"A Review on Holes, Aerial TRF7960/70 Tuning and misunderstandings"
Hello Josh and Daniel and others. I think I have a "clearer" understanding on this "hole business"...
Your previous comments were/are valuable to me. The consequence of my/our confusion/curiosity has had a reward as we have now used it to devise a "better" way to read distant FDX tags at 134.2KHz. (even though our focus is still on HDX)...this Full Duplex concept has "issues" common to 134.2KHz and 13.56MHz systems.
Indirectly this work follows from our new understanding about the "dual Receiver input" concept found in all TI design variants using the TRF7960/7970 (including the experimental 4Watt RF amplifier).
My proposition here is a bit long winded and a bit convoluted and the full story cannot be done justice in a few paragraphs however any comments that follow are worth considering as even now there are still some missing links in my understanding.
The holes you speak of are different: they "reside in" the LCR network between the 50 Ohm aerial and the RX_IN1 and RX_In2 pins of the TRF7960. let us not be concerned with Phase-Shift-Modulation for the moment , it is a distraction but beautifully addressed by TI and I think the 45 degrees is an artefact/requirement of the PSK
demodulation process...TI tried to get one chip to be multi functional and well there are consequences.
Your forum and TI documentation references is not(only) to spatial holes at all, BUT about "NOTCHES" in the frequency response behavior of the interface network AND THEY MOVE AND MORPH AS YOU APPROACH WITH A TAG...that is, as the coefficient of coupling between the Tag and the Receiver Loop changes so does the behavior of this clever network also change...(spatial null angles are still in play, and explain your curious reference to "bands of holes)
This is the real reason why there are two receiver channels and it is cleverly conceals a way to mitigate false detections by "noise"... by comparing the two RSSI values and then "MUX switching/transposing"...this is all driven by a "hole/notch" that is formed as the tag approaches...
TI have clever circuit network "hole/notch" and unless it is in the right place then poor or no tag detection will be a consequence.
This seems to me to be a most brilliant thing as the TRF7960/70 can now be made (by external micro software) to ignore every thing until a certain signal to noise ratio is achieved....ie the tag distance must be right (coefficient coupling about 0.2 for the EVM) before the decoding process begins.
What this then means is that far distant radio interference etc cannot cause spurious/partial reading attempts..the system simply will not provide open access until a card is physically present and even then this card must be physically closer than a certain distance....this is an amazing piece of foresight and another level of
identity/theft/security, even before all the clever coding/passwords and crc etc...a first level of defence..all done with a hand full of simple "but precise" components.
I keep in mind the general concept of TRF7960/ISO15693/14444A/B is that of a Full duplex solution: that is the transmitter is still "annoyingly" emitting while at the same time the Receiver(s) is straining hard to hear a faint reply from an approaching tag...
This clever network "nulls out" the feedback from the transmitter and enables a "quiet place/zone/null/hole/portal" through which the receiver can listen better for approaching Tags/Cards.To be more precise, with no tag in the vicinity, there is a notch [node or "hole"] centred at 13.54MHz
The TRF7960 can "ADAPT" itself, as when a faint card gets suitably close it switches (MUX) all it's best resources over to cleanly recovering the Tag/card telegram...though I think you have to tell it how and when to perform this transformation.
One channel is good for verge of perception and the other channel is better for demodulatin this is an "A" before "B" piece of logic...a very clever man devised this process.
To "cut to the chase"... for those struggling with Aerial Manufacture/Tuning for the TRF7960/70. May I suggest you do/try it this way.
1. Do not attempt to peform tuning by replacing the existing/new coil with a 50 ohn termination resistor
2. In isolation, continue to Tune/Match your aerial to present 50 ohm, as per the well explained use of Smith-
Charts... for those with a network Analyser etc this can be confirmed.Now connect it to the EVM.
The standard/classic TI circuitry between the TRF7960/70 and the EVM PCB aerial or your own home made variant DEMANDS
you MUST meet this requirement. There is a notch waiting and you must find it.
3. Either use a dual channel oscilloscope (use X10 probes) to simultaneously watch RX_IN1(pin5) and RX_IN2(pin9)
or software devise to watch the two RSSI values....you are looking for a null response at RX_IN1
4. Now well away from all metal surfaces/loose metals etc and no tag present, now "tune/fiddle" the coil's parallel resonating capacitor +/-30pF until a notch is found (other components could also be varied but not just yet)
5. Bring closer now a known good card and you should "see" the notch" disappear from" RX_In1 and "re-appear" over in RX_IN2...as one channel falls down the other goes up.... this is morphing behaviour ....,,when their magnitudes are equal the card should be about 10cm away and the RSSI numbers should be equal and if you have the software it should at this point tell the MUX to "swap over the inputs"....now the Main Receiver is "looking" at the best channel to identify the Tag/Card.
Anyhow, I would encourage those with Spice to model the TRF7960 recommended/mandatory "duplexing network" and ignore or confirm my entry proposal...My suspicion is that the TRF760 cannot live without it...it is a baby only a mother could love yet it excels in clever.
If you want I have some support documentatin...I use too many words...sorry
Regards
Ray