[C5515EVM] Unexpected power on USB_CPU_3V3_USB line during RTC-only mode

Hi kaka-san,

Will get back to you on this, I was looking at your other thread on leakage current with respect to SPI_TX pin.

Regards

 Vasanth

Hi Kaka-san,

This is what we have observed in the past,  from VBUS to 3.3V there is a weak path through weak resistors. Hence if 3.3V is down an VBUS is up, there is a residual voltage on 3.3V supply. But for the scenario you had mentioned (where VBUS is completely OFF ),  I need to discussed this further with USB expert and unfortunately he is out of office and would take some more time.

Meanwhile If I need anymore details I will get back to you.

Regards

 Vasanth

Hi Vasanth,

Thank you for your update.
Just in case, I summarized my observation for this phenomenon. Please check out this.

C5535_Result.pdf

If you will get more information please let me know.
Best regards.
Kaka

Hi Kaka-san,

  Thanks for the details.

  Is it also possible to share your code, you could share code it to my TI email ID. This will help in replicating the issue.

Regards

 Vasanth

Hi Vasanth,

Yes, I can share my code.  But I do not know your TI Email ID....

I attached this code to this post.Wakeup_interrupt.zip

If you have any questions, please let me know.

Best regards.

Kaka

Hi Kaka-san,

I am trying to reproduce the issue, but I do have few more questions.

1. Is this issue reproduced in any other hardware platform like customer board etc ?

2. In C5535 result.pdf you had mentioned in point #4

             Removed the jumper of JP15

          -> Not supplied for USB_3.3V line

    I believe you meant JP19 is my understanding correct ?

Could you Pl let me know on the above.

Regards

Vasanth

Hi Vasanth,

Thank you for your updates. I answered for your questions.
1.Is this issue reproduced in any other hardware platform like customer board etc ?
[Kaka]
Yes. We have two C5515 EVMs and we could reproduce this issue with both boards.
In addition to this , my customer board (It used the C5535) also confirmed this phenomenon.

2. In C5535 result.pdf you had mentioned in point #4
Removed the jumper of JP15
-> Not supplied for USB_3.3V line
I believe you meant JP19 is my understanding correct ?
[Kaka]
Yes. That's correct. JP15 is typo. I apologize for your confusion.

If you have any more questions, please let me know.
Best regards.
Kaka

Hi Kaka-san,

Thanks for confirming my questions.

I tried to reproduce the issue on two different platform.

1. C5515 EVM
2. Power board platform (TI internal H/W platform this will not have peripherals connected and used for power testing/measurements).

Here are my observation when verified with these two H/W platforms. 

• When tested on power board, I did not observe any residual voltage on USB 3.3 and 1.3V supply rails (when they were OFF). The voltage measured on these supply rails were close to zero volts.

• On C5515 EVM I did see the residual voltage, although measured voltages were not similar when compared against your data. 

I do have few more questions based on the above tests.

 1. How different is customer board when compared to C5515 EVM (Specifically looking at USB )?

 2. Once the power was switched OFF to USB, how much time you allowed before measuring ?

Regards

Vasanth

Hi Vasanth,

Thank you for your confirming. Please let me confirm about your test result.
To summary your result :

[Issue.1]
There is leakage voltage on USB 3.3V line during RTC-only mode.
Summary the power domains as below.
[Power domain condition]
LDOI, DCDDRTC: 3.3V
DPS_LDO: OFF
--> CVDD:OFF
CVDDRTC: 1.3V
ANA_LDO: OFF
--> VDDA_PLL, VDDA_ANA: OFF
USB_LDOO: OFF
--> USB_VDDA1P3, USB_VDDA1P3: OFF
DVDDIO. USB_VDDOSC, USB_VDDPLL, USB_VDDA3P3, VBUS: OFF

[Issue.2]
If supply the USB 1.3V during RTC-only mode, the leak voltage will increase compared to USB1.3V line is off.
[Power domain condition]
LDOI, DCDDRTC: 3.3V
DPS_LDO: OFF
--> CVDD:OFF
ANA_LDO: OFF
--> VDDA_PLL, VDDA_ANA: OFF
USB_LDOO: OFF
Exrernal 1.3V On
- -> USB_VDDA1P3, USB_VDDA1P3, CVDDRTC
DVDDIO. USB_VDDOSC, USB_VDDPLL, USB_VDDA3P3, VBUS: OFF

[Result]
1. You did not observe those issues when you used the your power board platform.
2. You can observe those issues when you used the C5515EVM. But there are some differences between your result and my result.
Is my understanding correct? If possible, would you please share the waveform which you observed?

Also I answer your question as below.
1. How different is customer board when compared to C5515 EVM (Specifically looking at USB )?
[Kaka]
My customer used the LDO (Parts number: S-1333, It is supports Discharge shunt function) for USB 3.3V line.
datasheet.sii-ic.com/.../S1333_E.pdf
So, this line will be off(0V, this line will connect to GND) during RTC-only mode by enable pin of LDO. And if the USB1.3V line will not remove while RTC-only mode, there is leakage on USB 3.3V line and it short to GND by LDO. This is difference between customer board and EVM. Also you can make the similar situation by shorting the USB 3.3V pin(of C5515) to GND.

2. Once the power was switched OFF to USB, how much time you allowed before measuring ?
[Kaka]
I apologize that I could not understand your question. Would you please explain about your question in detail?

If you have any questions, please let me know.
Best regards
Kaka

Hi Kaka-san,

Below is my response:

[Result]
1. You did not observe those issues when you used the your power board platform.
2. You can observe those issues when you used the C5515EVM. But there are some differences between your result and my result.
Is my understanding correct? If possible, would you please share the waveform which you observed?

Vasanth :  Yes that's correct on Issue 1.

Also I answer your question as below.
1. How different is customer board when compared to C5515 EVM (Specifically looking at USB )?
[Kaka]
My customer used the LDO (Parts number: S-1333, It is supports Discharge shunt function) for USB 3.3V line.
datasheet.sii-ic.com/.../S1333_E.pdf
So, this line will be off(0V, this line will connect to GND) during RTC-only mode by enable pin of LDO. And if the USB1.3V line will not remove while RTC-only mode, there is leakage on USB 3.3V line and it short to GND by LDO. This is difference between customer board and EVM. Also you can make the similar situation by shorting the USB 3.3V pin(of C5515) to GND.

Vasanth :  I would try to create similar scenario and get back with results.

2. Once the power was switched OFF to USB, how much time you allowed before measuring ?
[Kaka]
I apologize that I could not understand your question. Would you please explain about your question in detail?

Vasanth:  I was suggesting you to wait for some time, before you measure the voltage (residual voltage) on USB 3.3 and 1.3 supply rails, this is to ensure no discharge  being observed on these rails.

Regards

 Vasanth

Hi Kaka-san,

With Issue 1, I am not seeing any leakage on both USB supplies on power board.

I will re-verify on the Issue 2 you had  mentioned and get back to you .But with respect to this scenario there exists is a direct path between USB1.3v and USB3.3v through few transistors. Hence when USB 1.3V supply is up and USB 3.3V supply is off then these transistors will be ON and leak into 3.3V supply. I wll reconfirm on this.

Regards

 Vasanth

Hi Kaka-san,

Below is my response with respect to power board tests (marked in brown):

For issue1, I could confirm that there is 0.3V on USB3.3V line when the USB3.3V did not supply during Active state.
i.e.
1. Remove the JP19
2. Power on the EVM
3. Measure the voltage on USB3.3V line.
-> You will observe 0.3V on USB3.3V line.

Vasanth: I see around 0.25V when power both supplies are off, but eventually after a minute or so will reduces to almost 0V.

For issue2, I also reproduce this without entering RTC-only mode.
1. Remove the JP19
2. Change the short pin of JP15 from 1-2 to 2-3
3. Power on the EVM
4. Measure the voltage on USB3.3V line.
-> You will observe 0.7V on USB3.3V line.

Vasanth: My observation are similar with respect to this case  (on power board).

Regards

 Vasanth