Wireless power transfer
Tom540, Tue Sept 18 2007, 04:45PMOkay recently A friend asked me to help him with a wireless power transfer project. It Is supposed to charge some small robots that are encased in a plastic box the top is covered so not much light gets in. Naturally I looked to Tesla coils.
What I've been trying is basically a DRSSTC without the secondary and for the receiver (schematic attached) I am using a ferrite tube inductor and a cap in parallel with some Schottky's to rectify it's output. This works okay but the FET I'm using get hot due to power in the primary being absorbed back into FET's.
I am able to charge a 2200uF cap in about 3 seconds from ~2.5' away. The efficiency pretty much sucks, especially with the heating etc.
Anyone know a better way or have any ideas on how to make this a more feasible plan? I was thinking maybe a lower q winding and smaller cap. Haven't had the chance to try that yet.
]1190133893_487_FT0_rx.pdf[/file]
Re: Wireless power transfer
Sulaiman, Tue Sept 18 2007, 04:58PM
The transmission of wireless power is such an old topic that I'm sure google will give answers,
what I'm more concerned with is perception.
If people think that 50/60 Hz powerlines and/or cellphones are a health risk
what will they (?) make of your pulsed-rf power transmitter?
What will the FCC position be? Especially since your transmissions are deliberate.
In this case I think that you should research the legal side before the technical
Sulaiman, Tue Sept 18 2007, 04:58PM
The transmission of wireless power is such an old topic that I'm sure google will give answers,
what I'm more concerned with is perception.
If people think that 50/60 Hz powerlines and/or cellphones are a health risk
what will they (?) make of your pulsed-rf power transmitter?
What will the FCC position be? Especially since your transmissions are deliberate.
In this case I think that you should research the legal side before the technical
Re: Wireless power transfer
Marko, Tue Sept 18 2007, 05:41PM
Hi Tom
First thing is, you don't need pulsed inverter for this. And it's actually less efficient.
For magnetic coupling, your ideal circuit is parallel LC. It presents high impedance to driver and low impedance to receiver and is thus excellent for transmission of such kind.
My personal weapon of choice was current fed royer oscillator, because it achieves ZVS on switches and allows use into insane frequencies.
Receiver is simple air-cored coil with resonant tank cap tuned to frequency of the inverter.
At 1Mhz I ran over 50% efficiency while pumping about 10W into two incandescent bulbs over about 7cm. This included about 5W dissipated in gate pullup resistors
I'm confused by the ferrite core you used - as I think you need no core at all in this application. How did you wind the winding onto a ferrite tube? Going through it toroid-style winding will do nothing, you would need to float the ends like a ferrite rod. And I don't know if it would be of much help.
Also take care not to overvolt diodes in your receiver when it's not loaded.
If you must go off it might be a good idea to include a relay to disconnect the coil when not needed.
Voltage will be unstable but you can use Cuk or SEPIC converter to stabilize it.
What distance and power are you aiming?
The circuit of this size will emit tiny amounts of longwave radiation. So if it's tuned to legal frequency and properly filtered I think it wouldn't have problems even with FCC.
BTW, what do all our SGTC's do?
Marko, Tue Sept 18 2007, 05:41PM
Hi Tom
First thing is, you don't need pulsed inverter for this. And it's actually less efficient.
For magnetic coupling, your ideal circuit is parallel LC. It presents high impedance to driver and low impedance to receiver and is thus excellent for transmission of such kind.
My personal weapon of choice was current fed royer oscillator, because it achieves ZVS on switches and allows use into insane frequencies.
Receiver is simple air-cored coil with resonant tank cap tuned to frequency of the inverter.
At 1Mhz I ran over 50% efficiency while pumping about 10W into two incandescent bulbs over about 7cm. This included about 5W dissipated in gate pullup resistors
I'm confused by the ferrite core you used - as I think you need no core at all in this application. How did you wind the winding onto a ferrite tube? Going through it toroid-style winding will do nothing, you would need to float the ends like a ferrite rod. And I don't know if it would be of much help.
Also take care not to overvolt diodes in your receiver when it's not loaded.
If you must go off it might be a good idea to include a relay to disconnect the coil when not needed.
Voltage will be unstable but you can use Cuk or SEPIC converter to stabilize it.
What distance and power are you aiming?
What will the FCC position be? Especially since your transmissions are deliberate.
The circuit of this size will emit tiny amounts of longwave radiation. So if it's tuned to legal frequency and properly filtered I think it wouldn't have problems even with FCC.
BTW, what do all our SGTC's do?
Re: Wireless power transfer
Tom540, Tue Sept 18 2007, 06:14PM
Hi guys,
This isn't going to be a product or anything just a display and I am hoping to get maybe 6' distance with at least 50% efficiency if possible. 5 or 10 Watts on the Rx side would be great. If it's efficient enough I might want it to run continuously to actually power the bots or maybe only at night to charge them. Would a Royer circuit run a lot cooler than what I have?
So are you saying the TX and RX should both be parallel LC circuits? I tried rearranging some things and it seemed like havin the TX LC series and the RX LC parallel got the best results. Of course that was with the inverter. The frequency I am testing with this is only 130KHz.
It's a long skinny ferrite tube 2.5" long. I wound some turns around the outside of the core not through the center like in gate drive transformers. I'll post some pics tonight when I get home. The receiver coils need to be kinda small less than 3" diameter or so.
As far as FCC goes it will be in Europe and the frequency is low. it shouldn't be a problem.
-Tom
Tom540, Tue Sept 18 2007, 06:14PM
Hi guys,
This isn't going to be a product or anything just a display and I am hoping to get maybe 6' distance with at least 50% efficiency if possible. 5 or 10 Watts on the Rx side would be great. If it's efficient enough I might want it to run continuously to actually power the bots or maybe only at night to charge them. Would a Royer circuit run a lot cooler than what I have?
For magnetic coupling, your ideal circuit is parallel LC. It presents high impedance to driver and low impedance to receiver and is thus excellent for transmission of such kind.
So are you saying the TX and RX should both be parallel LC circuits? I tried rearranging some things and it seemed like havin the TX LC series and the RX LC parallel got the best results. Of course that was with the inverter. The frequency I am testing with this is only 130KHz.
I'm confused by the ferrite core you used - as I think you need no core at all in this application. How did you wind the winding onto a ferrite tube? Going through it toroid-style winding will do nothing, you would need to float the ends like a ferrite rod. And I don't know if it would be of much help.
It's a long skinny ferrite tube 2.5" long. I wound some turns around the outside of the core not through the center like in gate drive transformers. I'll post some pics tonight when I get home. The receiver coils need to be kinda small less than 3" diameter or so.
As far as FCC goes it will be in Europe and the frequency is low. it shouldn't be a problem.
-Tom
Re: Wireless power transfer
Marko, Tue Sept 18 2007, 08:02PM
Good luck about this, you'll need godlike resonators to achieve such efficiency. You'l need enormous Q's and very high drive voltage. I think 6 inches would be much more realistic.
It's simply because they configure the impedances nicest way. You could use series too, but then your inverter would need to supply all the tank current, and would be unable to ZVS.
Parallel LC circuit needs to be current-fed, either via DC link inductor or series inductor on a bridge (LCLR). LCLR is a ZCS-only design but allows higher operating voltage, so it's your pick. I haven't tried anything as such yet.
Regarding royer, it's still in heavy development from my side, so I don't know how much can I help. I want to drive the gates negative and test it with bigger variety of higher-Q resonators.
Marko, Tue Sept 18 2007, 08:02PM
This isn't going to be a product or anything just a display and I am hoping to get maybe 6' distance with at least 50% efficiency if possible.
Good luck about this, you'll need godlike resonators to achieve such efficiency. You'l need enormous Q's and very high drive voltage. I think 6 inches would be much more realistic.
So are you saying the TX and RX should both be parallel LC circuits? I tried rearranging some things and it seemed like havin the TX LC series and the RX LC parallel got the best results. Of course that was with the inverter. The frequency I am testing with this is only 130KHz.
It's simply because they configure the impedances nicest way. You could use series too, but then your inverter would need to supply all the tank current, and would be unable to ZVS.
Parallel LC circuit needs to be current-fed, either via DC link inductor or series inductor on a bridge (LCLR). LCLR is a ZCS-only design but allows higher operating voltage, so it's your pick. I haven't tried anything as such yet.
Regarding royer, it's still in heavy development from my side, so I don't know how much can I help. I want to drive the gates negative and test it with bigger variety of higher-Q resonators.
Re: Wireless power transfer
Sulaiman, Tue Sept 18 2007, 10:02PM
I'm with Marko on the 'royer' approach,
I would try to make the primary of copper tubing and as large as feasible,
because you're dealing with near-field coupling.
Example; 2 turns of 6mm (or more) copper tubing, 12" diameter
polypropylene film/foil capacitors,
12 V lead-acid battery or power supply.
The lower the frequency the better because you want a very large current circulating in the primary
so you're looking at more than 1 uF of high-current capacitance.
Basically it's amp-turns in the primary that you want to maximize.
The 'transmitting' resonant circuit needs to be as high 'Q' as practical for efficiency
the receiving circuit doesn't need to be high 'Q' or even resonant.
Keep iron/steel far away as it will absorb a lot of energy.
I'm waiting for some high current mosfets (SUM110N05) for exactly such a circuit
Vishay say I should have them by Oct. 2nd - so I'll know more then!
Sulaiman, Tue Sept 18 2007, 10:02PM
I'm with Marko on the 'royer' approach,
I would try to make the primary of copper tubing and as large as feasible,
because you're dealing with near-field coupling.
Example; 2 turns of 6mm (or more) copper tubing, 12" diameter
polypropylene film/foil capacitors,
12 V lead-acid battery or power supply.
The lower the frequency the better because you want a very large current circulating in the primary
so you're looking at more than 1 uF of high-current capacitance.
Basically it's amp-turns in the primary that you want to maximize.
The 'transmitting' resonant circuit needs to be as high 'Q' as practical for efficiency
the receiving circuit doesn't need to be high 'Q' or even resonant.
Keep iron/steel far away as it will absorb a lot of energy.
I'm waiting for some high current mosfets (SUM110N05) for exactly such a circuit
Vishay say I should have them by Oct. 2nd - so I'll know more then!
Re: Wireless power transfer
Tom540, Wed Sept 19 2007, 01:01AM
Part of the problem I had was rectifying what I got on the Rx side. I had a bridge rectifier made from a ton of schottky in series since they were only 40volt. It seemed to work okay but needs some high voltage fast diodes.
I don't know about the royer thing. Id rather just use an H-bridge since I already have pre made boards for those.
My transmitter was running from 60 volts off a bench supply. Tank circuit was in series just like the primary of a DRSSTC and the primary winding was 6 inches diameter. I was getting about 350vpp pulsed on the Rx since I had to use an interrupter because my fets were getting very hot. I think I will try a lower tank cap this weekend with much wider, more inductive coil with copper tube like Sulaiman said and see how that works. I'll run it off the wall with a voltage doubler and current limit it. I'm not happy with this thing needing and interrupter and huge fans.
Tom540, Wed Sept 19 2007, 01:01AM
Part of the problem I had was rectifying what I got on the Rx side. I had a bridge rectifier made from a ton of schottky in series since they were only 40volt. It seemed to work okay but needs some high voltage fast diodes.
I don't know about the royer thing. Id rather just use an H-bridge since I already have pre made boards for those.
My transmitter was running from 60 volts off a bench supply. Tank circuit was in series just like the primary of a DRSSTC and the primary winding was 6 inches diameter. I was getting about 350vpp pulsed on the Rx since I had to use an interrupter because my fets were getting very hot. I think I will try a lower tank cap this weekend with much wider, more inductive coil with copper tube like Sulaiman said and see how that works. I'll run it off the wall with a voltage doubler and current limit it. I'm not happy with this thing needing and interrupter and huge fans.
Re: Wireless power transfer
Sulaiman, Wed Sept 19 2007, 01:32AM
If you have a series-resonant Inductor-capacitor driven by a voltage source at it's resonant frequency
when there is no load the current will be Ipri = Vpri/Zpri
and since at resonance Zpri is VERY small (series resistance of the inductor basically)
the current will be enormous/huge/massive/excessive.
If you're going to use a series-resonant system then use a LOW voltage and run CW (continuous)
I still think a parallel-resonant 'royer' (current-fed parallel-resonant) is your best choice
Sulaiman, Wed Sept 19 2007, 01:32AM
If you have a series-resonant Inductor-capacitor driven by a voltage source at it's resonant frequency
when there is no load the current will be Ipri = Vpri/Zpri
and since at resonance Zpri is VERY small (series resistance of the inductor basically)
the current will be enormous/huge/massive/excessive.
If you're going to use a series-resonant system then use a LOW voltage and run CW (continuous)
I still think a parallel-resonant 'royer' (current-fed parallel-resonant) is your best choice
Re: Wireless power transfer
Marko, Wed Sept 19 2007, 10:59AM
Really not too low because total transmissible power will be proportional to frequency.
I use rather small caps, in order of few nF. Too high is still not too good either since you run into practical problems like gate drive. Low Mhz seemed to be sweet spot.
I actually had an idea to use really high-inductance resonator and a small few-hundred pf variable cap to tune it.
For these frequencies copper tube may also be somewhat too thick; you'l have better ''copper efficiency'' with foil or litz wires.
PCB may be useful in some cases.
You can wrap up aluminium or copper foil strip with Teflon or PE tape to make a really robust coil. It would have some self-capacitance, and if big enough it may not even need external tank cap at all.
You could also use ferrite, depending on wanted directionality.
Marko, Wed Sept 19 2007, 10:59AM
The lower the frequency the better because you want a very large current circulating in the primary
so you're looking at more than 1 uF of high-current capacitance.
Really not too low because total transmissible power will be proportional to frequency.
I use rather small caps, in order of few nF. Too high is still not too good either since you run into practical problems like gate drive. Low Mhz seemed to be sweet spot.
I actually had an idea to use really high-inductance resonator and a small few-hundred pf variable cap to tune it.
I would try to make the primary of copper tubing and as large as feasible,
because you're dealing with near-field coupling.
For these frequencies copper tube may also be somewhat too thick; you'l have better ''copper efficiency'' with foil or litz wires.
PCB may be useful in some cases.
You can wrap up aluminium or copper foil strip with Teflon or PE tape to make a really robust coil. It would have some self-capacitance, and if big enough it may not even need external tank cap at all.
You could also use ferrite, depending on wanted directionality.
Re: Wireless power transfer
Tom540, Wed Sept 19 2007, 05:14PM
I think I'll try that. Finally, an excuse to grab one of those cool tube variable caps.
Tom540, Wed Sept 19 2007, 05:14PM
I think I'll try that. Finally, an excuse to grab one of those cool tube variable caps.
Re: Wireless power transfer
Slava, Sat Jan 12 2008, 07:03PM
your SSTC primary method is the best bet, that is what MIT used to achieve record distance power transfer to a light bulb. Both sides of the circuit should be tuned though and should take into account the load in order to work well over a distance.
Slava, Sat Jan 12 2008, 07:03PM
your SSTC primary method is the best bet, that is what MIT used to achieve record distance power transfer to a light bulb. Both sides of the circuit should be tuned though and should take into account the load in order to work well over a distance.
Print this page