TPS65988: PJT setting for TPS65988DH

Hi Antony,

You will keep PP2 enable in 0x27 and assign the GPIO event to the external switch. Use PP2 as the switch source in 0x32.

So this external switch is only going to provide 5V and the 20V is from PP4, is the original 5V source still providing 5V or something else?

Regards,

Peter

Hi Peter,

There're two external switch (source path) for port-2.  One for 5V and the other for HV.  Based on the pjt I provided, originally the 5V source path is going through PP2 internally, but now they want to change it to the external 5V switch controlled by GPIO.  If I still keep PP2 enabled in 0x27 and also use PP2 as the switch source in 0x32 for 5V, there'll be two 5V path (PP2 and the external switch) running simultaneously in parallel.  THis should not be the result we want.

Antony

Hi Peter,

After we confirmed PP_HV2 is unconnected, we change the PD FW PJT as attached to configure GPIO0 CH4 in the waveform below) as "port-1 source PDO-0" to enable the 5V switch.  As you can see below, we can use GPIO0 to enable the power switch as expected when a sink device is plugged in, but for unknown reason GPIO0 is driven low shortly and VBUS2 drops accordingly.  What could be the reason for such result?

CH1:VBUS2

CH2:CC2

CH3:CC1

CJ4: GPIO0

Thanks,

Antony

TPS65988DH_6_1_1_New.pjt

Hi Antony,

The pjt file looks correct.

Do you see current limit event? What's the RDO current? GPIO0 is only connected to this external switch - correct?

FYI, there's a new GUI 6.1.2 in the website. Please make a plan to upgrade that as well.

Regards,

Peter

Hi Peter,

I just found out we can't define this GPIO as  "Port-1 source PDO 1 Negotiated" for customer's HW design.  Based on my measurement on EVM, the behavior of this GPIO is like below.  It will rises high for around 100ms and then low for awhile , in the end back to high level.

This is aligned with customer's measurement previously.  Since they're using this GPIO to enable the power switch to provide 5V to VBUS, when the GPIO become low in the middle, VBUS will become 0V and there'll be a reconnection repeatedly.

I also tried to use another GPIO even "PP2 switch ON" event, but still not worked for them since their PPHV2 is unconnected.  ( We confirm this event would become high only when there's a valid 5V on PPHV2.)

So, the conclusion here is, I can't find any suitable GPIO event to meet their requirement.  May I have your comment here regarding any other possibility to achieve this?

Thanks,

Antony

Hi Antony,

Peter is currently out of office, he will be able to provide feedback once he returns next week.

Thank you,

Hari

Hi Peter,

Do you think you can provide comments today your time?

Thanks,

Antony

Hi Antony,

I don't think that's a correct behavior. The GPIO should not toggle like that when used as IO driver.

What do you mean by measurement on EVM? Are you using a default pjt or customer pjt and what the EVM IO is driving/connected to?

I use the default pjt and monitor just the PB_PDO_0 (R158 on EVM) it does not toggle ON-OFF-ON when switching from 5V to 20V even though it's with TT setting it should behave as shown below.

Hi Peter,

I'm using a default PJT with GPIO0 (@R173) configured as "Port-1 source PDO 0 Negotiated" when I connected to a HDMI dongle which supports 5V sink only.  (Not support 20V).  I can see the same behavior on TPS6599x EVM as well.  Please help to clarify this using a 5V-only sink device.

Antony

Hi Antony,

So where is this GPIO0 connected to and how do you configure the pjt to use GPIO0? The EVM inputs are from the variable DC-DC and configured with the GPIO14 and 15 with TT configuration. I used power duo for 5V sink and obviously the R158 (PB_PDO_0) will not toggle because it's 5V.

HI Peter,

I load the GUI template and only change the definition of GPIO0 (see attached PJT).  GPIO0 is connected to R173 only on EVM as I know.  Could you double check it?

Thanks,

Antony

TPS65988DH_Advanced_v6_1_2.pjt

Hi Antony,

I used your pjt file and probe on R173, I do not see the double ON state as shown below when connected to 5V sink.

Hi Peter,

Your time scale is 100uS in the waveform screenshot.  Can you try to make it bigger like 100ms/div?  Based on my measured result previously, this GPIO would stay high for around 100ms and then become low.

I tried to test it on 994AD EVM using the default template from GUI6.4.17 with GPIO4 configured as event 14 (sourcepdo1contract_port2).  The results as below seems to be the same as 988DH.  That's why I think it should be related how we configure this GPIO event behavior in the FW.

For type-c 5V case (non-PD), this GPIO4 is never back to high.  I think it's because there's no PD contract established.

For PD 5V case, as mentioned above it will be high shortly around 100ms and drop to low and in the end back to high

Antony.

Hi Antony,

I see them, I need to debug further. I will let you know.

Regards.

Hi Antony,

In the customer design, can they use PP4(EXT) instead of PP2 (INT)?

In this setting, GPIO0 does not toggle twice.

Hi,

PP4 is already used for the source path with 20V source PDO, so it can't be used for 5V source path control.

Antony

Hi,

Maybe this will work since all you need is the trigger to enable the external switch. You can use PP3 and assign to VBUS2. They are not using port 1 - correct?

Regards.

Hi,

PP3 is already assigned as sink path for VBUS1, and GPIO16 is also used for the control of this sink path already.  It seems not workable to me.

Antony

Antony,

OK in that case, I will need to submit a FW Jira for this.

We will get back to you.

Regards,

Peter