Schematic help - Complementary transistor pair

[mobbarley]

Hi! The following excerpt is taken from microchips 125khz rfid reader reference design:




A 125khz signal is injected through the filter around L1 and current is pulsed through the main coil L2 creating the 125khz carrier signal.
Why is this a pair of transistors? Does C15 charge and then discharge through the coil on the other half of the wave form? I was hoping to drive something similar from a PIC's PWM hardware using mosfets, do I need this push pull arrangement? is it more efficient?

Thanks!

[mobbarley]

Full app note: http://ww1.microchip.com/downloads/en/devicedoc/51115f.pdf
Page 101 ' FSK Reader Schematic '

[Simon]

what exactly are you trying to achieve. To simply drive something with PWM from a pic you just need one mosfet for the simplest setup and usually this is a logic level mosfet (IRL540 being a nice choice)

[mobbarley]

I'm hoping to build a small RFID reader. I am trying to work out why the circuit uses this npn/pnp pair, as I have only seen this with used with +- power supply rails - I am assuming that it helps promote the resonance as L2/C15 are tuned to be resonant at 125khz. \
I was wondering if I could drive the coil with a single N-ch fet as you said, would it be less efficient? would it break the fsk conditioning (starts bottom left)?

[Psi]

Maybe so pushing from gnd to vcc and pulling from vcc to gnd are both nice and fast.  If you had one transistor or fet then (depending which way you did it) one direction would use a resistor and wouldn't be as fast.

Being an inductor, L2 will try to oppose you changing the voltage on it, so you need a high current to force the change quickly. A fet pulling it low and a resistor pushing it high would result in the pushing being current limited by the resistor.

[am2pgs]

I am not familiar with the application but:

It appears to me that the output of the push-pull stage is a ~sine signal (not a square signal). output of the processor is filtered, ac coupled and amplified by the class AB push-pull stage. 

The push pull stage is really better if you want a linear output rather than simply on/off. of course you could use something like a class A amplifier, but that wouldn't be very efficient.

[jimmc]

It doesn't help that there are errors on the circuit diagram, L2 is 1.62mH (and L1 1mH ,as per text and parts list). L2 (1.62mH) resonates C15 (1nF) at 125kHz meaning that the amplifier has to drive a low impedance.
Since the drive to the amplifier before the filter /matching circuit (L1, C9 & C12) is a square-wave C15 will charge to half the 12v supply negating the need for split supplies.

With the push-pull drive the output impedance is low over the whole of the cycle, a single MOSFET would either be short or open for half the cycle depending on configuration (shunt or series feed), either would destroy the Q of the resonant circuit.

Note the Q of the resonant circuit increases the received voltage across C15 and hence sensitivity.

Jim

[mobbarley]

Thanks Jim - Did a sim and saw what you said. And there are plenty of errors in that document!