Optimizing current for tag detection (ST25R95)

[Jhern.1]

Hello,

I am trying to optimize the current consumption in tag detection mode for my battery reader (based on ST25R95).

I have followed the application note for keeping current at minimum with tag detection as suggestred below:

https://community.st.com/s/question/0D53W00000BsxIdSAJ/approach-for-low-power-design

1)

In the AN (section 2.4 "Using LFO frequency settings to reduce power consumption") says the the LFO at 4kHz will reduce energy, but I have tested and energy is not reduced.

As far as I know the LFO clk is the same for different frequencies, the change is for the presecaler.

My question: ¿Does a lower LFO freqquency really save energy?

2)

I know the tuning for the antenna is very important for reaching good performances at reading level, but I would like to know how a good tuning impact on the current consumption.

3)

Other option I am assessing is to decrease the OSC and DAC start time (recommended to be 3 ms). How risky is to decrease this time under 3ms?

----

FYI:

I am currently using RF bursts of 2,3us and I am having a consumption of 50uA (as expected). I can not reduce this time because it wouldn´t detect tags otherwise.

POWER at 3,3V (both main and TX).

Any suggestion will be welcome.

Regards, Ramon.

[Brian TIDAL]

Hi Ramon

regarding your first question, I also believe that the value of the LFO frequency has a very limited impact on power consumption in Tag Detection mode configuration

I will let an RF expert answer your second question as this is a complex one.

For question 3, OSC start time depends on the cristal used on your board. I would suggest you measure the oscillation stabilization time and then adjust the OscStart accordingly. For the DAC stabilization time, I have no specific recommandation but one may try some lower value and check whether the tag detection is still functional. I would suggest to first measure the oscillation stabilization time and reduce the OscStart accordingly (with some margin...) and then try step by step to reduce the DAC Start value.

Most of the power consumption occurs during the tag detection bursts (in your case I believe you have SwingCnt=1Fh =>2.3 µs). You can reduce the overall power consumption by increasing the WUPeriod period between consecutive bursts.

Can you share the content of the Idle command sent by your application?

Rgds

BT

[Jhern.1]

Hi Bryan,

Thanks for your superfast answer.

2)Let me know the answer from your RF expert.

3) I am not sure how to measure the oscillation stabilization on my xtal. Should I hook an scope to the xtal and check when the 27MHz signal is stable? Can you give further details aout this?

I include the calibration process and the final idle command for detecting (in bold below). LFO freq used here is 32kHz.

RFID sleeping and detecting...>> 0x 07 0E 03 A2 00 B8 01 18 00 00 60 60 00 00 1F 01

<< 0x 00 01 02 

>> 0x 07 0E 03 A2 00 B8 01 18 00 00 60 60 00 FC 1F 01

<< 0x 00 01 01 

>> 0x 07 0E 03 A2 00 B8 01 18 00 00 60 60 00 7C 1F 01 

<< 0x 00 01 01 

>> 0x 07 0E 03 A2 00 B8 01 18 00 00 60 60 00 3C 1F 01

<< 0x 00 01 01

>> 0x 07 0E 03 A2 00 B8 01 18 00 00 60 60 00 1C 1F 01 

<< 0x 00 01 02 

>> 0x 07 0E 03 A2 00 B8 01 18 00 00 60 60 00 2C 1F 01

<< 0x 00 01 02 

>> 0x 07 0E 03 A2 00 B8 01 18 00 00 60 60 00 34 1F 01

<< 0x 00 01 01 

>> 0x 07 0E 03 A2 00 B8 01 18 00 00 60 60 00 30 1F 01

<< 0x 00 01 01

||DEBUG:||Calibration OK

>> 0x 09 04 68 01 07 10

<< 0x 00 00

>> 0x 09 04 68 01 07 00

<< 0x 00 00

Going to idle with Tag detection. DAC_REF = 2C >> 0x 07 0E 0A 21 00 38 01 18 00 40 60 60 24 34 1F 01

Do not hesitate to ask any further detail.

Regards, Ramon.

[Brian TIDAL · Accepted solution]

HI Ramon,

For the oscillation stabilization time, you can probe XIN/XOUT with a scope: you should see something similar to the yellow part of figure 3 from AN3433. I would use a configuration where the HFO is woken up thanks to IRQ_IN, so that IRQ_IN edge can be used as a starting point for the measurement.

If I am not wrong, in you application the WU Period is around 528ms (this can be easily measured: you just need to probe the tag detection bursts with a scope). I would not recommend to increase it too much in order to keep a good user experience (I believe 700ms is still acceptable but more would lead to a bad user experience, this probably need to be refined). The Idle command for tag detection seems fine to me.

I believe your MCU also enter low power mode and is woken up by the IRQ_OUT from the ST25R95?

Rgds

BT

[Jhern.1]

Hi Brian,

I will be checking this soon again and I would like to know if you manage to get an answer from your RF expert about the impact on current consumption due to detuning (question 2).

Regards, Ramon

[Brian TIDAL]

H iRamon,

here is the answer from RF expert:

Optimizing the the antenna tuning consists in adapting antenna impedance to the reader RF stage output impedance to maximize the power transferred to antenna. Under best tuning condition, current delivered to antenna is maximum, and reader current consumption is maximum. antenna tuning circuit is optimized for power transfer when its input impedance matches the reader RF stage output impedance (27 ohm at VPS_TX=3V). Choosing an input impedance above 27 ohm reduces the current consumption without impacting too much the transferred power (the current decreases and voltage increases). Choosing an input impedance below 27 ohm is not recommended because the transferred power decreases in combination with a high increase of current consumption. A typical input impedance between 27 and 60 ohm is recommended.

Rgds

BT