Choosing opamp as small power supply with 1 Mhz signal inpt

[naveenlp]

Hi,

I am currently doing a undergraduate project in my university and need some help rgd my circuit design. I am trying to experiment with resonant inductive transfer to wirelessly transmit power.

The setup is:
I have a load which is an LC circuit with a capacitor connected in parallel to a copper coil. I need to drive power into this LC circuit using a 1mhz signal at around 6.8V p-p so that the power delivered to the coil is around 100mW. This gives a current requirement of around 80mA p-p if I am not wrong. This amt of power can approximately light up 3 LEDs connected in series.

I am using a Colpitts Oscillator (CO) to generate a signal at 1mhz with 3.8V p-p. This can possibly be increased to 6.8V p-p. Now I cannot connect the CO output directly to the coils because one, the BJT I use cannot supply high currents and two, the oscillation will be affected because of the reactive loads (which might affect the LC filter in the CO) So instead, I am trying to connect the CO output to an inverting opamp (or alternatively a class AB amplifier, but I cannot find power MOSFETs that respond at high frequencies) so that the CO output looks into a high impedance gate/base of transistor. Also, the current reqd can be supplied by opamp instead of the CO

I initially used a LM318 opamp (available in lab). This gave a highly distorted output for 1mhz. But it did drive some power into the coils. Estimate at around 10-20 mW as LED lit up brightly.

I thought if I chose an opamp with better specs than LM318, I should get my results. I chose a high output current opamp, OPA561, which had 17 mhz GBWP and 50V/us slew rate, 50-70mA quiescent supply current and 1.2A p-p maximum output current. (15mhz, 50V/us and 10mA supply current for LM318) However, when I tested, output was again distorted and could not even light up a single LED.

Can forumers please advise me on how to choose opamps that are suitable for my application?

1. I have looked for GBWP >= 20Mhz, slew rate >= 50V/us and (output current >= 300mA OR supply current >= 15mA). Are the parameters AND values I am looking at correct or should I look at some additional parameters?
2. What is the significance of each of these parameters in my application?
3. What is the difference between supply current and output current and what values should these be for my purpose?
4. Why is there a high distortion for a 1Mhz input although the GBWP for both OPA561 and LM318 is about 15-17 Mhz?
5. Also, from the list below, could you please advise me if any of the opamps are suitable for my purpose?
6. Lastly, I am also looking at class AB amplifiers. However, these do not respond at frequencies greater than 100khz as there is heavy distortion in the power MOSFETs. How can I overcome this problem?


LM318 - http://www.mit.edu/~6.301/LM118.pdf
OPA561 - **broken link removed**

Other possibly suitable opamps. I looked under 2 categories, high frequency and high output current/power amplifiers. Checked in RS components and Farnell:-
LM8261 - **broken link removed**
TSH73CD/TSH74CD - **broken link removed**
OPA2604 - **broken link removed**
AD8017 - **broken link removed**

I know this is a long question (and my first). Thank you if you actually read through till the end! I shall try to post a video regarding this as soon as i can.

Thanks and regards,
Naveen LP

 

[karesz]

Hi,
Only a short impression:
Im not shure that its your right solution-because an parallel resonant circuit hase voltage maximum_HIGHER as for your OpAmp right is!
K.

 

[naveenlp]

Re: Choosing opamp as small power supply with 1 Mhz signal i

@karesz: Thanks for your reply.

I dont entirely understand why you say an opamp may not be the best solution here. I am not sure an opamp or class AB amplifier is the best solution, but as of now I can't think of anything else to use.

Also, as I have explained in my post, The existing setup using an opamp works with an LM318n opamp. However, the output waveform at 1Mhz is highly distorted and the power is arnd 3 times (estimate) lower than required values. My reasoning is that if I used an opamp that has a higher GBWP and higher output current/supply current, I should get the power output I require.

 

[FvM]

The question can't be answered without a rough estimation of the load impedance. You assumed some values in your first post, but it's completely unclear where you get them. In principle, an ideal parallel LC circuit has infinite impedance at resonance frequency, only the coil (and possibly coupled load) losses make up an input impedance. Are the values from a real measurement, an EM simulation or some kind of hand calculation? Did you tune the LC circuit? What kind of distortion do you observe?

 

[naveenlp]

Re: Choosing opamp as small power supply with 1 Mhz signal i

@FvM
I noticed you had answered a very similar question before. The differences are my power requirements (100mW) and distances (~6cm) are much smaller. Also I am using a colpitts oscillator connected to a opamp/class AB amplifier.

Rgd your question about the load, until now I was building the coils and colpitts oscillator and did not worry about the power transferred as I was earlier using a function generator for testing. But my attempts with the opamp were the first to actually transmit power.

I suggested around 10-20mW was being transferred solely because I had connected an LED in parallel to the receiver coil and it lit up brightly. I was told by my supervisor that an LED consumed around 10mW of power and hence the estimate.

My intended load at the receiver is a rechargeable 3.2V battery's internal impedance. Now, accd to wikipedia, this is very small and comparable to resistance of my coils. So I have taken readings of power delivered at receiver with no load by measuring voltage across capacitor and current through the coil. For a 7V p-p input FROM A FUNCTION GENERATOR, I measured a power of 15-25mW delivered to the receiver coil. (I measured 40mW without accounting for the power factor. So with the power factor, I expect around 15-25mW)

Also, I am not an expert in this field by any means. If you feel some important requirement has been left out in my analysis, please do let me know.


2 youtube videos showing my setup:-

Transmitter and receiver in resonance with LED
https://www.youtube.com/watch?v=ovvWAKJLowQ

Voltage and current at output on the oscilloscope. This is how I measured the power at output.
https://www.youtube.com/watch?v=1kIR1L0fw0w

Added after 15 minutes:

@FvM
Another thing is, I do not really understand how resonance affects the load impedance at the source side. I can see that the LC tank presents an infinite impedance overall. Also, I guess only reactive power should be delivered to the LC circuit. However, I do not really understand how this changes when you bring in a receiver coil at the same resonant frequency.

 

[malli_1729]

Re: Choosing opamp as small power supply with 1 Mhz signal i

You can try with more Slew rate Op amps with slew rate may be > 100V/us

i don't know exactly but with 50V/us and Vp-p = 6.8v , we can go upto 2.34 Mhz without distorsion......
it's not exact ..give a trial....