2
\$\begingroup\$

I am currently designing an NFC reader in the context of my PhD thesis. The NFC reader will use a PN532 chip from NXP (which is one of the requirements of my circuit). I plan to draw a rectangular loop inductor on the PCB (which is often referred to as an NFC antenna in wireless design) which would approximately occupy a 3x3 cm square (up to 4x4 cm). There will be other components on this PCB such as an LED driver, some LEDs, etc. The main micro-controller will stand on a separate PCB and will communicate with this board through an SPI link. Space is the main constraint: the volume occupied by the PCB has to be minimized as much as possible.

I read this document quite carefully (NXP Antenna Design Guide) and I have very specific questions related to the design of the PCB that will carry the NFC controller and antenna:

  • Could it be acceptable to put the components (NFC controller, LED driver) on the same PCB as the NFC antenna?
  • What would be the "best" position relatively to the antenna in this case? (i.e. one that would limit eddy currents and inductance increase)
  • How to efficiently shield the main controller board? Will a sheet of copper do the job (instead of ferrite)?
  • Is there an advantage to have different length and width for the antenna?
\$\endgroup\$
4
\$\begingroup\$

I'm no expert on 13MHz-carrier data-transmission but i do come from a background of transmitting power magnetically to rotating circuit boards and receiving modulated data back at 80MHz so if I've misunderstood anything please let me know.

•Could it be acceptable to put the components (NFC controller, LED > driver) on the same PCB as the NFC antenna?

Yes this is acceptable but keep circuits as far distant as possible so that the receive side stands the best chance of working effectively.

•What would be the "best" position relatively to the antenna in this case? (i.e. one that would limit eddy currents and inductance increase)

According to the very long-winded document you linked, the antenna Q needs to be a bit less than 30 and this is easily checked by using standard oscilloscopes and signal generators. Q is centre frequency (i.e. 13.56MHz) divided by 3dB bandwidth and for a Q of 30 bandwidth will be about 450kHz.

So you need to make sure your centre frequency is spot on and that you have chosen Rq values to limit the Q sufficiently or you'll be possibly breaking some regulations about emissions levels.

I'd position the antenna and compensate for eddy current effects (frequency and Q changes) with a slight retune then add or remove a little bit of Rq to give a Q of 25 to 30.

•How to efficiently shield the main controller board? Will a sheet of copper do the job(instead of ferrite)?

You may not need to do much (if anything) here but I'd recommend using the proper ferrite sheet (Ferroxcube used to sell it because I've used it). Using copper sheet may detune and slightly desensitize the antenna depending on how close to the antenna it is.

•Is there an advantage to have different length and width for the antenna?

Yes, the squarer the better. Bearing in mind this is magnetic coupled data and you should be looking for a loop with as much inductance as possible but not so much that it's self resonant frequency is lower than (say) 15MHz.

As a simple idea about inductance - inductance is defined as total flux produced per amp. The bigger the cross sectional area, the more flux is produced per amp. A long thin loop however has less inductance than a square loop of the same area.

|improve this answer|||||
\$\endgroup\$
  • \$\begingroup\$ Thanks for all the details! Unfortunately I can't use ferrite for shielding the rest of the circuit, because the shield has to remain in the PCB itself (I want to reduce the assembly time as much as possible). I wrote a squared loop antenna on the top layer and will follow your recommendations - tune the components around the antenna with the final PCB with a network analyzer. \$\endgroup\$ – Ant2N Oct 3 '13 at 16:11
  • \$\begingroup\$ Good luck with this and feel free to up-vote any answers and/or accept appropriate ones. That's the consultancy fee LOL \$\endgroup\$ – Andy aka Oct 3 '13 at 16:39
  • \$\begingroup\$ Haha terrific! Unfortunately I can't vote up because of my reputation but at least I accepted the answer! I have one more question though if I can abuse from your knowledge: would that be a good idea to split the ground plane under the antenna to separate the "shield" (square of copper) from the ground net? \$\endgroup\$ – Ant2N Oct 4 '13 at 8:59
  • \$\begingroup\$ Having a ground plane under your antenna will seriously reduce it's effectiveness - you are looking to receive only the magnetic field (not the electromagnetic filed) and therefore avoiding any large lumps of copper around the "antenna" coil will make it more effective. \$\endgroup\$ – Andy aka Oct 4 '13 at 9:04
  • \$\begingroup\$ For sure, but I have to protect the other components, as well as the antenna from the rest of the circuit. For instance there's a big connector just under the antenna (which position is, unfortunately, fixed) and there are 12 PWM-driven LEDs on the very same PCB. On the other hand, the maximum distance between the two controllers will not exceed 6mm and once the two parts are assembled, the two coils will be perfectly aligned. Does that seem acceptable to you? \$\endgroup\$ – Ant2N Oct 7 '13 at 14:26
0
\$\begingroup\$

Putting all the components on the same PCB is commonly done and should be fine. For an example of that, check out the schematic and PCB for: http://www.seeedstudio.com/wiki/NFC_Shield_V1.0 or when separated: http://www.seeedstudio.com/wiki/NFC_Shield_V2.0

As for shielding the board, the two above just have a copper pour layer on the side with the components. I have both boards and they work very well.

Good luck!

|improve this answer|||||
\$\endgroup\$
  • \$\begingroup\$ Thanks for the links! I'll try to keep the size of the antenna as small as possible to be able to put the components aside the antenna as in NFC Shield V1.0 (the antenna has to be centred on my PCB for symmetry). Is there a limiting factor for the size of the antenna, apart from the track width? \$\endgroup\$ – Ant2N Jul 18 '13 at 12:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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