from messing with and/or attaching themselves to mail sorting equipment
and so on.
I also don't know whether the nickel plating on the brand-name magnets is any better than what you get with no-name magnets. Flaky plating is the only real quality issue you need to worry about, especially for spherical magnets, which have very small, high-pressure contact points.
Magnet strength isn't important unless you've got very critical engineering requirements. Even cheap magnets are likely to be N38 or something, which is pretty darn strong.
ThinkGeek have a lot of cool stuff, and they're speedy and professional, but the price you pay is... the price you pay. You can get considerably more and/or bigger magnets for the same money on eBay. Most of the eBay dealers want to sell you a bunch of identical magnets rather than an assortment, but you can get assortments elsewhere.
The cheapie-vendors at DealExtreme also have a modest selection of NIB and ferrite magnets, all with worldwide shipping included in the price.
Magnets and meteorites
I have read on the internet that a magnet on the end of a walking cane will identify some meteorites, as they are attracted to magnets.
Do you have any experience or recommendations about what kind of magnet or combination of magnets might be useful, and how exactly to use them? Will using a rubber tip on the end of the cane, and as a protection for the more fragile magnet, diminish the attracting power of the magnet(s)? Is there a market for sales of a tested and proven product such as this? Do you want to go into the manufacturing partnership with me and get rich?
J
Answer:
A magnet on the end of your walking stick will collect any small magnetic objects it
touches, but most of them won't be meteorites. Many
micrometeorites aren't nickel-iron,
either; a magnet won't pick them up.
If you just want to collect magnetite particles, it's easy to do so with a bagged magnet - tie a tough plastic bag to a string, put a neodymium-iron-boron (NIB) "rare earth" magnet or three in the bag, drag the bag through sand or dirt, and lots of particles (and the occasional bottle top, bit of wire, et cetera) will accumulate on it. Doing this with a bag lets you pull the magnets out of the bag and have the magnetic stuff just fall off; if you use a bare magnet, you'll have a hard time getting the magnetic dust off it.
If you want to collect micrometeorites in particular, you can do it by harvesting them from your roof. A significant amount - though certainly not all - of the magnetic dust that washes off the average roof during a rainstorm is micrometeorites.
It's easy to sort magnetic micrometeorites out from the rest of the roof-crud. It's considerably harder to find the "stony" non-magnetic ones, but you can still coffee-filter the remaining roof-crud and go through it with a strong magnifier to find things that look melted. Note that you may still just be collecting volcanic ash and other stuff that just happens to look like little melted blobs.
You can also use a plastic sheet to do this.
If you want this to be a walking-stick-shaped product, then you could use a rod down the middle of the stick attached to a movable magnet inside a hollow end to the cane, so you can draw the magnet up away from the end of the stick by pulling on the other end of the rod. Then, anything collected by the magnet would fall off. Magnetic pick-up doodads that work this way already exist.
With a chunky enough NIB magnet in the end of the stick you could make the casing around it quite thick enough to protect it from normal walking-stick damage. Note, however, that the stick would also try to trip you over if it stuck to a manhole cover, or something.
Dropping volts
I live in a country where 220V is the common voltage on the walls. I've purchased a multi-effects processor for my guitar that works on 110V. No problem, I just have to get a step-down transformer right?
Now while I'm checking out my options, there are so many different sizes and ratings, and I was wondering HOW do I calculate how much I really need? I know that bigger is always better/safer, as far as having "headroom" is concerned - but of course we all know how HEAVY these transformers can get. Since I want to use this for live gigging, I need it to be pretty as light and portable as possible.
I'm thinking that it would be ok to get the lightest/smallest one since I only intend to use the transformer for that ONE unit - but of course I want to make sure MATHEMATICALLY what is the SAFE rating actually is for my particular situation.
Which is why I thought of going the extra mile and ask you exactly how you calculate these things as well. I'd love to be able to apply the "rules/equations" you plan share to any electrical situation.
Carlo
Answer:
A guitar effects processor will, as you say, not draw much power - maybe 25 watts, but
it could easily be less than ten. So just about any step-down transformer should work
fine.
Note that if the processor has a simple outboard AC adapter, it's a better idea to buy a matching 220V (or, these days, switchmode world-compatible) adapter with the same specs and output plug. Music gear that runs from plugpacks often wants nine volts DC, regulated, but there are various other possibilities; just match the output voltage and current rating of the plugpack the thing comes with and you're done.
If the processor plugs straight into the wall, this is of course not an option. But even if it has a complicated multi-pin, multi-voltage power supply, whoever sells that brand of music gear in your country may well have power supplies to suit.
Assuming that you have to use a step-down unit, bear in mind that:
1: Cheap step-down transformers ones can be unreliable. They're unlikely to blow up, but they may well just drop dead after not many hours of service. You can get some very cheap step-down transformers if you only need a low power rating, but the cheapest ones are not necessarily a good buy.
2: If your effects doodad has an earthed (three-pin) plug, make sure you get a step-down transformer with a three-pin socket, and also make sure the third pin is actually connected to earth (this should be mentioned on the box, but you can also just use a multimeter or other continuity tester - if the earth pin on the transformer's input plug is connected to its output socket, you're fine. Otherwise, using the transformer will un-earth your appliance, which is fine until the kind of fault that earthing protects you from occurs, like a live wire coming free and touching the inside of a metal casing.
For this application you don't really need to calculate how big a step-down converter you'll need, because even the weediest units should be able to handle it. But the important thing to remember when dealing with AC power is that while devices are rated in watts, what they actually draw is volt-amps (VA), and for non-resistive, "reactive", AC loads the VA will always be bigger than the watts (for DC, VA always exactly equals watts, by definition).
Many power-supplying devices, including step-down and step-up transformers, should be specified in VA but are instead specified in watts, often optimistically. A "250W" step-down transformer will only be able to run a 200-watt load, if that load has a "power factor" of 0.8.
I talk about this more in my old piece here, and on several other pages. Try this and this to start.
Great Scott! A genuine product!
Could you tell me if Brick Wall surge protectors seem legit?
I've been toying with the idea of a DIY UPS. But after Google-ing around and finding this, I'm now wondering if it'd be enough to buy one of the basic Brickwall models and a cheap UPS to plug into that? Since, you know, the landlord may not appreciate the beauty of a homemade power unit.
Mark
Answer:
Yes, the Brick Wall surge protectors look pretty kosher to me. They are not
magic rip-off boxes.
I don't have any personal experience with them, since they make almost no gear that's usable in 230V Australia. But once you wire-brush off the marketing paint, their product claims seem plausible enough. It is indeed possible to make a rock-solid power filter that actually does what all of the cheap plastic ones say they do, and the Brick Wall ones probably are that product.
Note that they are still unlikely to be as good as a proper ferroresonant conditioner. But they also don't weigh 30 kilos each.
While I wouldn't swear to it in court, I think the deal with Brick Wall is that they sell boxes made by the same people who make the Zero Surge boxes; I think Zero Surge started selling boxes under that name after OEMing for Brick Wall for a while. I think there's some commonality between those two brands and SurgeX, as well.
This doesn't mean there's some sort of scam going on (well, except for the marketing whenever companies that sell the same product call their version "unique") - I just mean that if you're thinking of buying one of them, you should check out all three and see if you can get the same numbers for less from one of the other brands.
Now - is it a good idea to buy a basic Brickwall, or similar, and plug a cheap UPS into it?
Well, that depends on how cruddy your power actually is. If you live somewhere where the mains power is up and down like a yo-yo and lightbulbs detonate on a daily basis, then serious power filtering is a very good idea (though nothing is going to save you if the lightning strikes right outside your house). If you live in a typical suburban area in a First World nation, though (which includes, oh, at least 60% of the USA...), then the not-so-great filtering included in a cheap UPS may well be more than adequate all by itself.
Apart from keeping regular gear running when the power's particularly lousy, the only reason to buy a monster line conditioner is to protect very, very expensive gear. There may only be a one in ten thousand chance of it being zapped by its mains supply in any given year, but if that'd cost you $250,000, you might well be very happy to drop a thousand bucks on a power filter.
This is why old crusty ferroresonant power filters keep showing up in second hand markets that're fed by universities and scientific establishments. Those conditioners used to be powering a 1973 minicomputer or a Tunnelling Birefringent Electron Twist Analyser or something. Big businesses used to have a lot of ferroresonant conditioners, too, but even the most awful working environments have by now unloaded their COBOL-era hardware (though not necessarily the software).
Anyway, this means that functional ferroresonant conditioners do show up on the second hand market reasonably often, though they're not necessarily correctly described in the eBay listing.
One big problem with old heavyweight ferroresonant conditioners, though, is that they can consume a lot of power just by themselves.
The big old Sola conditioner I talk about in this piece consumes, all by itself, something in the order of 140 watts, at least according to my cheap power meter. It uses this imposing amount of power to keep itself toasty warm (the top of the casing gets to more than 30 degrees C above ambient) and produce a rather loud hum.
Now, if you're using most of its 1000VA output rating, or if you're using it to protect some super-expensive piece of gear, then the extra power draw is no big deal. But if you're just running a computer from it, you could easily double the average power draw of the computer, not to mention make your office warmer than you necessarily want.
(If you're running an electric heater in the office anyway, though, then there's nothing lost.)
Modern lightweight power conditioners should all consume a lot less power than this. I'm not sure exactly how much less, though, and it can be tricky to measure.
I have, for instance, a couple of 1200VA Total Power Systems line conditioners (another cheap eBay purchase) that are far smaller and lighter than the old Sola and obviously don't waste much power, because they don't get much hotter than ambient even when they're feeding a hundred or so watts of load. (They don't hum, either.) But my cheap power meter reckons they consume 90 watts, with a power factor of 1.
This is because my cheap power meter can only measure power factor in one "direction", the inductive-load direction, and just tops out at 1 (or "100", on the percentage display it uses) when it sees a capacitive load like this conditioner. I suspect that the Total Power Systems conditioner just has a really lousy power factor when it's plugged in with no load, and that multiplies its actual ten-watts-at-most of idle power consumption up to 90W from the cheap meter's point of view.
You can tell that there's something screwy going on with an electricity consumption
meter if you conduct experiments with loads that you
What this means is that when the power conditioner gets itself a reasonable load, its capacitive components stop just flapping back and forth drawing lousy-power-factor "apparent power", and the conditioner actually adds no measurable power consumption to the load.
(You don't, by the way, have to pay for "apparent power"; the electricity meter in your building's breaker box can't even measure it. This is one of the reasons why the magic rip-off boxes, which claim to save you money by improving power factor, are a rip-off. You're not billed by power factor in the first place.)
Neither Brick Wall nor ZeroSurge seem to list power consumption in their product specifications. Neither do APC, for that matter, for their "Line-R" conditioners that're basically just the filtering components from a decent UPS.
SurgeX do, though; they say it's 10 watts for their SA1810, to pick one at random.
If you're only thinking of buying one of the basic Brick Wall units for $250 or whatever, then they look like a reasonably safe buy - but scan around for used lightweight power conditioners within cheap delivery distance of you first. Also bear in mind that if your local power isn't particularly sucky, then a UPS of reasonable quality may be all you need.
Nobody ever got fired for buying APC UPSes, which contain better surge-suppression hardware (well, on average, anyway) than you get in the typical suspiciously-cheap UPS. APC UPSes also, for that matter, show up on eBay all the time, sometimes with good batteries and sometimes without.
(Obviously, if you're buying one with dead batteries, you should try to buy it from someone who's smart enough to remove the dead batteries before they figure out the shipping cost.)
A question I'm sure we've all asked
Quick question about powerful lasers and Crookes radiometers.
How powerful a laser do you need to make a radiometer move? I assume they spin faster as the power goes up. You've obviously tested this from stuff on your site.
I've just got an arrangement of mirrors and radiometers that I think could be really cool if done right.
Adam
Answer:
Practically any laser pointer will turn a radiometer - my old 5mW red pointer did. They
just won't turn them fast. Even a 95mW laser doesn't
kick a radiometer around nearly as fast as even mild direct sunlight, or a humble
incandescent bulb hung over the radiometer.
I'm not sure why this is. It isn't necessarily because the laser lacks power.
The dot from a high-powered green laser can be much brighter per unit area
than direct sunlight, and thus very clearly visible even on a directly-sunlit surface.
But the dot is also very small, while sunlight can illuminate the whole area of each
of the radiometer's little "sails". Work out the numbers, though, and it's clear that
a modern handheld
terrorist tool laser can deliver something like the energy level
of sunlight, for radiometer-turning purposes.
The vanes of my radiometer are about 15mm square; that gives them an area of 2.25 square centimetres. With an unremarkable insolation of 500 watts per square metre, that's 50 milliwatts per square centimetre, or 112.5 milliwatts on an area the size of the single black side of one radiometer vane which is, on average, exposed to the sun.
Sunlight is unlikely to be perpendicular to the sails, of course; the real practical incoming energy per sail-area could easily be only 30mW or so. So you'd think a powerful laser pointer would do a good job of turning a radiometer, even if its official power rating is rather optimistic.
But, in my experience, sunlight or an incandescent lamp work much better than even beefy lasers.
This might have something to do with the laser emitting one tightly defined frequency of light, instead of broadband light with plenty of heat, like an incandescent filament or the sun. Or it could be the angle of incidence of the laser light on the vane, or some other thing I haven't thought of having to do with the unexpectedly complex means of operation of the radiometer.
I think the current Wikipedia writeup has it right, but I'm not sure. How radiometers work is one of those hard-to-pin-down things like income-tax brackets, planes on conveyor belts or that ultimate divider of families and nations, the Ralph Wiggum Viking Question.
If you're hoping to bounce one laser beam through multiple radiometers, you will of course face the problem of the beam being reflected and refracted by the glass bulbs. Still, nothing is impossible for the man with an optics bench, an awful lot of mirrors, lenses and prisms, and spare time.
