UCD9090-Q1: GPO output value not coherent with equations

Clément Letonnelier

Part Number: UCD9090-Q1

Hello,

We already exchanged on one of our design on the linked post with yihe.

We are encountering a very strange issue with that same design, on the equation of the TMS_GPIO21 output pin (HWRST_Ni on the schematic sent through our FAE Franziskus Hillreiner).

The configuration of the GPIO is the following :

All signals except the FB_LBK_HWRST_N are active high inputs.

We have checked on the board with an oscilloscope and all the inputs except the FB_LBK_HWRST_N are low. However, we would have expected HWRST_Ni to be high at board level and it is low.

We confirmed these levels with the Monitor task in Fusion Digital Power Designer.

This does not make sense.

We encounter this behavior after some tests which leads to the FB_LBK_HWRST_N toggling (due to UV generated on the MON inputs) between high and low with DIS_RESET_GEN low to ensure the HWRST_Ni is not set during these tests, when we set DIS_RESET_GEN back to high state after FB_LBK_HWRST_N being high during at least 40ms.

We do not encounter this behavior each time, we couldn't figure out a pattern of occurence, and it happens on two UCD9090-Q1 of our board.

Anything you see we could be doing wrong ?

Best regards,

Clément

10 months ago

0 yihe 10 months ago

TI__Guru 60800 points

Could you share the project file?

Regards

Yihe

0 Clément Letonnelier 10 months ago in reply to yihe

Expert 2436 points

Yihe,

I think you already have the project file, it was sent to you by Franziskus along with the schematic in Juy.

Do you still have it ?

Best regards,

Clément

0 yihe 10 months ago in reply to Clément Letonnelier

TI__Guru 60800 points

I received back to July and is the right file?

Regards

Yihe

0 Clément Letonnelier 10 months ago in reply to yihe

Expert 2436 points

Hi,

Hum no that's not exactly the last version, I thought it was.

I'll send it to Franziskus on monday so he can transfer it to you.

Best regards,

Clément

0 yihe 10 months ago in reply to Clément Letonnelier

TI__Guru 60800 points

Sure, I will check once receiving it.

Regards

Yihe

0 Clément Letonnelier 10 months ago in reply to yihe

Expert 2436 points

Hello Yihe,

Finally I am attaching it here.

Best regards,

Clément

project.xml

0 yihe 10 months ago in reply to Clément Letonnelier

TI__Guru 60800 points

Thank you for the file. I will review and provide response tomorrow.

Regards

Yihe

0 yihe 10 months ago in reply to yihe

TI__Guru 60800 points

The issue is that Pin24 GPIO15 can not be configured as the 6th GPI. You have to swap it.

The general rule for those low latency GPIs, those GPIs can not be assigned to based on the number(1-8) listed below.

For example: GPIO1 can not be configured the1st GPI. GPIO17 can not be configured as the 8th GPI

Regards

Yihe

0 Clément Letonnelier 10 months ago in reply to yihe

Expert 2436 points

Yihe,

Where is that information stated in the documentation ?

Does it mean that the GPIO state change isn't taken into account because it's low latency ?

Also, we only have #7 GPI, so we can't swap it so that it goes above 1-8, will that work ?

Clément

0 yihe 10 months ago in reply to Clément Letonnelier

TI__Guru 60800 points

When the GPI is configured as low latency GPI, the state of the GPI may misbehave.

what you need to do is to not configure GPIO15 as the 6th GPI. you can configure 1st, 2nd. 7th. but not the 6th.

Regards

Yihe

0 Clément Letonnelier 10 months ago in reply to yihe

Expert 2436 points

Hi,

Ohhhhhhhh, gotcha. We will try.

I would suggest that you implement a warning in the software then if that's the case.

Best regards,

Clément

0 Clément Letonnelier 10 months ago in reply to Clément Letonnelier

Expert 2436 points

Yihe,

We switched #6 with #7, made the sure the equations were good after the change and the behavior is now even more erratic.

The state of the GPIO15 just does not seem to have any impact anymore on the output result of the GPO #5 equation...

Best regards,

Clément

0 yihe 10 months ago in reply to Clément Letonnelier

TI__Guru 60800 points

That's impossible.

You may need double check LGPO setting and also you have to power cycle the board to let the new settings take effect.

Please send your new files.

please ensure that the file hast the proper name. the previous file does not have name and it is hard to follow

Regards

Yihe

0 Clément Letonnelier 10 months ago in reply to yihe

Expert 2436 points

Yihe,

What do you mean by double LGPO setting ?

I'll get back to you on Friday.

Clément

0 yihe 10 months ago in reply to Clément Letonnelier

TI__Guru 60800 points

I mean double check your LGPO setting to ensure that the right GPI is set.

Regards

Yihe

0 Clément Letonnelier 10 months ago in reply to yihe

Expert 2436 points

Hi,

The issue was a side effect from another function following the working change.

So your suggestion did solve our problem.

One question remains, where is that constraint/issue documented in your technical documentation ?

Clément

0 yihe 10 months ago in reply to Clément Letonnelier

TI__Guru 60800 points

Thank you for the feedback. will will take action for the future releases.

Regards

Yihe

Power management

Power management forum

UCD9090-Q1: GPO output value not coherent with equations