TPS65950/30/20: VINTANA2 pad connection and power saving

Currently one of customer not turning off USB3V1. Reason might be this comment in TRM: "During the WAIT-ON power state (the VUSB3V1 regulator is off), nonzero dc voltage must not be applied on pins CTSI/ADCIN3, TXAF/ADCIN4, RXAF/ADCIN6, RTSO/ADCIN5, DP/UART3.RXD, and DN/UART3.TXD. This is to avoid damage to the TWL5030 caused by injection of large drain-bulk current on the speaker amp output stage (for the DP/UART3.RXD, DN/UART3.TXD, and RXAF pins) and on switches connected to the four MCPC pins (CTSI/ADCIN3, TXAF/ADCIN4, RXAF/ADCIN6, and RTSO/ADCIN5)."
AD readings in ADCIN4 become 0 if 3V1 is disabled.
The warning above says 3V1 should not be turned off to avoid damage.
Kindly, provide info on above CAUTION  mentioned in TWL5030 TRM section 15.4.11.3.

Hi,

     In the Wait-On state ( OFF State), AD conversions are not possible. Also, Audio and USB functionalities are not available. For all these to be possible, we must Switch On and be in ACTIVE state.

In the ACTIVE state all the above functionalities are possible and also USB3V1 is ON. Hence no issue. The warning actually does not say that 3V1 should be OFF to avoid damage. It says that Non-zero voltages should not be applied on the pins because 3V1 is OFF in Wait-ON state and can damage the chip.

Please let me know if I was clear. Kindly provide more details of the use-case used by customer. We woud  like to know why the customer wants to apply non-zero voltages on these pins in WAIT-ON state.

Kartik

Thanks for quick reply.

Ok. But does that mean if USB3V1 is OFF in ACTIVE state, we can still apply Non-zero voltages to those pins ?

Use case like, Not in WAIT-ON state. Gaia is in ACTIVE state, but the USB3V1 resource is off.

Regards,

Sumeet

Thanks for the clear use-case. Now i get your point. No it does not mean we can apply non-zero voltages if USB3V1 is OFF.

To avoid the leakage ( and damage ), we must enable USB3V1 resource. Hence, if non-zero voltages need to be applied to these pins, we must have it ON. The note ( caution) is just to indicate that by default, when GAIA goes to WAIT-ON from ACTIVE , USB3V1 turns OFF and in this case, non-zero voltages should not be applied.

Also, by default, when GAIA is in ACTIVE, USB3V1 will turn ON in sleep mode consuming less current.

To summarize, irrespective of the state GAIA is in, USB3V1 must be ON if we need to apply non-zero voltages on those pins.

A slight modification to the very first note:

If VINTANA2 is switched OFF then the consumption will not reduce by 1mA. This value really depends on what is active on the device. Since the reduction isnt 1mA it is recommended that you put the resource in sleep mode. In this mode the consumption/leakage will be the ground current.

Do I need to follow specific sequence to put these resources in a sleep using a broadcast message IN TritonT2 ? like as mentioned above for VINTANA2.
Resources are VAUX1, VINTANA1, VINTANA2, VINTDIG

No, there is no specific sequence needed. For broadcast message you will have to group your resources accordingly.

In case you are using the NSLEEP signals for this then you can refer the TRM for Active to Sleep sequence.

In terms of sequence of power off VINTANA2 and VUSB3V1, if I use broadcast messgae to turn of both of them for either S2A and A2S sequences, what would be the side effects?  Current leakage?

 For another case that I need to put  VINTANA2 in active mode (for audio application) and VUSB3V1 in off mode (to save power) during GAIA sleep, is there any side effect?

Antony

If broadcast is used to turn off both of them, then there would not be any side effect.

For the second case, there is NO side effect if  VUSB3V1 is switched OFF and VINTANA2 is ACTIVE but since you have mentioned "Audio application", you must take care to not apply non-zero dc voltage on the pins: CTSI/ADCIN3, TXAF/ADCIN4, RXAF/ADCIN6, RTSO/ADCIN5, DP/UART3.RXD, and DN/UART3.TXD. 

Kindly go throught the  "CAUTION"  mentioned in TWL5030 TRM section 15.4.11.3.

Kartik