TXS0104E: I2C RTC hang issue in RF immunity with use of TXS104E

Jagdish Karamchandani

Part Number: TXS0104E
Other Parts Discussed in Thread: TXU0102, LSF0102

Hello TI support,

We are facing strange issue of RTC getting stuck during RF immunity testing with 20V/m field strength. We have some other product without use of level translator having same RTC (as attached in schematic) which doesn't face any issue upto 30V/m. Please advise what could go wrong with level translator & are there any suggestions to be tried when we go to lab. Just FYI, we already tried below methods but didn't succeed,

1. Adding 22pF/100pF capacitors at RTC side on both I2C lines which show DNI in schematic.

2. Decreasing value of external pull-up resistors from 10K to 3.3K

3. Adding 100pF+SP4021 diode in supplies of RTC.

4. Adding 22pF capacitor in supplies of level translator.

5. Adding 22pF/100pF capacitor in 1.8V supplies of BG95 module.

Awaiting your response at earliest. CLK frequency is 100KHz.

Regards,

jagdish

over 1 year ago

0 Clemens Ladisch over 1 year ago

Guru 317370 points

The TXS has edge accelerators that trigger when it detects what looks like the beginning of a rising edge. This can be sensitive to noise.

For translating I²C, I would use the LSF0102, and for the SPI signals, the TXU0102.

0 Jagdish Karamchandani over 1 year ago

Expert 1420 points

Hello Clemens,

Got it, do we have any test report for TXS family for RF immunity to know the levels it could sustain? We would access our product for those levels & see its acceptance by customer, if not then will spin the PCB with your suggested ICs.

Regards,

jagdish

0 Clemens Ladisch over 1 year ago in reply to Jagdish Karamchandani

Guru 317370 points

I've heard the threshold is about 30 % of V_CC. The actual RF immunity depends on how good an antenna the SCL trace is.

0 Jagdish Karamchandani over 1 year ago

Expert 1420 points

The SCL trace length is even smaller than the SCL trace in product passing 30V/m. Apart in whole circuit there are various peripherals with comparable trace length with non failing in RF immunity even at 30V/m, meaning at 20V/m & 30V/m RTC hangs but other peripherals are functional. What I meant was any bare testing carried on the IC as part of reliability testing?

0 Clemens Ladisch over 1 year ago in reply to Jagdish Karamchandani

Guru 317370 points

RF testing can be done only on a complete board (a bare pin might be able to pick up noise, but this is insignificant compared to a trace.)

You already know that you board has too much EMI. I do not think there is any simple solution.

0 Jack Guan over 1 year ago in reply to Clemens Ladisch

TI__Mastermind 41216 points

Hey Jagdish,

Unfortunately, RF Immunity was not something we categorize at the time of designing/ testing this device.

Regards,

Jack

+1 Jagdish Karamchandani over 1 year ago in reply to Jack Guan

Expert 1420 points

Hello Jack,

Thanks for this clarification.

Regards,

jagdish

Logic

Logic forum

TXS0104E: I2C RTC hang issue in RF immunity with use of TXS104E