TPS65218D0: TPS65218D0 off-state problem

Part Number: TPS65218D0 Hello, my customer is using PMIC TPS65218 in their application and it was running well. He switched to the new TPS65218D0 now and got into a problem which stopped the production! The problem is that in a specific configuration when the application is switched off still a parasitic current occurs on the BIAS and DCDC input terminals of the PMIC (see attachment). Now the D0 PMIC is going into the OFF state after a period of 1-20s and states there, no chance leave this state. Only the push button or a reset helps. The non-D0 version wasn't going into this OFF state. - Is there a difference in the D0 and non-D0 version which leads to this problem? - To ensure that the PMIC comes out of the off-state the customer has to make sure that DCDC5 and DCDC6 are also active in the off state. To do so the FSEAL- bit has to be configured. Is this possible withouth a Coil Cell Supply or is the bit reseted in the off-state. Regards, Holger

Holger, Sorry, but what am I looking at in this scope shot? What pin are you probing and when are you probing it? I don't really have a point of reference here. It would be nice to see PWR_EN pin and at least one DCDCx output voltage in the same capture. If you are saying keeping DCDC5/6 on helps, it would also be helpful to show DCDC5 or DCDC6 (a) when you are experiencing the issue, and (b) when there is no issue. Please use all 4 channels on the scope to provide some context. Leaving the OFF state depends on 4 events: VIN_BIAS > (VUVLO + hysteresis) & (PB low || AC_DET low || PWR_EN = high) In order to determine if the issue you are seeing is expected behavior, I need to know the state of these 4 pins at the time when you expect the PMIC to turn back on.

Holger, To answer your questions: HS WG-FAE - Is there a difference in the D0 and non-D0 version which leads to this problem? No, feel free to review the TPS65218 (-B1) to TPS65218D0 changes, attached here and also in our FAQ section. 2577.TPS65218xx_Rev_history.pdf HS WG-FAE Is this possible withouth a Coil Cell Supply or is the bit reseted in the off-state. If there is no voltage present on the CC pin, it will reset to 0b automatically after you set to 1b. However, a physical coin cell is not required. You can also use a resistor divider to generate 3V nominal at the CC pin from the main input voltage. 3V = 5V*R2/(R1 + R2) --> R2 = 100k, R1 = 67k