RFID 13.56MHz Antenna using the TRF7960EVM 4W pwr amplifier

Hello Vladimir...

Allow me to make some comments on your efforts to achieve further
detection range...my domain is in the Low Frequency 134.5KHz spectrum, however
perhaps any help is better than none.

Your theme is "reciprocity" and it is a matter that also concerns us. We have some common issues.

1. we find coils wound with significient radiused corners give more range than sharp
corner square PCB etched coils...in the extreme these radius revert to circular
coils and along their eigen axis you will achieve better stimulation and reception range.

2. Square coils have "hot-spots" in each corner, where the intense flux curls
quickly and stays local ...that is, it is 'wasted energy" as far as distant tags are
concerned. Round coils will detect a bit further.

3. Coils wound in a "flat-pancake" manner, will always exhibit less range than coils
wound in "solenoid" manner...mentally follow the flux lines emitting from a pancake
coil and you will see the "net" flux is predominatetly directed out
sideways...again, less eigen flux for an approaching RFID tag/card....Revert to
solenoid winding and achieve more range... but loose the simplicity of a PCB aerial.

4.There are formulas, but the larger the aperture of the Transmit/Receive coil, then
the emitted flux is more intense along the eigen axis, and the with this larger
aperture the "simultaneous receiver" intercepts/captures a larger proportion of the re-emitted
influence/signal of the replying tag...together the bigger diameter works well for
both functions...more range results.

5.We found the RFM007 module for 134.2KHz has got reciprocity just right, when using
their commercial coil...it must have been no mean feat to have found the receiver
gain setting to achieve this matching reciprocity...in it's raw state I suggest your
TRF7969EVM pre-manufactured board also received good matching of transmission and
reception engineering...however we like you choose to push the barriers just a bit
further....and issues arise.

6. More turns is good, for a given diameter coil, as more transformer/induced voltage from the
fixed-power transmitting tag is achieved...however the inductance changes greatly and
you must re-resonate and lower the new Q-Factor to again not greatly slice off the
received modulation sidebands...you will need to recalculate...smith chars help here.

7. We found with FDX 134.2KHz cattle tags, if the receiving tuned circuit was "off-
tune" either above or below the nominal carrier (134.2Khz for us and 13MHz for you)
then a strange thing happens, and this is if the phase shift of the received  AM sidebands
approaches 90 degrees the side-band doubling occurs...roughly then...if a 1 KHz tone was
modulating the tag,then the off-tune receiver starts to demodulate/see 2KHz...a spectrum view of the
tag emissions will say all is well but the RFID receiver say's "no, I don't understand you"...side-band phase
distortion does occur and it is just one other reason why HDX is far superior tho FDX for cattle tags.

8. Concerning Parallel or series resonance, both have the aim to achieve "amplified
circulating current " swirling/oscillating around the transmitting coil....and therefore emit the
greatest magnetic flux...think in terms of primary and secondary of a transformer
(as this mode of RFID is more that, than radio transmission)
In parallel mode your capacitor/active devices are also parallel/across the coil and its peak/peak
voltage can go into hundreds of volts...the components will soon fail.
In series resonance the coil is "current -driven" and all that's needed is a low
source impedance driver...the RFM007 uses a step down transformer with I think a one
turn secondary and this is then in series with the coil...much better idea.

9. Concerning "mutual coupling" between Reader-Coil" and "Tag-Coil"...there is a
problem for you more so than for me...your 13MHz coils are comparable in size to the
Tag/card coils...so up close the coefficient of coupling is very high...where-as for
134.2KHz, even up "very close" the RFID tag still has a very low coefficient of
coupling...all this relates to coupled tuned circuits and in your case a close tag
may in fact not be read because the relationship of the modulating sidebands will
corrupt the shape of the recovered video... wherein resides the Tag Identification....

10. I suspect this matter has some role beyond that of "data collision" and all your
software algorithms will never solve this...when two of your tags are presented
one on top the other, neither will or can be cleverly separated.

11. just some things to keep your eye on Vladimir...more range is not just more brute
force power...as we too have learned...what are you proposing to do now?

Kind Regards

Ray

Hello Ray,

Thanks for this info as it makes more sense to me and just re-confirms my findings. I might go with a multi-layer PCB to kind of mimic the solenoid as much as possible and round my corners. I need a rectangular shape antenna so I can maximize the field coverage as the tag moves from one side of the antenna to the other. (Parallel motion) Antenna size is about 150mmX75mm.

Problem is that when trying to maximize the number of turns of the antenna the inductance increase a lot like you stated previously. Going with a parallel or series resonance I have to keep my inductance value below around 1.5 to 2uH or else I can’t match the output impedance of the 4W amplifier (50 ohm antenna). Are there other topologies I can approach so that using an inductor value higher than 2uH won’t be an issue? I would like to increase the turns as much as possible. Just having 2 full turns at 150mmX75mm is already ~1.5uH.

I would like to go series resonance but that means have to re-calculate the 4W amplifier output impedance to be lower, closer to 10 ohms or less. If I’m not mistaken the output impedance of the amplifier from TI EVM is more or less 50ohms? Since the low pass and band pass filters were designed to be optimized at 50 ohms.

I have also noticed that even If I completely tuned an antenna at 50ohms my RFID carrier frequency (13.56MHz) slightly distorts but if I load a pure 50ohm resistor the output has minimal distortion.

I also designed a 6 turn spiral antenna at 100mmX100mm where it is impossible to tune the circuit to 13.56MHz at 50 ohms using parallel resonance topology. Therefore I just place a cap value to try to remove as much as possible any reactance and to my surprise this was the best result for reading distances.

Vlad

Hello Vlad

I see your preference is to achieve "best detection" when the RFID card approaches from one side of the Reader-Coil...not from the front?

Am I correct that your application is like a "swipe card reader as seen for train or bus payment"... you "hope" or want every patron to conform to this manner of presentation.

If "every" presentation proceedes from left to right or right to left and "every" presentation has the card levitating in the same plane as the PCB-ReaderCoil, then on reflection I suggest the pancake spiral might be a good choice...the solenoid is probably better for a "frontal" card approach.

You face several peculiar dilemmas ...and they can be seen through the vector representation of the "curling" flux field of the "transmitter (and Receiver)

There are two graphs that will help you visualize this matter and see the null response angles (sinX)/X and 1/X-(CosX)/X ...these not exactly right but close for now.

If side swiping is a "must have" ... then I suspect your card in fact gets "Successfully read" just before it crosses over the outer perimeter of the closest long side of the coil...perhaps you could confirm that for me.

A concept I find useful is to visualize a card/tag as approaching from infinity...a card/tag does not "magically" just appear conveniently in front of a Reader.

This leads you to recognize that "detection" is an horizon event, that occurs on the threshold of perception...and once done it is now irrelevent how much better the illumination and reception of the approaching card gets...

The public or housewife "thinks" they must carefully center their card on top of the place provided...little do they realize it has been read long before they have nestelled it into its placeof best visual-fit

If the act of "swiping" is essential for you...I recommend you have a long thin loop aerial-coil...say 30cm by3cm and remove the temptation of the public user from trying to "dock" his/her card on top ....their thinking is like a hellicopter landing on a off-shore oil rig?...force them to conform....make your Reader box a different shape?

Perhaps? ...perhaps...I am still uncertain about the two receivers in the 7970...and still passing curious that these receiver demodulators are little more than "cat's whisker crystal sets"... however this is for another time.

Your project is of interest to me, for later I will migrate up to 13.56MHz.

Regards

Ray

Hello Ray,

Achieving best detection is surely my goal and in my case the actual distance from the card/tag to the reader will be constant as the tag travels from one side of the antenna to the other. I would like to achieve the minimum read access time from one card to the next. I will consider designing an antenna shaped like a ruler and see how that works out.

As for the receiver demodulator,  I’m  still not 100% sure how the 2 Rx inputs help eliminate holes in the field. Why having 2 RX inputs with a 45deg phase delay from each other improves on reading tags (eliminating holes in the magnetic field of the antenna)?

BR,

Vlad

hello Vlad...(and Raghu)

I had some spare time and used SPICE to look at your problem...above you on this form is "Raghu" who is somewhat also in your orbit....I attach a pdf binder to give you both an idea of what I found....tuning of your 13MHz Readers must be "DEADLY"... and still I cannot understand the 45 degrees, but I now read somewhere it gets use once for transmission then again for reception (simultaneousness action) to eventually cause a 90 degree phase shift to the input to the Phase Detector....what a maze...

Any how I am trying to attach this pdf file for you both...(proving difficult for me though) I show stepping of Reader coil inductance and tuning to get a sense of how sensitive/critical things are.

I think I like 134.2KHz better...just what are you attaching your RFID tags too?

Watch out with your 4W Amplifier, I see they on purpose detune the Reader-Coil off resonance so as to NOT achieve high breakdown voltages...when you added capacitance you just move back ino dangerous waters....whatch out for blue smoke....interesting...still can't do attachments.

Regards

Ray

trying to attach a pdf file ...again