Dan Zimmerman N3OX wrote:
>> In the case where they are inductively coupled there is very little
>> radiation, but if they even charge more than a couple inches away it's
>> going to take a substantial signal and it certainly wouldn't be green.
>>
>
>
> Mostly I think they use a pair of coupled mag-loops.
> They can be a couple loop diameters apart and still achieve 50-70%
> efficiency.
>
My headsets use a pair of matching "pancake" coils of 8 to 10 turns
(give or take).
I've never traced it out, but there is very little to the circuit. It
does shut off when the battery is charged.
The headset remains off, until you place it on your head which pulls the
support band tight and then it turns on. It appears that it takes a bit
for the transmitter in the holder to turn on. IOW the green transmit
light doesn't come on as soon as you place the head set on your head.
> You can model the effect in EZNEC. Build a couple identical resonant
> magloops, put a resistive load in one and a source in the other.
>
> Compare the radiated power (average gain test) to the "load loss," which of
> course, in the case of resonant wireless power transfer is the whole useful
> bit ...
>
> http://n3ox.net/files/eznec/wireless_power.ez
>
> It's kind of a neat effect. Just near field coupling that you could build
> at home at the basic level. I assume that the actual devices will be fairly
> sophisticated... measuring impedance of the primary circuit with very low
> power levels and then cranking it up when a load is detected (look at the
> source impedance in EZNEC with and without a load.... almost purely
> inductive without, very resistive with... that's part of the key to the
> method).
>
> Of course, if you tightly couple them (fraction of a loop diameter) you can
> get essentially 100% efficiency. This isn't actually true for a pair of
> single wire loops stacked right on top of each other without resonating
> them. If you try to build a non-resonant transformer that way (that is,
> without a core) it doesn't work very well. Too much leakage.
>
I would think, if the pancake coils are resonant and at a high enough
frequency they would work very much like a transformer without needing
an iron core to increase flux density.
73
Roger (K8RI)
> If you very tightly couple them, you get great power transfer and they
> really can be considered non-radiative. The back EMF on the load loop
> induces a current 180 degrees out of phase with the current in the source
> loop, and from a radiation standpoint you have a parasitic phased array very
> close spaced of antennas with 1% radiating efficiencies anyway, so it ends
> up really low gain.
>
> I built a pair of resonant coupling rings to transfer ultrasound signals on
> and off my rotating Ph.D. experiment:
>
> www.n3ox.net/files/us_ring.jpg
>
> The EZNEC file of that one shows 99+% of the power going to the load and an
> "average gain" too low to list (<-99.99dBi)
>
> You move 'em apart by a couple loop diameters and they end up more like
> -40dBi and 70% efficient if I recall correctly.
>
> That could still be a RFI generator, but I'm hoping that the fact that *very
> sharp resonance* is the entire key to the operation will limit the RFI
> potential to hams.
>
> I get kind of a chuckle out of the hype.
>
> I've been using this sort of thing to couple ultrasound signals to rotating
> equipment for a few years now, ever since I got the idea from the Tenth
> Edition ARRL Antenna book (circa 1964), Figure 3-74, which shows how to use
> large inductive loops to both match to and couple open wire line to your
> rotary beam antenna ;-)
>
> I still haven't tried to light up a lightbulb on the other side of the shack
> but I could, and you could too.
>
>
> 73
> Dan
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>
>
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|