Designing a band-pass filter for 13.56 Mhz (detect RFID readers)?

[hyperzap]

Hi,

As a project to get into some analog electronics (done arduino/ARM stuff to death) I thought it would be fun to build a thing that can tell 125khz RFID readers and 13.56Mhz RFID readers apart. I think I can build the 125khz one fairly trivially with an Op-amp based filter, but 13.56Mhz is too fast for hobbyist Op-amps. I know I could probably do this with just passives because of the magnitude of the input signal, but I would rather get some confidence in active stuff.

I'm not sure about how to build the 13.56Mhz stage, but I have fleshed out a high-level plan below. Questions are at the bottom.


Plan for the reciever coil
As RFID readers produce powerful fields which I can get as close to as I like, I was hoping not to care about impedance or matching or coil details. I am planning on just putting a big circular coil on the PCB, connecting one end to ground through a large DC-blocking capacitor (100nF should do the trick?), and the other to the input of my 125khz detector stage and my 13.56Mhz detector stage.


My plan for detecting 125khz signals (does this make sense?)
Looking at the datasheet for the LM358:



It looks like I will have ~16dB of gain at my target frequency. So if I build a circuit like this one on allaboutcircuits:



Then feed that into a diode envelope detector, then buffer, than an LED, I think it will work.




But hobby op-amps are too slow for 13Mhz, how can I build a filter?

Cheap op-amps loose their usefulness at this frequency, so my reasoning is I should make this circuit with a BJT. The 2N3904 has a CBP of >300Mhz, which sounds cool because its a lot higher than my target frequency. Additionally, it has this graph in the datasheet:

Which is like 30dB gain at my target frequency.

This is where my problems begin. Whilst building a circuit around op-amps is easy because they have awesome properties, BJTs have wierd bias and dropout and saturation regions and other stuff - easy for digital but now I want to operate them in a linear region I'm a little confused about what I need to do.

I found this circuit on google images:


And I think if I whack a RC high-pass before its input, and on its output connect a envelope detector -> buffer -> LED, I will be able to get the LED to light in the presence of a 13.56 Mhz field.

Questions

1. Is my component selection sane?
2. How do I select the bias voltage in my BJT amplifier stage?
3. I guess I have to start worrying about impedance. How do I select the values of Re, Rl, R1 & R2 in light of impedance but without screwing up my bias voltage?
4. I'm guessing I need to pick the value of C2 to produce a low resistance path @13.56 Mhz?
5. Any other recommendations? is this going to work?

Thanks! 

[nugglix]

Hi!

There is a whole series of videos showing a 13.56MHz rfid detector:


If it were only for filtering and amplification you could use a NE592 video amp.
http://www.onsemi.com/pub/Collateral/NE592-D.PDF

And of course a reference to Alan's nicely made videos:


Btw, there is a list of all of his videos as pdf:
http://www.dorkage.com/youtube/W2AEW_video_index.pdf

Hope you get some ideas!
Cheers

[tggzzz]

Look at how Tayloe mixers and superheterodyne receivers work.

[orolo]

In the MHz range it's normal to use passive filters. The reason passives are not usually employed in audio frequencies is that inductors are too big at these frequencies. Instead, op amps, resistors and capacitors are ingeniously used to substitute audio inductors.

The problem with passive RF filters is that you need to wind your own inductors, and for that you need adequate toroids and some practice. If you are aiming for a bandpass filter, you'll need also to tweak the filter, since component variations will not land you at the desired frequency once you go from calculations to practice. You'll need either trimcaps, trimming inductors, or a lot of luck.

Perhaps the best in this case is to try a tuned amp. This is very similar to the common emitter amplifier you posted, but with RL changed by a resonant tank. This amplifier gives you gain only at your band of interest, and it's rather easy to tune and understand. It is very selective if done right, which will help you filter out a lot of interference in your magnetic sensor, particularly AM stations.

[danadak]

Use the free TI tool to design your filter.


http://www.ti.com/tool/filterpro&DCMP=hpa_amp_general&HQS=NotApplicable+OT+filterpro


Regards, Dana.