TPS65982: tps65982 connection issue with Surface go 2

Hi Gary,

I've assigned this issue to the relevant expert who will get back to you shortly

Hi Gary,

Can you try with the attached pjt file?

Regards,

Peter

v1p12p9_4p8A_new.pjt

Hi Peter,

 We use this PJT and customer's dock is become to power sink so alt mode / DP is ok, but we can't charge Surface now... I think it's not a good solution to solve the problem? I see "initiate swap to source" flag is disable in your pjt...?

Regards,

Gary

Hi Gary,

So the project I sent works, correct? It should be able to accept power role swap as well to become power source.

In the competitor's log, they are also started as power sink and UFP then follows with PR_Swap.

Do you have the project GUI? You can configure for them and try on their setup real time.

Regards,

Peter

Hi Peter,

 Your new setting let the dock be power sink always without PR swap. Please check the log as attached. Thanks.

Best regards,

GaryNo PD Charing log_20200817.zip

Hi Gary,

I have updated the project file to initiate as source and verify it on my setup.

I can see the PR_SWAP and enter into DP mode.

Please try this.

Regards,

Peter

2437.v1p12p9_4p8A_new.pjt

Hi Peter,

 We test the customer board and Surface go 2 again and we think maybe our PD disconnect connection due to protection...

1. Surface go 2 can be charge in high battery power (>85%)

2. If we short Rsense in PPEXT path, we can charge surface at any time.

3. All type-c docks using TI solution in customer have issue to charge surface go 2, but thunderbolt dock works ok.

 We can't see abnormal in waveform about I/V/CC in our test. Would you please helping to check any problem in connection and we can try to cover it from FW setting? Thanks for your kindly help.

Failed connection in current dock.

Good connection with TI TBT dock

Snapshot file

20v.zip

Best regards,

Gary

Hi Gary,

So this is a separate issue from your previous thread, correct?

From the scope captures, it shows a current limit condition occurs and when you short Rsense it's basically not sensing the current limit and allows you to charge.

From the debug register file, it shows the PD is using PP_EXT 20V 3A PDO to source the max 2.25A RDO, so why it hits current limit is something to debug further. When the VBUS current is below 2A VBUS stays at 20V, but there's also time it doesn't like the last cycle. What signal is channel 3?

Can you send the project file and PD log?

Regards,

Peter

Hi Peter,

 It's the same project as previous PD log/PJT. 

 Do you have idea how to speed up the debug process in the case? I take 987 evm and there's similar result in charging surface go 2. ... It's MP product and end customer is worry about the failed condition because we can't explain why the failed happened. Thanks for your kindly help.

Best regards,

Gary

Hi Gary,

This looks like an issue with ILIM configuration in FW, I sent an email to Jeffery Lee copy Zoe Ko to help with the FW debug.

Can you confirm that these are the files you refer to?

• Project file: v1p12p9_4p8A.pjt
• PD log: surface_dud_fail.usb

This will help them with the debug.

We need the FW state machine, what GUI version are you using? Are you seeing "PD State Machine Trace" under Commands tab in debug mode?

If you are not, since you are able to reproduce this with TPS65987 EVM; you can use that and follow the steps below to capture the FW trace and work with Jeffery to close this issue.

Regards,

Peter

State machine capturing steps.pdf

Hi Gary,

I talked to Jeffery and confirmed the FW is setting the OCP based on RDO operating current, so you need to set the Surface sink operating current to what it needs e.g. 2.25A instead of 0.75A.

Regards,

Peter

Hi Peter,

1. I use GUI v6.1.1 in current project and I don't see state machine function on this GUI. So I have to use other version (maybe for TBT, 99x or...) GUI to run state machine function?

2. I think I can't set Surface's sink current as 2.25A because it's MS's product....

 I'll get the machine at tomorrow morning (Taiwan time). Please help to advise what I have to test for this case. Or do we have any comment about the failed condition to end customer?

 Thanks for your kindly help.

Gary Teng

Hi Gary,

There's no PD State Machine for TPS65982, you can get the state machine for TPS65987 since you see the same issue with TPS65987.

But then, the cause of the problem is still the Surface operating current which is not correct and we can't change the FW for that.

You can inform the customer that the Surface is not configuring the correct operating current thus causing current limit in PD device.

When you say it works with TI TBT3 dock, do you know what TI PD device is in there?

If you do know, run the test and get the debug, configuration file and PD log on its EVM for the working PD?

Regards,

Peter