TX amplifier design

[sled]

I want to build a simple TX coil driver using a DAC followed by an amplifier for a VLF IB detector. However I'm not sure how much current an average IB metal detector puts through the coil.
What I intend to use is a either a headphone amplifier or a small loudspeaker amplifier (single-ended or bridge-tied-load).

System parameters:

TX inductance: 0.95mH
TX resistance @ DC: 4.3 Ohm
TX capacitance: 150nF
TX tank type: Series RLC
TX tank resonance: ~13.33kHz
DAC Output voltage: 0..1Vrms (2.8Vpp)
Supply Voltage: +5V / +3.3V

The low resistance of the TX coil causes me some headache because most headphone/small speaker amplifiers are only specified for 8/16 Ohm loads.

My question boils down to how much power / current does the amplifier need to deliver

[Skippy]

I've just done some 'back of envelope' calculations. I reckon driving your circuit with approx. 600mV pk-pk ( 200 mV RMS) will give a drive level comparable to commercial machines. Drive power will be around 10 mW, drive current 50 mA RMS.
You could lower the Q slightly, to 15, for example, with little detrimental effect, by adding 1.5 Ohms in series with the coil. You would then need to apply more drive, eg. 800 mV pk-pk, at the same sort of current. So there's not much to be gained by increasing the resistance.

The load resistance specified for the amplifiers is what they can drive to full voltage levels. They could happily drive 4 Ohms, it's just that the max output level will need to be lowered.

[Altra]

You maybe able to adapt this prototype front end I built for a dsp vlf. The ad5933 has a bulit in dac that generates a 3V sinewave output. This output is ac coupled through C8 and R17 to the non inverting input of a TCA0372 power opamp. The Tx gain is adjusted for a symetrical, non distorted output on TP (approx. 10V). R19 and C15 filter high frequecy noise, which I did on a trial and error basis depending on the tx frequency and coil. The TX coil should have inductance in the 500 to 1000uh range. C13 on the tx virtual ground caused uncontrolable oscillations as drawn in the schematic. It should be increased to 100uf or disconnect from pin 6 and connect to pin 2. The circuit actually worked well and was able to gb and target ID. But I was using a TFT display and it turned into a user interface project and then got side tracked on something else.

[sled]

@Altra: Thanks a lot for sharing your schematic, really interesting project - very close to what I'm trying to achieve I didn't know such high current output op-amp exist, could be a good candidate since I don't need any amplification at all, a buffer would be good enough.

What I've found in my parts bin is a TPA6111A2 (http://www.ti.com/product/tpa6112a2) which is able to drive 150mW into an 16Ohm load.

Interesting fact is the datasheet of the TPA6111A2 has a section for a 6 Ohm load (I'm not sure whether this is a typo):



So my plan is to add a voltage divider to the DAC's output to linearly map the 1Vrms to an achievable range for the amplifier.
Question is whether the TPA6111 is unity gain stable

By the way, do I need a flyback diode around the coil? Should I drive the TX coil into resonance or a little below/above the resonance frequency?

[Altra]

I would calculate the ac reactance of your TX RLC at 13.33khz. This will tell you what your driver needs to deliver. It will probably be over a 100 ohms. I used a power amp in case it was needed. In reality it ran on very low current. The TX was surprisingly flat from 4 to 23khz. What really changed with frequency was the RX phase and amplitude. As far as driving your tx at resonance, the math say's yes. The RX is the one you want to run off resonance. No diode on the coil, that would just clip the sine wave. Post your progress.

[sled]

I would calculate the ac reactance of your TX RLC at 13.33khz. This will tell you what your driver needs to deliver. It will probably be over a 100 ohms.

If I remember correctly the reactance at resonance is equal to the ohmical resistance e.g Z=R, so it will be about ~4.3Ohm + some ESR from the capacitor. At the moment I'm coding a small DDS signal generator go get an adjustable sine-wave (time to buy a proper signal generator...)