Fun with wireless power

[Lightning On Demand]

Hi All,

Here's a short video of the new 'Tesla Roadster,' a small vehicle powered completely by ambient fields as per Tesla's original patent.

A solid-state 'sending' Tesla Coil produces an ambient electric field, which a small 'receiving' Tesla Coil on the roadster harvests, using the aerial above the driver.

/>
There's some wireless history at the beginning, test starts at 2:22.


Cheers,
Greg Leyh

[Max]

Love the sparks on the chain! What‘s the efficiency of the setup? How much input power do you need for how much power on the wheels?

Kind regards,
Max

[Lightning On Demand]

Hi Max,

The transfer efficiency varies widely with range, about 10% on average during this test.  The overall efficiency would be higher with more vehicles operating at the same time.  I didn't have metering on the roadster during this test, but I think the motor was consuming about 200-300 watts.

[Mads Barnkob]

Hi LoD and welcome to HVF!

Will there be an updated paper on the wireless transmission, as the one you have on your page is from 2008? http://lod.org/misc/Leyh/Papers/NAPS2008Final.pdf

[Lightning On Demand]

Hi Mads,

Once I'm able to set up some outdoor tests with multiple vehicles I might have more some data that would interest folks.

I need to add some better metrology to the equipment to collect data.  So far I've been just doing sims, followed by scope msmts on the bench.  Wonder what kinds of affordable battery-operated data logging gadgets are available now?

The sims yield some interesting interactions.  On a wireless racetrack, it appears that a car with a phase/freq agile receiver could steal power from neighboring cars. 

[acmq]

I imagine that if the system is optimized to maximum power transfer with one receiver at a certain distance, if more receivers are added at the same distance the same maximum power will be split among the receivers, that would then receive a smaller fraction of the transmitted power. A consequence of the maximum power transfer. Some of the power that was being lost can be captured, but not much.

[T3sl4co1l]

Cute!

Interesting that the driver goes in circles, never approaching or getting too far away from the primary coil...

Yup, the coupling between resonators is all over the place, so you get double peaking, or N in general, which reduces amplitude at a given frequency.  To some extent this can be compensated by regulating the resonator voltage.

Resonator adjustment is something already done today with mains power, just to a much smaller extent: as load varies, the voltage drop and phase shift due to a long transmission line varies, and has to be compensated properly with reactors.  The reactive power is less than, say, 10% the total load, so the Q is very low in this case.  A high Q resonator means that it's just very much more sensitive to changes in load demand and environment (number and size of loads, proximity of anything).

Even with very large towers, you'd need them spaced evenly along e.g. a highway; you get the same problem as the various infamous Solar Roads projects, that you're spending billions on just a tiny bit of infrastructure (some miles of roads?), but it's costing you billions more in operating costs, and only earning millions in actual useful effort (in this case, not even power generated, only transferred, with pitiful efficiency at that).

And even with defensively smart loads, I don't think you can prevent someone from "hot rodding" their resonator, allowing abuse both in terms of excessive power consumption, and denial of service to other users.  Something that, again, has parallels in low frequency mains, just not nearly as bad: a heavy or shorted load will brown out the local branch, but only by so much, and its fuse will clear quickly.  Crank up the load sensitivity, and remove the ability to disconnect offenders, and what will you get?

Tim

[Steve Ward]

Nice video and demo.  Good to see you here, Greg!

I have to wonder how much more power you might receive if one of those sparks grabbed a hold of your aerial :-).

Maybe you can get crowd-funded to build a bumper car arena with intermittent lightning on demand!

On a more serious note... One question that comes to mind is whether it makes sense to operate as an "interrupted" or "pulsing" type coil, or go for lower peak fields with a continuous-wave type drive?  I haven't thought much about how efficiency changes with field strength. But maybe there's some reason the efficiency goes up with more voltage, so higher pulsed levels make more sense.  The rectification losses are non-linear, working in favor of pulsing i think, but all the IIR losses would rather go for a CW setup.  At least that's my take on it.

[Lightning On Demand]

Hi Antonio,

I think that's true for the unusual case where one receiver is coupling most of the power, say >50%.  This roadster uses about 5-15% of the available power.  If another roadster were added, I think the power delivered to the first one wouldn't drop much.  More power is coupled from the sending coil now, since the effective area of all the receive aerials has doubled, and less power is delivered to the room losses.   -Greg

I imagine that if the system is optimized to maximum power transfer with one receiver at a certain distance, if more receivers are added at the same distance the same maximum power will be split among the receivers, that would then receive a smaller fraction of the transmitted power. A consequence of the maximum power transfer. Some of the power that was being lost can be captured, but not much.

[Lightning On Demand]

Hi Steve,

I'd imagine the power would drop to near zero during a streamer connect, either because the safety gap shunts out the coil, or... because faults through a non-safety'd coil would shunt it as well!

And yeah, I think that CW operation would yield best RMS results, since breakout on the sending coil [or the receive aerial] is a limiting factor.  -Greg

Nice video and demo.  Good to see you here, Greg!

I have to wonder how much more power you might receive if one of those sparks grabbed a hold of your aerial :-).

Maybe you can get crowd-funded to build a bumper car arena with intermittent lightning on demand!

On a more serious note... One question that comes to mind is whether it makes sense to operate as an "interrupted" or "pulsing" type coil, or go for lower peak fields with a continuous-wave type drive?  I haven't thought much about how efficiency changes with field strength. But maybe there's some reason the efficiency goes up with more voltage, so higher pulsed levels make more sense.  The rectification losses are non-linear, working in favor of pulsing i think, but all the IIR losses would rather go for a CW setup.  At least that's my take on it.

[Lightning On Demand]

Cute!

Interesting that the driver goes in circles, never approaching or getting too far away from the primary coil...

Yup, the coupling between resonators is all over the place, so you get double peaking, or N in general, which reduces amplitude at a given frequency.  To some extent this can be compensated by regulating the resonator voltage.

Resonator adjustment is something already done today with mains power, just to a much smaller extent: as load varies, the voltage drop and phase shift due to a long transmission line varies, and has to be compensated properly with reactors.  The reactive power is less than, say, 10% the total load, so the Q is very low in this case.  A high Q resonator means that it's just very much more sensitive to changes in load demand and environment (number and size of loads, proximity of anything).

Even with very large towers, you'd need them spaced evenly along e.g. a highway; you get the same problem as the various infamous Solar Roads projects, that you're spending billions on just a tiny bit of infrastructure (some miles of roads?), but it's costing you billions more in operating costs, and only earning millions in actual useful effort (in this case, not even power generated, only transferred, with pitiful efficiency at that).

And even with defensively smart loads, I don't think you can prevent someone from "hot rodding" their resonator, allowing abuse both in terms of excessive power consumption, and denial of service to other users.  Something that, again, has parallels in low frequency mains, just not nearly as bad: a heavy or shorted load will brown out the local branch, but only by so much, and its fuse will clear quickly.  Crank up the load sensitivity, and remove the ability to disconnect offenders, and what will you get?

Tim

An interesting and useful technology escalation, one might hope.