Other Parts Discussed in Thread: TPS62825
Hi –
We have a design where we are powering an SOC which itself has 5 voltage rails:
- 1.8V AON (Always On)
- 3.3V
- 1.8V
- 0.8V
- 0.65V
At times, we may decide to disable #’s (2)-(5) to save power while leaving the 1.8V AON supply up. Our SOC has a PWRGOOD input in its Always-ON (AON) section that will be used to determine when the SOC can be safely released from RESET after all the supplies are back up.
For the 5 supplies, we are using the TLV62084A. I’ve attached a simplified sketch of how these are wired in our schematic. Though it seems a bit more complicated than need be, there are other parts of the system on the board that require this complexity. The basic function of the supply is such that the 5V supply and the 1.8V Always On supplies will be powered up all the time. But, the other downstream supplies will be enabled/disabled depending on demands of the SOC.
We want to make sure we are using the PG pins properly in relationship to one another and our SOC. In the datasheet, it states:
“The PG terminal is an open drain output and is specified to sink typically up to 0.5 mA. The power good output requires a pull-up resistor which is recommended connecting to the device output.”
Because we have 4 of the supplies that are effectively AND-ing their PG outputs together to feed our SOC’s PWRGOOD input, we’re concerned by this recommendation. Our intention, as seen in the diagram, is to add pull-ups to each of the PG outputs to the 1.8V Always-ON rail. Because the drive capability of the PG output is 0.5 mA, we selected a 178 kΩ resistor for each. And because of the guidance at this video: https://training.ti.com/combining-power-good-signals, we elected to have a physical AND gate tying them together.
Because all of these supplies are having their PG outputs AND’d together, we need to keep the PG logic-high output at the same level as the AND input and our SOC. Therefore, we are not following the recommendation of tying the power-good output pull-up resistor to the device output. Instead, we are tying the pull-ups to a common 1.8V supply and utilizing the open-drain aspect of the PG output.
Could you please review the diagram and let us know if this is wired properly and if you see any concerns with the PG output logic?
Are there any other provisions that would help make this solution more robust?
Thanks in advance!
Tom