TPS65982: TPS65982 negotiation problem

Hariprasad Bhat

Part Number: TPS65982

Hello,

We are using TPS65982 in one of our custom design. We generated firmware and updated to flash. We are seeing some problem in one board as the negotiation is continuously happening.

I have used Type-C to power the PD under dead battery (Vin_3V3 is not valid). Also, I checked HRESET and MRESET pins for any toggling. There is no toggling @ those pins.

I have used totalphase PD analyzer to capture the logs and I attached the PD controller logs of working and non working platforms.

Is it possible to check the attached logs and suggest what could have gone bad in the non working platform?

/cfs-file/__key/communityserver-discussions-components-files/138/2068.PD_5F00_Debug.zip

With Regards,

Hariprasad Bhat

over 7 years ago

0 Eric Beljaars over 7 years ago

TI__Mastermind 36751 points

Hi Hariprasad,

It appears that the device is sending a soft reset in the PD logs. In the TPS6598x Utilities GUI, can you read back register (0x40) PD Status?
There you can read back the SoftResetType bits and see why the device is sending a soft reset. From there we can begin debugging that issue.

Thank you,
Eric

0 Hariprasad Bhat over 7 years ago in reply to Eric Beljaars

Intellectual 630 points

Hi Eric,

Thanks for quick response.

I am seeing 3 different status while reading through Aardvark. Hence, attached 3 different snapshots while reading the PD_status.

Also, if I probe VBUS, then I see 3 VBUS levels, 0V, 5V and 12V and again 0V, 5V and 12V....

With Regards,

Hariprasad Bhat

0 Eric Beljaars over 7 years ago in reply to Hariprasad Bhat

TI__Mastermind 36751 points

Hi Hariprasad,

Could you potentially take a scope capture of the CC lines and VBUS during this failure?
The Soft Reset message that you are seeing might be related to the rise time of VBUS from 5V to 12V.

Thank you,
Eric

0 Hariprasad Bhat over 7 years ago in reply to Eric Beljaars

Intellectual 630 points

Hello Eric,

I have captured the plots of VBUS and CC lines. Working_PD_VBUS_0V_5V_12V.png is the capture of working PD. CC_VBUS_Plots.zip is the collection of plots at different points as mentioned by file name of the snapshot.

/cfs-file/__key/communityserver-discussions-components-files/138/6278.CC_5F00_VBUS_5F00_Plots.zip

One more observation regarding the IC temperature @ 27C ambient. Working PD die temperature is 30C when powered through Type- C cable. Non working PD die temperature is ~40C when powered through Type- C cable.

With Regards,

Hariprasad Bhat

0 Eric Beljaars over 7 years ago in reply to Hariprasad Bhat

TI__Mastermind 36751 points

Hi Hariprasad,

What are you connecting to the Type-C port when this error occurs?
Is it possible that you are seeing an over-current trip?

Thank you,
Eric

0 Hariprasad Bhat over 7 years ago in reply to Eric Beljaars

Intellectual 630 points

Hi Eric,

I have a test PC with thunderbolt connector which supports Type-C USB, Displayport and PD functionality. I am using same port for both failing and working boards.

How can I know whether over current trip has occurred ?

With Regards,
Hariprasad Bhat

0 Eric Beljaars over 7 years ago in reply to Hariprasad Bhat

TI__Mastermind 36751 points

Hi Hariprasad,

Could you possibly try to read register 0x40 again?
The values that you saw for SoftResetType are not valid values and I believe it has something to do with the device glitching upon the reset.
I would recommend plugging the device in, seeing the reset occur. Then disconnecting your test PC and then reading register 0x40. You should see a value between 0 and 24 for the SoftResetType.

Thank you,
Eric

0 Hariprasad Bhat over 6 years ago in reply to Eric Beljaars

Intellectual 630 points

Hi Eric,

If I power ON my entire board(VIN_3V3 is present) and try to check the VBUS, then it is stable 12V.

VBUS toggling is observed only if I power the PD through Type-C only.

So, with powering on my entire board, VBUS will not toggle and PD Status says no soft/hard resets.

Observation with Powering on the entire board:

In working boards (4 boards) I am seeing the die temperature ~60C @ 27 ambient, while with the faulty board I am seeing the PD temperature is ~85C.

(Previously mentioned Thermal observations are with powering from Type-C cable only, not entire board)

Should I go for replacing the PD controller?

With Regards,

Hariprasad Bhat

0 Eric Beljaars over 6 years ago in reply to Hariprasad Bhat

TI__Mastermind 36751 points

Hi Hariprasad,

I have a few follow up questions:
Do you have any other HW changes between the working and non-working boards?
Do all boards contain the same FW configuration?
If so, could I get your PJT file to review your configuration?

Thank you,
Eric

0 Hariprasad Bhat over 6 years ago in reply to Eric Beljaars

Intellectual 630 points

Hi Eric,

Please find my comments below.

Do you have any other HW changes between the working and non-working boards?

<Comment> No hardware changes.

Do all boards contain the same FW configuration?

<Comment> Same Firmware.

If so, could I get your PJT file to review your configuration?

<Comment> Attaching here

/cfs-file/__key/communityserver-discussions-components-files/138/4428.SEDIM_5F00_CARD_5F00_SNK_5F00_EXT_5F00_PATH.pjt

With regards,

Hariprasad Bhat

0 Eric Beljaars over 6 years ago in reply to Hariprasad Bhat

TI__Mastermind 36751 points

Hi Hariprasad,

I noticed that you have 2 configuration sets with different Sink PDOs and also "Shared Device Settings" with only one Sink PDO.
I would suggest adding the 12V sink PDO in the "Shared Device Settings" as maybe you are loading the incorrect configuration in the failed cases.

Thank you,
Eric

0 Hariprasad Bhat over 6 years ago in reply to Eric Beljaars

Intellectual 630 points

Hi Eric,

I will update the 12V in the shared device setting and share the observation with you by tomorrow.

I continued the debugging with faulty board and observed followings.

VBUS toggling is still present if I connect Type-C only(VIN_3V3 not valid)
Previously I observed that with Powering On the entire board, the functionality was not proper as compared to working boards (Alt mode entry etc)
Today I checked again with Powering On the board to test the functionalities. Now, I'm able to enter into Alt mode and all. VBUS stays @ stable 12V .
But, what concerns me is the PD die temperature which is almost 25C more (~85 @27C ambient) than the working boards (~60C @ 27C ambient).

Is there anything you can help me to debug this Die temperature issue?

With Regards,

Hariprasad Bhat

0 Eric Beljaars over 6 years ago in reply to Hariprasad Bhat

TI__Mastermind 36751 points

Hi Hariprasad,

The PD die temperature should not vary that much from board to board. Do you have an EVM available to test this image on to see if you see similar results?

Thank you,
Eric

0 Hariprasad Bhat over 6 years ago in reply to Hariprasad Bhat

Intellectual 630 points

Hi Eric,

I tested with firmware configured with 5V and 12V in the shared device setting.

Same result as before.

With Regards,

Hariprasad Bhat

0 Hariprasad Bhat over 6 years ago in reply to Eric Beljaars

Intellectual 630 points

Hi Eric,

As I mentioned earlier if Powering the PD only with Type-C port
"
One more observation regarding the IC temperature @ 27C ambient. Working PD die temperature is 30C when powered through Type- C cable. Non working PD die temperature is ~40C when powered through Type- C cable.
"
Due to the component placement and routing in the custom board, with Powering the entire board with main supply, we observe 50C to 60C in working boards, but in non working board PD die temperature will be ~85C.

With Regards,
Hariprasad Bhat

0 Hariprasad Bhat over 6 years ago in reply to Hariprasad Bhat

Intellectual 630 points

Hi Eric,

New Observation :

With the binary generated from attached .pjt, I am not seeing the toggling. But Voltage stays @5V which is ok for us.

/cfs-file/__key/communityserver-discussions-components-files/138/5305.1258.SEDIM_5F00_CARD_5F00_SNK_5F00_EXT_5F00_PATH.pjt

With Regards,

Hariprasad Bhat

0 Eric Beljaars over 6 years ago in reply to Hariprasad Bhat

TI__Mastermind 36751 points

Hi Hariprasad,

The PD controller should not have such a large variant in operating temperatures from board to board. Are there any other variables that may cause one board to heat up much more than the other? The PD controller could potentially be going into Over thermal protection.

Thank you,
Eric

0 Hariprasad Bhat over 6 years ago in reply to Eric Beljaars

Intellectual 630 points

Hi Eric,

I was looking into other parameters for overheating. So, I planned to remove all the discrete components and then check. Now there is no discrete components except input decoupling, still the IC is overheating. Now, I will replace the PD IC and remount all the discrete components with new parts and check.

At the same time I observed following issues in one more board. Can you check with the factory regarding the same?

HRESET pin resistance to ground is around 26-ohm (Some K-ohm in working boards)
GPIOs mapped to AMSEL and "Either USB or DP" Events stooped responding. [These events switch as expected if we map to some other GPIO]

With Regards,

Hariprasad Bhat

0 Eric Beljaars over 6 years ago in reply to Hariprasad Bhat

TI__Mastermind 36751 points

Hi Hariprasad,

Did changing the PD controller on that board fix the issue?

Thank you,

Eric

Interface

Interface forum

TPS65982: TPS65982 negotiation problem