4
\$\begingroup\$

I'm currently working on building my own simple RFID reader. I've already build a small circuit in which my coil is fed from 5V current source or is connected to ground dewpending on the 125kHz square wave signal from the microcontroller.

I've measured the frequency of the signal between the coil and capacitor C1 - it's exactly 125kHz. D1, C2, R3 are a kind of AM demodulator.

What I wanted to achieve after bringing the transponder closer to the coil was a change in amplitude, but not of the whole signal, just in some of the peaks. What I got is the whole signal amplitude decrease by about 2V.

Can anyone tell me what is wrong with my setup? I'm attaching my schema and an oscilloscope measurements between L1 and C1 (yellow) and after D1 (blue).

Thank you! Schema Measurements

UPDATE (20140323):

The same signal, but I've zoomed out in time domain (without transponder): Signal without transponder near the coil

The same signal, but I've zoomed out in time domain (with transponder): Signal with transponder near the coil

If we zoom in the signal after the diode, with transponder near the coil, we can see something like this (with and without software low pass filter):

With filter Without filter

The same signal without a transponder near the coil: enter image description here

\$\endgroup\$
4
  • \$\begingroup\$ That sounds about right. I made something similar a while back. Zoom out in the time domain and you should be able to see the modulation. \$\endgroup\$
    – s3c
    Mar 23, 2014 at 10:06
  • \$\begingroup\$ Where did the circuit originate from? \$\endgroup\$
    – Andy aka
    Mar 23, 2014 at 11:03
  • \$\begingroup\$ @s3c, I've attached some new screenshots from the oscilloscope. I can see the modulation, but I don't understand the signal shape after the diode (the blue one). Shouldn't it look like this: radio-electronics.com/info/rf-technology-design/am-reception/… ? \$\endgroup\$
    – Rubid
    Mar 23, 2014 at 14:36
  • 1
    \$\begingroup\$ It does look like that! What's the problem? You're clearly getting 1.95 kHz (Fcarrier/64) modulation from the tag; now you need to figure out how to decode it. It's possible that the combination of the low-pass characteristic of the AM detector, plus your software filter, are wiping out the relevant timing nuances; try opening up the bandwidth a bit. \$\endgroup\$
    – Dave Tweed
    Mar 23, 2014 at 15:16

2 Answers 2

1
\$\begingroup\$

The large, slow amplitude variations occur because the tag must necessarily consume power from your transmitter in order to operate at all.

The data you're looking for will be in the form of low-amplitude (a few mV), fast (typically 3.9 or 7.8 kHz) square waves. You'll need a high-pass filter after your detector in order to separate the two kinds of variations.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks @DaveTweed. I'm not sure if my detector is working correctly. I've attached some new screenshots from the oscilloscope. \$\endgroup\$
    – Rubid
    Mar 23, 2014 at 14:42
0
\$\begingroup\$

There are better circuits that what you have proposed and I would recommend this one from MAXIM because it is more linear and you'll have less trouble detecting the data: -

enter image description here

This one is shown for FSK but the same circuit (the important areas) are the same whether you use PSK or AM.

You can find the circuit and a lot of good tips and chips in this document. The document contains circuits for PSK, FSK and AM.

The basic problem with your circuit is lack of sensitivity - you have no gain following you AM detector and if you are serious about this you do need gain. Also, you are driving the coil with a half-hearted square wave when it's always going to be more effective with a push-pull driver and a sinewave - note the push-pull driver in the diagram and the filtering of the oscillator signals before they enter the push-pull amplifier's input.

\$\endgroup\$
2
  • \$\begingroup\$ FSK refers to how the data is encoded, not the carrier frequency. It's all AM as far as the detector is concerned. For different kinds of data encoding, you'll need to change the logic that follows the slicer (U8:A) at the output of the filter/gain chain. \$\endgroup\$
    – Dave Tweed
    Mar 23, 2014 at 12:19
  • \$\begingroup\$ @Andy aka, thanks for your suggestions. When it comes to gain, it will be the next step, after the current circuit will work. I will use a coupling capacitor and put the signal into an opamp. For now, I want to understand what I'm missing here. \$\endgroup\$
    – Rubid
    Mar 23, 2014 at 14:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.