afinchum

modified 11 years ago

THANKYOU! Inductor driver for wireless power transfer

1387

1d 00:46:19

The transmitter and reciever for the transfer of wireless power, simulated here by means of a loosly coupled transformer. The 100uf cap is a filter, the wave generator is simulating a 555 and the right side has all the stuff for a reciever including output overvoltage protection with the zener diode and resistor. The calculated transmission and recieving coils are pancake coils with 70 turns of 24 AWG sized wire (I.D. =.5 cm, O.D. = 4 cm) @ 10 uH. 

The problem now is the recieving resonating cap, its acting like a pass filter so no current even gets to the bridge. Not what i was expecting..

EDIT: Transformer prob fixed, i didnt realize you need a ground on both sides. Thank you all for the appreciated help!  
EDIT: Well now i really have not a clue on how to properly drive a mosfet. So many learning curves! Thank you to everyone who is taking the time to teach me though! Very much appreciated!

Okay, Now this little project is starting to look like a reality! Thank you to all who have helped me with this little endeavor!

This project is working well in here so its time to try it in reality!
I will post back when I try it.

published 12 years ago

MidMan

12 years ago

You need Ground on the right side too.

adreno

12 years ago

Yep, ground both sides and it should simulate.

afinchum

12 years ago

Oooohhhhh!!! That makes alot more sense! Thank you much! :D

hurz

12 years ago

Next is, mosfets are voltage controled and you need defined dc levels at gate, replace ac source with rectangle and remove cap (dc at gate is undifined) at gate. Next, 1um is a very tiny mosfet for integrted logic circuits. Increase to lets say 200um. Next, 10H transformer will be hard to drive with 30kHz and quite big as air coil in reality. Lower the inductivity to lets say 500uH... next, update the ?resonance? caps at in and output (i just guess they are for resonnance). next, your rectifier is not correct, anyway a full rectifier is not needed as you not drive the transformer in bridge mode. But you can keep it anyway. Depends on orientation you allow to couple a device in your magnetic field.

mocito

12 years ago

You need a diode with cathode connected between the transformer and the resistor and anode connected between the transformer and fet because of the inverse currents generated by the transformer coil too. Put a rectangular wave generator, you need to change the bridge rectifier, it's not correct. And change the transformer like they said above.

afinchum

12 years ago

@hurz: Thank you very much! Your knowledge is greatly appreciated :D duh! My silly rectifier >.< and the mosfet i didnt know where to begin so it was just kinda in there, those are supposedly resonance caps but im not quite sure if they will give the desired effect, if they do produce a clean sinewave then the full wave brige is neccessary (right?), i calculated a small 20 turn 24 awg aircoil to be in the nanohenry range, which probably greatly reduces the amount of energy that can be transfered.. i also calculated a resonace cap for 30 khz to be quite large since the coil is so small and the resonating frequency is relatively low...

afinchum

12 years ago

@mocito: doesnt the resonating cap stop that kickback and if i do that all of that stored energy in the coil gets wasted and cant be used for resonating, right or not? Im still new to this stuff and appreciate all of these comments!

hurz

12 years ago

A few tens of turns on a toilettpaper are probably around >10uH

afinchum

12 years ago

A 5 cm cake at 70 turns gives 10 uH (an I.D. Of .5 cm). what size would you recommend to give a fair amount of power transfer? My ultimate goal is to have the recieving coil the size of a credit card (and of course this will only be for low power applications)

hurz

12 years ago

With 10uH your frequency will be around 1MHz. Xl = 2×pi×f×L and Xl must be low enough to transfer some energy and has to match somehow the impedance for best power.... think about to stop this project because you are on the way to build a radiostation!

afinchum

12 years ago

How do you figure the frequency? I used a resonant calculator to determine C at 30kHz, because i want the operating frequency to be between 25-45khz so you cant hear it and so i do not end up broadcasting ... And on Xl how low would be low enough?(do you know of a website or table that shows certain reactivity strengths?)

hurz

12 years ago

Xl depends on frequency, unfortunately 30kHz is to low to come close to your 50Ohm resistor. xl/(2*pi*L)=50/(6.28*10e-6)= ~800kHz. Go with 5Ohm you are still at 80kHz!

hurz

11 years ago

BTW move the ground to zener anode, this makes it easier to measure the voltage. And increase the mosfet gate driver voltage to about 10V to get a lower RDSon. And 5Ohm is ok, higher is less power and lower is also less power.

hurz

11 years ago

No more progress here?

afinchum

11 years ago

There is, somewhat... I guess i dont understand the reactance formula,is the resistance the total resistance in all of the circuit or is it the inductors dc resistance or something like that, beacust the at50 ohm resistor was just supposed to be a current limiting resistor but as soon as i turned it down to the resistance on my Xl calc it works alot beter

afinchum

11 years ago

Is vto the gate driver voltage? Will i have to increase the timer if i increase that?

hurz

11 years ago

It is a good idea to have R as current limiting. However, if it is to high Z will limit the power. Z is geometrical added by R and Xl. So Xl is givven by inductivity and frequency. Z = sqrt( R^2 + Xl^2)

hurz

11 years ago

For the mosfet Vt is a point were it starts to let current go but it is far away from low RdrainSource. To get a low RDSon you need a much higher voltage delta between Gate and source than just 3.3V

afinchum

11 years ago

Ohh so i need a driver vircuit after the '555' timer then? (and i am Horrible with transistors so ill Reeeeaaallly need help with that driver trans for the mosfet..)

afinchum

11 years ago

Is there another factor/equation that will determine where the sweet spot for Z is?

hurz

11 years ago

forget the driver thing for now. It will be an issue but later on. Keep the EC functiongenerator as driver and increase to voltage as I said.

hurz

11 years ago

cleanup a little bit the circuit. replace the fullway rectifier with a single diode. because it seems you have problems in orientation were is plus and minus ;-) so were to place ground and how to mount zener and LED. 2. whats happend to the mosfet properties? VTO is now 10V, its unrealistic, go back and use e.g. 2V. If i talked about higher voltage then i was talking about the generator you set so far to 3.3V high level.

afinchum

11 years ago

Okay i just replaced that mosfet i had with a fresh one and set the width to what was mentioned earlier

hurz

11 years ago

Zener upside down. And 3.3V from generator is definitive to low. Increase to lets say 10V to reduce RDSon and bring down the voltage over drain and source as close as possible to 0V. 200um width is to small if have come from 1.89 or ~ 2Ohm. Increase to 500um.

afinchum

11 years ago

Okay doke, i changed those things around and instead of having a bleeder diode for overvoltage protection i just made did a regulator circuit instead. Thank you very much for all of your help so far :D

afinchum

11 years ago

I see a potential problem now, the mosfet appears to be just burning through useless energy because the timers pulse width is too wide,but when i change it, the tank circuit starts throwing harmonics and i dont want to start broadcasting...

hurz

11 years ago

The mosfet is driven with rectangle and this produce a mass of harmonics anyway! To not prodcue overtons you have to put much more effort into the generator. Sinewave and linear amplifier. For 30kHz you can probably think about a normal audio amplifier ;-)

hurz

11 years ago

Put a 1mOhm resistor in serie to primary coild and check the current waveform. Its seems to be best if receiver cap is off and transmitter is set to 2.81uF. However, there are still lots of harmonics, and anyway 30kHz is LW radio for submarine ;-)

afinchum

11 years ago

Ohh yes, I forgot about the square wave harmonics

afinchum

11 years ago

okay it looks like when i properly balance the reactance to resonance it doesn't throw harmonics and a stable wave forms that doesn't look like it is wasting power

afinchum

11 years ago

a side not: In this case i would like that 1uf cap on the reciever to be charged with at least 5 volts and it seems like i get a higher voltage when a cap is connected to the reciever coil, it doesnt have to be a resonating cap

hurz

11 years ago

This is because your voltage regulator is cnfigured to low in voltage. Set zener to 2.7V and it is even without 500nF high and a nice sinwave.

hurz

11 years ago

Yes, and i would like to see if you try it now in real. I think this is a good start. There will in practical a lot more upcoming problems but nothing you can't solve, i am sure. You are open minded and it seems to me you love electronics. It was a pleasue to help. Keep it up!

afinchum

11 years ago

Thank you very much! I really couldnt have done it without you (because i wouldve given up XD). I look forward to breadbording this and seeing if anything happens or if my mosfet burns up! XD

afinchum

11 years ago

Ill try this soon, within the week so ill post back then!

afinchum

11 years ago

Good News! I do not need a driver circuit because a 555 timer delivers approximately (Vcc - 1.7 Volts) worth of current :D

hurz

11 years ago

The 555 is propably not the best one for this. What you need is more a PWM generator and 555 can do that but not as good as a simple 2 transistor astabile multi can do. 555 has two external resistors in serie to charge a cap, timeconst (ra+rb)*c and to discharge rb×c, so changing the mosfet ontime always change the frequency in your circuit and the frequency must be stable to have resonance. 555 is more doing a PWM mixed PFM (puls frequency modulation) but give it a try if you have one by hand

afinchum

11 years ago

Ohh true, what if i tie the reset pin to after the mosfet, to keep it aligned? i do have a few other random pwm ic's from salvaged powersupplys but ill have to get their datasheets and see if one of them fits this application better...

afinchum

11 years ago

Okay, update time: I have matched all components nearly perfectly, I am using the W9NK90Z Mosfet from ST(i kept it on the sink it was on out of a switching supply), and my current limiting resistor is 2 Ohm @ 10 watts. I cannot keep it on for more than 20 secs because that limiting resistor gets very hot, but it does its job (no heat from the mosfet observed) :D I used my meter and its running at 30kHz with a width of 12.6% The coil I made is calculated at nearly 10 uH (10.2something) But I feel like that is too small, which is why the resistor is getting so hot, although it doesnt get hot even though it is made of very small magnet wire... I am not good in a situation like this because it exceeds my knowledge... Help is always appreciated! >:D

afinchum

11 years ago

OOOOOHHHH I had a 2 watt on it not the 10W ... Wow.. XD

afinchum

11 years ago

Ohh, no, this one gets hot too...

hurz

11 years ago

And the PWM is fine? Combine 2*2Ohm to 1Ohm and than addition 1Ohm in serie, if you have 4x2Ohm

hurz

11 years ago

if you have a sinwave generator you should first measure the inductivity. try to find the serial resonace with a little cap 10nF around 500kHz

afinchum

11 years ago

Yea the width seems fine, i have the 555 controlled by pots so its not exactly where i want it but ill put it on more precise ones later. The mosfets a voltage controlled device and acts like Capacitor, so wouldnt that mean theres just a very slight current drop on the output of the ic? Also 555's have a (small) driver transistor which i would think to somewhat 'isolate' that output side from the timing side?

afinchum

11 years ago

How should i measure the inductivity? I have a scope a frequency meter, a variable frequency source and a resonance calculator, so do just hook up the tank circuit to the scope, sweep the frequencys slowly, and see where the voltage increases the most, read the frequency, and the capacitance the calculate the inductor value? Ooooorr is there a waaaay simpler way of doing that?

hurz

11 years ago

Yes a powermosfet does has a capacity at gate and it is quite relevat for your driver circuit. Even if mosfet is voltage driven you have to charge and dischage the gate cap with any 30kHz periode. Suppose the gate has about 1nF or even more. tau=r*c to fully recharge you need about 4tau. for 1%pulse worst case width tau is about 300ns. (300ns/4)/1nF = 75Ohm and this is probably to low for a 555.

hurz

11 years ago

For the inductivity, i would put top down 100Ohm 10nF and the coil in serial and measure at resistor the frequency were the lowest swing appears. Use the coil together with cap as notch to find the resonating frequency.

afinchum

11 years ago

In tau=r*c, are those in base units(ohm and farad) or like kohm and micro farad? And what does notch mean? (notch filter?)

hurz

11 years ago

To calculate use all values in Ohm and Farad. e.g. 100Ohm * 10nF = 1e2Ohm * 1e-8Farad = 1e-6 seconds = 1us

afinchum

11 years ago

ooOOoo ¡okay!

infamousjeff

11 years ago

hurz get your circuit made. I was looking at Wikipedia and see that you're design is superior. go kick some asš.

hurz

11 years ago

I would be proude to find our design on wikipedia. Link?

sauzz

11 years ago

This circuit can charge mobile phone if some modifications done in it?

imtyrone

10 years ago

Dude you rock!

afinchum

10 years ago

Im sorry that i fell off the face of the planet guys, school got the best of me! I have just recently had time to start tinkering again and came across multiple circuits for induction heating (i cast aluminum parts too, so many projects so little time!) but in this process i found zero voltage switching supplys and was wondering what it would take to turn that pulse timer into one of those? That way when i actually build this circuit i dont need an external signal source. I would like to thank you again hurz, for the electrical genius!

hurz

9 years ago

Long time no see. I have to thank you, and have fun with electronic!

add comment in editor