TXB0304: The power consumption of TXB0304

Sam Huang

Part Number: TXB0304

Hi,

I encounter a issue about current, the schematic is as follows,

Story:

Step 1. When the voltage of VDD_EXT changes from 1.8V to 0V, the current of TXBN0304 will be as high as 400uA(20pcs out of 100pcs chips will have this issue).

Step 2. I release the voltage of 3V3 to 0V and power it on again after 1s, the current of TXBN0304 drops to a few microamperes.

In addition, I saw that the control flow of OE was marked in the specification.

TXBN0304RUTR requires that OE should be tied to GND to maintain high-impedance state during power on and power off sequences.
If this is not observed backflow through the device can occur resulting potential damage.

Do you think these defective chips may be caused by the above situation?

However, I had tried to follow this workflow, keeping OE low during power on and off, but can not solved the issue of sleep current.

But I can solve this issue when I follow the step 2(release the voltage of 3V3 to 0V and power it on again after 1s).

Could you please help to explain?

Thanks.

3 months ago

0 Clemens Ladisch 3 months ago

Guru 261610 points

There is no backflow through OE. But if VCCB is too low, a current can flow into VCCA and out of VCCB; see Biasing Requirements for TXS, TXB, and LSF Auto-Bidirectional Translators.

0 Sam Huang 3 months ago in reply to Clemens Ladisch

Prodigy 10 points

Hi Clemens,

Thank you for your answer. I can understand that the voltage of VCCA should be less than VCCB, but I still have a little confusion.

1. Why does the current consumption issue not occur when I release the power supply of VCCA for 1s and then supply 3.3V again?

2. Why is this issue not 100%, the defective rate is 20%.

Thanks.

0 Clemens Ladisch 3 months ago in reply to Sam Huang

Guru 261610 points

Are you actually forcing the supply to 0 V? If not, then the decoupling capacitors will keep it at some higher voltage.

0 Sam Huang 3 months ago in reply to Clemens Ladisch

Prodigy 10 points

Hi Clemens,

Yes, I am sure to force V_CCA to 0V.

Test method:

1) Cut off the line of V_CCA and connect it with two wires, so that V_CCA and 3.3V power supply can be disconnected or connected at any time.

2) When the power supply of V_CCB is from 1.8V to 0V, I will disconnect V_CCA from 3.3V and then connect it again after 1~2 second.

3) At this time, it was found that the current of TXBN0304 got smaller.

Thanks.

+1 Jack Guan 3 months ago in reply to Sam Huang

TI__Mastermind 20941 points

Hi Sam,

This seems to be a known issue with the TXB0304 that has been documented in the past with increased leakage current seen during partial power down mode. There is a possibility that an internal node becomes floating when one of the power supplies are powered down, resulting in a slight supply increase. To minimize the leakage current seen, it is suggested to power down both power supplies.

Regards,

Jack

0 Sam Huang 3 months ago in reply to Jack Guan

Prodigy 10 points

Ok, thanks for your reply.

Logic

Logic forum

TXB0304: The power consumption of TXB0304