• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 34 subscribers
  • Views 144 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TXB0304: There is an issue that uses the TXB0304RSVR as a level shifter to convert SPI from 3.3V standby power to SPI 3.3V main power.

Cho Lena
Prodigy 50 points
Part Number: TXB0304
Other Parts Discussed in Thread: TMUX1511, SN74CBTLV3126, SN74LV125A

Tool/software:

Hi, Sir,

We have a motherboard (CPU: Intel Core i5-1335UE) that uses the TXB0304RSVR as a level shifter to convert SPI from 3.3V standby power (3.3VSB) to SPI 3.3V main power (3.3VCC).

However, we’ve encountered an issue with this configuration.

Schematic:

When the ATX power pulls PSON# low to turn on the system, the TXB0304RSVR's OE pin is high, which causes interference to the SPI source (3.3VSB). As a result, the BIOS flash is affected and the system fails to boot.
In our experiment, if we connect both the VCCA and OE pins to 3.3VSB (Circuit shown below), the system is able to boot normally.
(Note: During the failure, the connector had no device attached — meaning the A-side SPI lines are floating.)

Questions:

  1. Is it appropriate to use the TXB0304RSVR for translating SPI from 3.3V standby power (3.3VSB) to SPI 3.3V main power (3.3VCC)?
  2. Does the above schematic (first schematic) have any design issues?
  3. Since this is a same-voltage translation, which side (A or B) is recommended to be connected to the SPI source? (In our current design, the source is connected to the B-side.)

Thank you for your support.

  • 0
    Guru 316690 points

    You do not need a level shifter when you do not want to shift levels. And the auto-bidirectional architecture of the TXB (which never allows an I/O pin to float but forces it to either high or low) is likely to interfere with other signals on the bus when OE is high during power on.

    Please specify exactly how the MISO signal should behave. If it is shared between multiple SPI devices, then it must be disabled when the SPI slave is not being accessed. So you need a three-state buffer like the SN74LV125A (controlled with /CS), or an analog switch like the SN74CBTLV3126/TMUX1511.

  • 0 in reply to Clemens Ladisch
    Prodigy 50 points

    The connector is connected to an external TPM module, and the TPM module's power is supplied by 3.3VCC. If the power of SPI source is 3.3VSB and is not converted to 3.3VCC, the TPM may experience leakage current, which causes the system to fail to boot. Therefore, isolation of SPI signals is required.
    Our application is that MISO signal is shared between multiple SPI devices. I will check datasheets of analog switches. Thank you for the suggestion.

    Additionally, I would like to ask:
    Since a customer want to connect an external SPI module from a connector of board, we are concerned that the extended SPI trace length may cause signal attenuation, affecting signal quality.
    We are considering using the following circuit to strengthen the SPI signals in our board:
    Circuit:

    Would this configuration be suitable?
    If not, do you have any recommended ICs for this purpose?

    Thanks.

  • +1 in reply to Cho Lena
    Guru 316690 points

    The TXB has a drive strength of 20 µA; it will not strengthen but weaken the signals.

    An analog switch will not buffer the signals. Use a buffer like the SN74LV125A (powered by either 3.3VCC or 3.3VSB); it will isolate the two sides, even when powered off. The MISO signal must be disabled with the /CS2 signal. (Any input pin that might be disconnected must have a pull-up or -down resistor to prevent it from floating.)