TPS65987D: Prevent sink on power up

Hi Sam,

What are you ADCIN1/2 pin configurations? It may be a default configuration its going into after the power off.

Thank you,

Hari

Hari,

Boot mode = Infinite Wait

ADCIN1 = 1.095V  DIV = .33, SPI_MISO = 1

Our tablet is powering our board.  You would think it would not accept power.  Any idea how to keep it off?

Thanks.

Hi Sam,

Please refer to Table 6 of the datasheet to see the other options for dead battery mode. In your current configuration, it looks like the power path is enabled to sink power. You may need to switch the setting to BP_NoResponse or BP_NoWait to disable the power path.

Thank you,

Hari

Hari,
I had originally tried ADCIN1 = gnd, MISO = 1. Table 5 says BP_NoResponse and Safe Configuration. However, the board is still powered by the tablet.
Any other ideas?
Thanks.

Hi Sam,

I may need to see your board schematic to see if anything else could be causing this. Could you upload it here?

Thank you,

Hari

Hari,

It appears it is recognizing there is no board power, and is setting Dead Battery flag.

Based on ADCIN1 = 0V and MISO = 1, that should be BP_NoResponse.  V3P3 is derived from V5P0.

However, the PP1 switch is enabled, which clearly it should not be.

Here’s what the hardware seems to be doing:

Ch1 yellow = 5V

Ch2 cyan = 3.3V

Ch3 purple = CC2

Ch4 green = Vbus

Why might PP1 be enabled (and the board powered), when it the configuration says disabled?

Thanks.

Hi Sam,

I'm thinking since this setting also corresponds to the Safe Default Configuration which shows that the device will operate as a Legacy sink, this is what turns on the power path. Could you try changing it to use one of the other default configurations by changing MISO = 0?

Thank you,

Hari

Hari,

It appears BP_NoResponse is required to do nothing during dead battery mode.  And, Infinite Wait is required to do nothing when normally powered.  There appears to be no configuration that does both.  So, I’m not sure how to implement your suggestion.

What should a BP_NoResponse and Reserved configuration do?

Why is BP_NoResponse enabling a power switch, when it says it does not?

Also, I floated PP cable, with no help.

Thanks.

Hi Sam,

The Configuration 1 that I mentioned would only advertise sourcing capability so you would just need to GND the SPI MISO pin, instead of tying it to V3P3 like it is now. The BP_NoResponse won't enable the power path only until the VIN_3V3 is on, which from your scope capture seems to be turning on. It looks like it initially goes to about 500mV, I'm not sure why that is. Could you tell me what it's connected to? 

There could be other circuitry that could be interfering with the power signals as well which could cause issues. I will need to discuss with my team to see if there's anything missing. Have you tried testing on the EVM? I would also recommend referencing the Dead Battery Mode Application Note in the meantime.

Thank you,

Hari

Hari,

Configuration:  MISO = 0, ADCIN1 = 0  undefined state

Close to BP_NoResponse, Configuration 1, which starts at ADCIN1 = .1

Didn’t try this.

Configuration:  MISO = 0, ADCIN1 = .1  BP_NoResponse, Configuration 1

Does not ‘dead battery’ power device.  Tablet states that it is charging the device.

Does not power normally either.  App Customization Tool recognizes the device, but nothing can be changed, ever.

Configuration:  MISO = 1, ADCIN1 = .5  BP_ECWait_External, Infinite Wait

Does not ‘dead battery’ power device.  Tablet states that it is charging the device.

Does not power normally either.  App Customization Tool recognizes the device, but nothing can be changed, ever.

No, we do not have the EVM.  This seemed like a simple application…  We are running your App Customization Tool with the Aardvark I2C interface.

A TPS62742DSSR generates 3.3V from 5V. 

Thanks.

Hi Sam,

I will discuss with my team to see if anything else could be missing. Do you happen to be using VBUS to power the TPS62742 that provides VIN3.3 to the PD controller?

Thank you,

Hari

Hari,

The V3P3 buck converter is powered from V5P0 (which is also PP_HV1, PP_HV2, and PP_CABLE), not VBUS.

Thanks.

Hari,

So, I acquired an EVM.  The syspwr, uB bus, and var DCDC leds are on.  It seems to be able to sink.  I reinstalled the gui to ver 6.1.2.  It seems to see the FTDI.  However, it does not see the device.  Doing a sweep address range, it finds nothing, “I2C Address Sweep failed read for all addresses in range 0x0..0x7F”. 

SPI_CLK is active, and I2C_SCL is silent (SCL and SDA both go low, but no activity).  S1, 1-6 all on.  The same happens with only S1 5-6 on.

Any idea why the I2C is silent?

Thanks.

Hari,

Thanks for confirming what seemed to be happening. 

Do we know which power path will be enabled in dead battery mode?  If so, maybe we can float that path.  Otherwise, the TPD6S300 would appear to be an option of last resort.

A different port fixed the FTDI driver issue.

Thanks.