• Resolved

LSF0204: Is sensitive to..?

Prodigy 30 points

Replies: 5

Views: 54

Part Number: LSF0204

Good day,

I have two LSF0204 in my design. They are used to translate UART and I2C levels: from 5v (MCU) to 3.3v (sensors/other MCU).

I'm having a problem that they stop functioning. I ordered 2 spares but they have also now died.

The devices are solder on to a PCB with lots of other ICs and it's only the LSF0204 that keeps breaking down. 

My question: Is there a specific design criteria that I have missed?

Best Regards

Niklas Löfgren

  • Hi Niklas,

    UART translation can use LVC2T45(or TXB0104) as they are push pull bidirectional signals. 

    https://training.ti.com/tlm-voltage-level-translation-uart-interface?cu=1134826

    I2C can use LSF as they are open drain bidirectional devices. 

    Refer to the LSF videos on its functioning:

    https://training.ti.com/introduction-voltage-level-translation-lsf-family

    Please send out the schematics and what do you mean by 'stop functioning' and 'has died'? Any scopeshots will be helpful on the Vcc and IO ports. 

    -Thanks,

    Shreyas

    Watch the Introduction to AXC Family

    Read the app note on Glitch-free Power Sequencing for the AXC family

    Evaluate the AXC8T245 EVM and watch the translation using the EVM

  • In reply to ShreyasRao:

    Thanks for your fast reply, that is much appreciated.

    The schematics is exactly as figure 6 in the data sheet. Rpu = 1kohm on both sides, located less than 5mm away from unit. PCB track from MCU to LSF is around 25mm and from LSF to sensor it's around 1200mm (1.2m). The second LSF have only 20mm from unit to sensor(located on the same board).

    Table 2 in the same data sheet indicates that UART is possible with LSF0204. Is that wrong?

    Please see attached scope for 'broken' behaviour. Blue is MCU(5v) side, Yellow is sensor(3.3v) side. The unit is not able to hold the signal low.
    They where perfect copy of each other when the unit was working.

  • In reply to user5820644:

    If I had to guess, I'd suspect that some overvoltage or overcurrent broke the device.

    But we don't know where that might come from. Please show the actual schematic.

  • Problem solved.

    A small background if someone googles this in the future. I have two LSF0204 in my design: one was never working and the other was working perfectly for around one week before it died.
    After replacing the unit with a new one it was working but around one week later died again. Due this we thought there was a problem with the LSF, it was not.

    The actual problem was a floating enable pin (neither connected to ground or Vrev_A) which caused this.

  • In reply to user5820644:

    Excellent, thank you for your valuable feedback. I beleive it will be helpful for the community.

    Yes, the active CMOS inputs must never be floating due to this reason. 

    -Thanks,

    Shreyas

    Watch the Introduction to AXC Family

    Read the app note on Glitch-free Power Sequencing for the AXC family

    Evaluate the AXC8T245 EVM and watch the translation using the EVM