EMI filter for CR95HF and more

[MikeObs]

Hello,

I am making a project with CR95HF on a custom PCB and I have been researching some information about the antenna for the transceiver operated by the early mentioned chip. I have couple questions to clear up your documentations for CR95HF and X-NUCLEO-NFC03A1.

1. In the documentation AN5248, you mention in the calculation for EMI cut-off frequency that fc= 1 / 2*PI*sqrt[(2*L0) * (C0/2)] = 14.6 MHz, however my calculation brought me a result of fc = 15.04 MHz. What causes this almost 0.5 MHz jump?
2. In the same document AN5248, you use this drawing of schematic part close to antenna: There you mention Low pass filter contains 2 capacitors, C01 and C02, that are equal in value. However in the schematic of X-NUCLEO-NFC03A1 (for example UM2479), there are 4 capacitors in parallel (C9, C10, C13 and C14).What is the difference between using 2 and 4 capacitors in the EMI filter. According to the fc formula, it should not change the cut-off frequency so I have been wondering what is the purpose of doubling them in parallel?
3. Last question I have been wondering about is that in the CR95HF Antenna Design Query post that was made earlier you mentioned that X-NUCLEO-NFC03A1 antenna is 34x47mm/4 turns/0.8ohm~1120nH antenna. I downloaded the gerber files and put the parameters into your eDesign suite to come up with inductance result of 1260 nH. Could that difference be caused by rounded corners of the actual antenna on X-NUCLEO-NFC03A1, while the design suite counts with "perfect" rectangle? Or there is a difference because of "theory versus the real thing" reason? Or possibly something else?

I have to design a multiple antennas on a custom PCB and test their functions, so I am trying to clear up the information about EMI and antenna design, before actually designing the prototypes. Thank you so much for help and clarification!

[Henry Crane]

​Hello Mike,

1) you highlited a typo in the document; you are right calculated tuning frequencyof low pass filter is 15MHz.

2) using multiple capacitors has no effect on calculation resut as you pointed out, it offers however more flexibility to reach the desired capacitance value from standard values availbale on the market. Both inductances and capacitances are have standardized values. using two caps in parallel alloweded to reach 200pF while only 220pF is availbale on the market.

3) the correspoinding error is in a range of 11%. EdesignSuite antenna module precision is supposed to be in a range of 5% for rectangular antenna. Round corners reduce a bit the parasitic capacitance of the antenna with the result that imaginary part of impedance is a bit lower. This could be an explanation of difference between calculation and reality. I cannot say that "real life vs simulation effect" does not play a role too ;-). The 6% difference between expectation and measuremet is not an issue since the tuning of the antenna is trimmed with external components.

feel free to ask more if you need,

Best Regards,

Henry Crane RFID/NFC support team. 

[Henry Crane]

​Hello Mike,

a last point I did not mentioned in my respoinse below: eDesignSuite calculates the rectangular coil inductance with no feed line.

Antenna impedance used in CR95HF and ST25R95HF application note is measured close to tuning circuit, at the input of the 5mm feed line. it is also a contributor to difference.

Best Regards,

Henry Crane RFID/NFC support team.

[MikeObs]

Hello,

Thank you for your fast answer! You definitely cleared up some things for me.

I am working on a project where I will be testing multiple antennas in the set environment to find the best working one. I need them to be able to work in Tag detection mode (one we designed before did not work in the Tag detection mode), while providing decent read range. Do you think that I should stick to the inductance around 1uH like the antenna on X-NUCLEO-NFC03A1, or what would be the best inductance value you recommend for the CR95HF IC?

I want to try different shapes and conductor width/spacing, not sure what inductance suits the IC the most because I found different values from 0.5 to 4 uH for multiple different IC's . (Sorry but antenna design is not my strongest skill)

Thank you for your help,

Mike

[MikeObs]

Hi, @Henry Crane_O​ any idea about my questions? Thank you so much

[Henry Crane · Accepted solution]

​Hello Mike,

sorry for the delay in the answer.

focusing on 1µH inductance has the advantage of not changing the values of the tuning circuit components.

Increasing the inductance will lead to very small capacitor values for C2 (capacitance in parallel to the antenna), typically few pF. it requires the use of 1% tolerance capacitor, and maybe you could not be able to tune the antenna. from another side, large inductances offer a wide band matching which is good for communication. High inductance means larger number of turns and a better coupling factor with tag.

decreasing inductance is also feasible (our M24LR-Discovery kit features a CR95HF transceiver board using a 2 turn 424nH antenna). the resulting tuning is narrow band, in combination with a large Q factor: regarding communication, it is is good for ISO15693 communications, but less suitable for ISO14443 communications. the generated field strength is high, thanks to a lower antenna impedance and the narrow band tuning theoretically eases tag detection compared to wide band tuning.

as you can see, the antenna exercise has always to do with trade-off, and 1µH antenna is a good compromise. In case you plan to work only in ISO 15693, you can decrease inductance to lower value.

best regards,

Henry C.