Slashdot Mirror
How Do 'Singing Magnets' Work?
dpbsmith asks: "Singing magnets are available at all of the usual geek-toy emporia, and, for all I know, ordinary toy emporia as well. They consist of a pair of magnets made of a polished substance with the general appearance of hematite. What is surprising, pleasing, and unexpected is that when the magnets strike each other under their own power, they produce a sharp, loud buzz that rises in pitch. The sound lasts a good fraction of a second and climbs somewhere into what sounds like the 200-500 Hz range. The exact sound and its duration are somewhat unpredictable and depend on how the magnets happen to strike each other. It is a little like the sound that you get when you mash a pingpong ball against a pingpong table with a paddle. What physics are involved in the production of these sounds?"
"Google searches turn up some forum postings that indicate that it is a synthetic magnetic substance similar to hematite that's available cheaply in China as an industrial byproduct. The singing magnets are a little larger than size of olives; the shape is similar to a (U. S.) football but slightly more elongated. Their major axis is about 5 cm long, their minor about 1 cm. They are fairly powerful and will jump together when placed on a desk about three inches apart. They can distort the colors on a CRT display from a distance of over 20 cm.
Contrary to expectation, the poles of the magnets are oriented along one of the minor axes of the ellipsoid, not the major axis.
Neodymium magnets in 'ordinary' shapes produce boring 'plinks' when they snap together. Something about the shape of these magnets makes the sound much longer-lasting and entertaining. It is not simply the bounding rebound of two objects made of stiff-but-elastic material. Transfers of linear to angular momentum are clearly involved.
If course, I'd love to know whether these things were 'invented' or 'discovered', and by whom, trying to do what.
Contrary to expectation, the poles of the magnets are oriented along one of the minor axes of the ellipsoid, not the major axis.
Neodymium magnets in 'ordinary' shapes produce boring 'plinks' when they snap together. Something about the shape of these magnets makes the sound much longer-lasting and entertaining. It is not simply the bounding rebound of two objects made of stiff-but-elastic material. Transfers of linear to angular momentum are clearly involved.
If course, I'd love to know whether these things were 'invented' or 'discovered', and by whom, trying to do what.
It's also true. I had these when I was a kid, and by pushing them towards eachother, I could make the sound louder, and by muffling them with my hands, I could damp them out. That has to be vibratory acoustics. Plus the sound TOTALLY trips out my son, makes him giggle like you wouldn't believe.
I like to place meaningful quotes in my sig, so people will know that I know what meaningful quotes are.
I have a pair, and this is the only explanation that makes sense to me. I imagine that the effect would also happen on a whole range of rounded magnets of this type, but with the length of the "buzz" varying - the stretched football shape is probably just one of the more effective ones.
If you try to isolate the system by throwing them up in the air so they pull together and strike each other while airborne, they will generally buzz for quite a long time - generally, it stops because the motion was dampened by your hand (or whatever else they land on) rather than coming to a stop on its own.
The whole effect is made even more fun because if you throw them in the air, they will spin around each other like a cat in zero gravity.
Go here: http://liquidmetal.com/news/dsp.multimedia.asp
View their "Ball Bouncer Demonstration"
I have a hunch that these "singing magnets" are coated with a Liquidmetal-like material which preserves some 99% of the impact force and returns it quite energetically, causing the magnets to come together, be pushed apart, come together, be pushed apart, fast enough to create a "tone" of sorts.
When the ball gets to the end of the ball bouncer demonstration, it does largely what the original poster was asking about with regard to the sound it produces.
Reinvent the wheel only at either a lower cost, greater effectiveness, or your own personal enrichment and satisfaction.
For those unaware of what a "singing sword" actually is (and they are real, though not magical) it refers to a sword forged in such a way that the crystalline structure of the metal is highly organized. This wasn't easy to accomplish with low-temperature forges and beaten steel, so such swords were highly prized.
Today we call this metal "stainless steel", and blades made from it are found in just about every kitchen.
"What part of DIGITAL don't people understand?"
In my experience (and I bought one of the earliest iMacs with USB & FireWire, so I've had more than most) the gague of cable does have a significant effect on the operation of some bus-powered peripherals. If the cable run is too long (or the gague too small), series resistance can cause voltage drops that may prevent the device working properly. Bus powered hard drives exhibit the most obvious symptoms (refuse to spin up), but there have been occasions where I've found other devices (Tascam US-224 for example) show unreliable behaviour which has been cured by using more solid cable*.
There's also the issue of grounding; even though the signal down the USB cable is digital, any voltage drop across the ground connection will appear as a signal at the analog input/output of the device connected with respect to mains ground (a condition better known as an "earth loop"). Bad news for unbalanced devices, which cannot operate with "braid breakers" or floating ground.
That being said, power problems can be avoided by simply using decent cable (Belkin is pretty good IMO), and ground problems can be avoided much more cheaply by simply plugging appliances into mains outlets in decending order of power consumption (highest power device closest to mains outlet, a simple form of "star" earthing**). Just avoid using cables the diameter of dental floss, and you should be OK.
I agree that Monster USB/FireWire cables are a pointless waste of money.
Just like Monster speaker cable. If you consider cable impedance as a whole (induction and capacitance as well as resistance), Monster cables are considerably worse than 10 Amp mains flex. The sheer physical size of Monster cable means it has much greater capacitance (a capacitor is two large conductive surfaces seperated by an insulator, which is a pretty good description of Monster cable) than ordinary mains cable. Experiments performed by Ian Hickman published in Electronics and Wireless World (1997, IIRC) showed that Monster's capacitive effects start to be significant at frequencies an order of magnitude lower than frequencies that exhibit the "skin effect", which is the main justification for having such a large conductive area (if 10 A mains cable can deliver 2400 Watts and still need mains filters, it can handle 100 Watts from stereo easily, so clearly low resistance isn't the reason for a large conductor area). These capacitive effects manifest themselves as phase errors (group delay) in high frequencies; so Monster cables do sound "different", but that doesn't automatically equate to "better"***.
*Headphone drivers can consume up to 1 Watt, or 200mA@5V. If a cable has a resistance of 5 ohms (not unreasonable) that translates to a voltage drop of 1V (V=I*R). That's an AC signal, which is then re-injected into the analog circuitry via the power supply. If the power supply rejection ratio (PSRR) of the analog circuitry isn't high enough, this signal will feed through to inputs and outputs, a property the US-224 has in spades.
** OK, this isn't exactly star earthing, but at least any ground voltages are superimposed equally on all devices. The point is to make sure that all devices see the same ground reference voltage.
***"But what about the damping factor", I hear the audiophiles cry? Well, damping factor is measured at the PCB, not at the speaker terminals. When you add the resistance of internal wiring, solder connections and wire oxidation inherent with clamp or screw terminals, your damping factor will be shot to hell long before the cable run to the speakers has any influence. A combination of voltage and current feedback will have a far greater effect on damping factor, and cost less in components than 2 metres of Monster cable.