ST25R3911B matching impedance target

[RPalm.3]

In the app note an4974, ST recommends a matching impedance of around 16-20 Ohms, this part is reasonable and makes complete sense.

In my configuration, I'm using differential with AAT. According to the app note, the target impedance should be at the bottom of the loop, near the real axis (again, this part makes sense) but it is not at the point closest to the real axis (which has the minimum imaginary component).

I have seen various matching recommendations, mostly with the loop either centered about the real axis, or the loop below the real axis and the match point at the point at the left most crossing of the real axis. This is because as you present a tag, the match point will shift clockwise. If you're at the center of the loop, then that shift will reduce your impedance. If the loop is below the real axis and you've matched at the left crossing, then a clockwise shift will increase your impedance as you move a card closer to the antenna.

Why does ST recommend putting the loop ABOVE the real axis and setting the match point to a point near the bottom of the loop. Also, I assume this match point is WITHOUT the AAT impedance included (since it's measured with the chip off and therefore the AAT is out of circuit)

Putting a card in the field moves the match point clockwise, which reduces impedance. Adding capacitance via the AAT does the same thing. This seems to be counter productive.

Thank you,

Robert P

[Travis Palmer]

Hello Robert,

The ST25R3911B offers one HW based algorithm to optimize for the maximum amplitude measured on the antenna (which will measured via the capacitive voltage divider on the RFI pins).

Also the internal measurement commands (measure amplitude, measure phase) can be used to do a specific algorithm fitting your needs.

In general there are two use-cases for AAT.

1.) Trim the antenna tuning during production. The AAT can be used to compensate component and production tolerances. The obtained value can be stored and used for the lifetime of the device.

2.) Continues adjust the antenna. Here the idea is to tune the antenna more often. Either every time during power on (as it is done for the ST25R3911B-DISCO - besides the option to manually trigger AAT) or you can select "Tune antenna when there are no tags" in the "Polling" tab of the GUI.

AAT should be used to optimize the reader if no tag is present. In this condition we want to optimize the output power to be able to read the tag as far as possible. It is not required to re-tune the antenna if a tag is in close coupling condition. Here the power transfer is anyhow good enough to read the tag.

Why is it not required to adjust the antenna during close coupling condition with a tag? A perfectly tuned reader without AAT would suffer from the same detuning and still can read the card.

Please also have a look at AN4914.

BR Travis

[RPalm.3]

Travis,

Thank you for your response. First off, my initial question was a little broad, and you did a good job addressing it. You did clarify that the AAT is really designed for environment and/or drift, not for tag detuning affects.

A more precise question would be:

• Why does ST recommend the loop be fully above the real axis of the smith chart AND have the match point at the lower left part of the loop (about 7 o'clock position).

I'm used to seeing the recommended matching be with the loop either centered on the real axis or below the real axis.

With the loop above the real axis and the match point (13.56Mhz) at the lower left, detuning from a tag will push the 13.56mhz point further clockwise, detuning the antenna to a much lower impedance and driving up the current.

Why isn't the recommended match point at the point on the loop closest to the center of the smith chart (the lowest point in the Return loss dip curve)?

Why isn't the loop recommended to be on or below the real axis?

[RPalm.3]

Follow-up...

By the way, the AppNote AN4914, for trim value 0 (no added capacitance) the 13.56Mhz point is where I would normally target the optimal match. The recommended matching corresponds to Trim value 7 (which matches what I test on the DISCO board) - this allows you to add OR remove some capacitance when the mounting environment adds capacitance or inductance, but the goal is always to get the 13.56Mhz point at the approximate location shown by Trim value 7 in the AppNote.

[Travis Palmer · Accepted solution]

Hello Robert,

In general we should try to have a matching close to the real axes. This means that the current and voltage in the MOS transistor of our reader is in phase.

The matching impedance is our main parameter to adjust the power consumption and power delivered to the matching network. In some cases you see only a little difference in reading distance between Z=15Ohm and Z=20Ohm. For a normal reader (e.g. access control) it may not be required to tune to Z=15Ohm and drive the maximum allowed driver current.

Other applications like EMVCo or reader exposed to metallic environment (center console readers in cars) may require to squeeze out the last mW of power to achieve the read range requirement or cover a volume without reading holes.

Using the tuning showed on our discovery boards, it is much easier to fulfill both use-cases.

Looking at the 3916(B), here it is much easier to tune the reader to Z~8.5Ohm.

Component and manufacturing tolerances are then adding an additional level of complexity.

Having a very small resonance loop (in the smith chart) with all these variations and achieving a constant performance across different devices may become complicated.

Please let me know, if this answers your questions.

br Travis