# Oscillator / Power Amp circuit voltage swing problem

1. Oct 28, 2012

### oso0690

Hey guys,

I'm designing a wireless cell phone charger that can be used inside a car for the senior design project. I'm feeding 12V DC into a colpitts oscillator which is fed into a B class power amplifier which will then be fed into the transmitting coil. Each one works great independently but when I connect them together, the voltage swing is off. It goes from about +4V to -8V and I would prefer it to be +6V to -6V.

Also, the peak voltages cut off... I'm guessing from the fact that the input voltage is the same DC input that the transistors run off. What is wrong with the circuit?

I attached the schematic as well as the output (and a zoomed in output).

Q1 is a general purpose BJT (2N2222)

Thanks for the help!

#### Attached Files:

File size:
31.2 KB
Views:
197
• ###### Cut Off.png
File size:
8.1 KB
Views:
120
2. Oct 28, 2012

### AlephZero

3. Oct 28, 2012

### oso0690

You're right. I didn't even see that. I changed it but it gives the exact same output as before.

4. Oct 28, 2012

### AlephZero

Try working backwards to see where it goes unsymmetrical - i.e. see what voltage you have at points 1 2 and 3.

Also, since your plot looks "ragged", try to plot enough points so you can see the actual waveforms. It could just be an artefact of how you plotted the output.

5. Oct 29, 2012

### sophiecentaur

It seems that you haven't actually applied any bias voltage to the bases of your complementary pair of transistors. Their operating point will just be defined by some sort of 'mutual agreement between them' and could be anywhere.

6. Nov 3, 2012

### oso0690

Thanks guys. I changed/added some things in the circuit that fixed the problem. Another problem was that the two oscillating capacitors were reversed. I ran into another problem though.

Here's what I'm doing:
Oscillator -> Power amp -> Transmitting coil -> Receiving coil -> Rectifier -> Cell Phone Battery

I'm trying to get as much power transferred from the transmitting inductor (L2) to the receiving inductor (L3). The power at the transmitting side is very high (seems unrealistically high) but the receiving side is much smaller. I tried matching the impedances through trial and error with the capacitor values (C5 and C7) but I feel like it should be better (k = 0.9!) Do you guys notice any immediate problems with the circuit?

File size:
54 KB
Views:
163
7. Nov 3, 2012

### sophiecentaur

The details of the coupling between your two coils is vitally important. A simulation can't help you there - you're into RF transformer design if you want a clue about that. It's going to depend on separation and physical size of the coupled coils.

BTW, what do you mean by "power at the transmitting side"? Are you referring to V.I or just VI? If V and I are not in phase there will not be as much power as you think. How are you measuring the variables?

8. Nov 3, 2012

### vk6kro

The secondary coil L3 has a capacitor across it which tunes it to about 21.6 KHz. Since the drive is at 166 KHz, this capacitor should be about 7.6 nF.

Also, the two transistors Q3 and Q4 should have about 1.8 volts DC between them because there are 3 diode drops in the base emitter junctions of Q3, Q2 and Q4. An easy way to do this is to have 3 silicon diodes in series between the bases of Q3 and Q4.

Last edited: Nov 3, 2012
9. Nov 4, 2012

### oso0690

@sophiecentaur - Okay, I assumed that the mutual and self inductances were all that was needed to describe the transformer.

I'm not sure what the differences are between V.I and VI? I was looking at Vrms and Irms on the transmitting and receiving coils and Vdc and Idc at the output.

Do you suggest I test the physical circuit to get a feel for what's going on in the transformer?

@vk6kro - That capacitor was the first thing I did but I saw very negligible voltage and current so I thought wo = 1 / sqrt(LC) wasn't useful. I played around with the capacitors and the ones in the schematic gave the best results.

10. Nov 4, 2012

### Enthalpy

O yes!

A few random comments (and I'll try not to rant at simulators):
- You lack capacitors at the power supply.
- 40nF and 450nH don't exist.
- You should check what precision you need at the LC circuits and at the oscillator frequency.
- Did you check if you're allowed to radiate at this frequency, with what frequency tolerance (the LC oscillator may be too imprecise), what power or field? How strong can the harmonics be? You need a better filter probably, with several stages.
- At 100kHz and with coils, don't use a circuit for audio amplifiers. Use a class C amplifier circuit for radio, with a single transistor, possibly a MOS.
- 12V and several Amps make a low output impedance. It should work at 166kHz but RF designers would take a higher supply voltage.
- Your Darlingtons lack a base-emitter path to discharge the base. I expect it's necessary at 100kHz.
- If you use BD139/140 at the second stage of a Darlington, the first stage should be smaller.
- Are 1N914 strong enough? You don't need a full bridge anyway: one diode is enough and drops less voltage, possibly two (with a mid-tap at the coil) to reduce H2 stray radiation. A Schottky would drop less.
- 120µH don't resonate at 166kHz with 40nF. Is that desired? I doubt.
- 120µH are shorted at 166kHz by 450nF. This can't be desired.
- 100µF is huge for 100kHz and induces big current peaks in the diodes and the coil.

My general advice is the same: give up Spice for a while, build the coils and measure them, including their losses and coupling. Coupling tends to be small between two loose air coils, that's the only reason why we resonate them both.

11. Nov 4, 2012

### oso0690

Wow thanks a lot guys. After I change things up and test the physical system, I'll let you know how it goes! As well as possibly more questions :)

12. Nov 5, 2012

### sophiecentaur

I was wondering how you were going to find the Mutual Impedance accurately. That could be a bit uncertain as performance will be very dependent on the actual construction - so building the circuit would mean that you can check on actual performance.

My V.I and VI comment was to draw the distinction between what you get when you multiply magnitudes of V and I and when you multiply then with regard to phase, which is a true measure of power. (It's precisely the same thing with AC mains circuits, where Power Factor is very relevant). But until you have a defined load, where is the power actually going? Once you get the secondary circuit resonating, you will be able to tweak the values to get a maximum of DC from your rectifier. Only then will you really have an idea of the useful power that is being transferred.

13. Nov 5, 2012

### oso0690

My professor gave me a way of finding the k value using the coils and an inductor analyzer. Using the same distance we'll be using in our final design, I measured the k value by first finding the value of the total inductance seen with the secondary coils shorted. He defined this as Ls. Then he found Ls = L1 - (M^2/L2), then k = sqrt(1 - (Ls/L1)). I haven't finalized the coils yet because I want to step up or step down the voltage as much as necessary at the final output to get around 5V (instead of having to design a voltage regulator).

Okay I understand what you mean about power. I'm not sure what load to use before attaching the coils. I've been using a 10 Ohm resistor for now since that's what I have in the toolbox. I figured that when I get the simulation to nearly match the physical circuit it will be good to proceed to the next stage (coils).

14. Nov 5, 2012

### Staff: Mentor

You didn't answer Enthalpy's question yet (that I see) about whether it is legally okay for you to put out RF power at your target frequency, and you didn't address his question about the spectral purity of your transmitted signal.

I suspect you are probably okay legally down at that low frequency, but I'm not sure. What frequency is used for ELF communication with submarines, for example? And your "antenna" is not very efficient at launching an EM wave at that low frequency, but still, legal (FCC) considerations should be part of the write-up of the report for your project. It would be incomplete without it.

Also, you should include in your project report a section on the human safety considerations of such a wireless charger. I believe that some wireless charging arrangements have had some health related safety issues, but I only have a hazy memory of that. A little searching should find you reliable sources on the allowed RF power levels versus frequency for near-human use.

15. Nov 6, 2012

### oso0690

Oops, missed that question. Thanks for directing to it, I overlooked this part of the project.

My only question on this is what is defined as the "final radio frequency stage?" The input to the transmitter?

I'm not sure about the spectral purity. My professor said that when the receiver is tuned (with the capacitor) it gets rid of most of the harmonics. I'll have to make sure that they are insignificant, but I don't want to do that until the receiver side of the circuit is completed.

I picked a lower frequency because it's easy/cheap to do with a breadboard. Going into MHz and I'd imagine the parasitic effects become more pronounced. Our coils are so close together they basically touch so I'm hoping there are no worries with that.

I thought submarines use a very very low frequency (in the 100s of Hertz?)

Health issues! I overlooked those too. I remember reading a similar project confirming no health issues but our project is not exact. It would be poor engineering to not research that.

16. Nov 6, 2012

### oso0690

I finally solved the main problem I was having with connecting the oscillator to the power amp. I needed a buffer amplifier between them as well as drop the biasing diodes in the power amp. My professor said that crossover distortion is nearly eliminated with a tuned circuit. Attached is the schematic and its output.

EDIT: Now that I'm thinking into the <1W constraint, how can I possibly deliver enough current to the cell phone? It seems that 1W itself cannot charge a cell phone battery (3.7V Li ion, 950mAh) very well let alone needing the signal to travel to the receiver AND go through the rectifier. What do you guys think? Should I find another frequency range that is allowed more watts? Am I not understanding FCC Regulations letter a?

#### Attached Files:

File size:
42.6 KB
Views:
110
• ###### Osc-PAmp Output.jpg
File size:
18.5 KB
Views:
98
Last edited: Nov 6, 2012