Actually, you see these things clamped to all sorts of cables on computer equipment. It looks as if the cable (be it a video, mouse or even sometimes power cable) runs through a fat piece of plastic usually clamped an inch or two behind the connector.
In the ‘olden’ days of computers I actually had to add one of these things myself and know there is more to the thing than plastic. It seem as if it is some type of metal. Still, I never figured out what these things were for.
Do they absorb radiation? Something else? They are passive in nature (i.e. they receive no power and do whatever they do just by sitting there). I don’t see why they are useful when they only cover maybe 1% of your cable though.
What gives with these things?
[sub](P.S. Nice to see the old board back all new and shiny!)[/sub]
Ferrite cores
Look closely. Basically what we have is a toroid ferrite core with half a winding of cable. And what is that if not a coil. Admittedly one with very little inductance, but still a coil.
And what are inductors good for? Blocking high frequency!
The little ferrite thingy clamped on the cable is nothing but an LowPass filter, designed to block high frequency noice, of the kind that you might get from the CRT, or just about anything else.
They also block high frequency emission, that might otherwise make it difficult to pass CE/FCC regulations.
Can I get a little more explanation on what is happening here Popup?
Assuming you are correct (and I see no reason to assume otherwise) why does the ferrite core work when clamped to only a tiny bit of your cable? Shouldn’t the cable be emitting EM energy all along its length? Why bother blocking EM interferrence along a tiny fraction of the cable’s length?
The basic premise is that the clamp is a coil with only half a winding.
What we have is basically an inductor in series with the cable. This blocks high frequency signals, that are picked up (or emitted) along the whole length of the cable.
Try hooking up your monitor with a cheap extention cable that does not have the shielding and you will see the difference-- the picture is a little blurry and will have a few mm “doubling” effect that you see in a lot of TV reception.
The electromagnetic waves travel ALONG the cable, going from one end to the other. The ferrite ring is both a wave-blocker and also a wave-absorber.
Often its the CPU bus which spews out nasty electromagnetic garbage. In that case, the ferrite rings should be next to the connectors plugged into the CPU case, to stop the garbage signals from continuing on to the monitor, printer, etc.
With the snap-on ferrite cores, try using an FM radio to listen for electromagnetic garbage near your computer. Snap the cores off the cables and see if the FM interference gets worse.
True, but missleading. What one tries to stop is the high frequency current fluctuation in the cable. The core/choke/clamp adds inductans to the cable, and thus attenuates the HF currents. (Remember: U=L dI/dt for an inductor). Which in turn reduces the RF interference emitted (and received).
Sure, but at the frequencies involved (100s of MHz), you’re looking at antenna physics and propagating voltage waves, not just currents in a wire. At low frequencies only the current would be important, then it wouldn’t matter where you put the inductor. Since yard-long propagating waves are involved (which are made of current and voltage, electrostatic and magnetism), you must put the inductor closest to the source of the waves. In most cases that’s near the CPU. In expensive equipment they actually put ferrite plates inside the multipin connectors on the back panel.
Also, these large ferrite “beads” are not inductors! The ferrite material is intentionally chosen to give strong hysterisis absorbtion at the frequency bands of interest. They act like resistors. True, they are magnetically coupled to the cable. They are like a transformer in series with the cable, but the floating secondary of the “transformer” is connected to a resistor. At lower frequencies this “resistor” vanishes, and the ferrite looks only like an inductor. But it’s the high frequencies (cpu-clock) they’re trying to eliminate.
I did a lot of EMI testing on industrial equipment for a couple of years. I was suprised myself when I found out that ferrite chokes aren’t inductors! I don’t think this fact is widely known.
Another example: those RF anechoic chambers with the big carbon-foam absorber-spikes on the walls? Some companies don’t use spikes. Instead they coat their walls with ferrite plates. The ferrite plates look like resistors, like “black cloth” for RF energy. I think it only works at high frequencies though, so if they’re doing testing below 30MHz, they need the big foam spikes.
Fascinating! The only time I’ve actually measured on these things we were trying to eliminate switching noise in the 100kHz region. Then a single bead had so-and-so many uH (I cannot even remember the order of magnitude). I remember measuring the voltage drop over a single choke, and it was on the order of 100s of mVs peak-to-peak. (The currents involved were enormous.)
I made the very common mistake of scaling outside the validity of the assumptions.
Thanks for pointing me in the right direction!