TPS25831-Q1: Not generating 5V

Michael Durian

Part Number: TPS25831-Q1

I'm using a TPS25831 in my design, but it is not behaving as expected. It isn't generating the 5V output when a device is attached to the USB-C connector. I've got it configured in DCP mode.

With no device attached, I measure about 0.48V on OUT. With a device attached, an iPad in this case, I measure about 0.25V on OUT. I also see short bursts of activity on SW: ~400kHz pulses every 300us, but that might be RF noise since it is present regardless of if a device is attached.

Since I don't need LD_DET or POL in my application, I did not route the signals and unfortunately, cannot check them to see how they respond when a device is attached. I do FAULT and THERM do not show errors.

I read 5.5V on VCC, which seems high, but I've checked two boards and they both show the same value. IN is 9V.

If anyone has any insight as to what I've done wrong, I'd appreciate it.

Thanks,

mike

over 2 years ago

0 Bob Ma over 2 years ago

TI__Expert 8980 points

Hi, Mike,

Thanks for reaching out. Can you let me know your cc1 and cc2 voltage when device connected and not connected.

Regards,

Bob

0 Michael Durian over 2 years ago in reply to Bob Ma

Prodigy 210 points

Hi Bob,

Thanks for the reply.

Unconnected: CC1 = 4.8V, CC2 = 0V

Connected: CC1 = 0.4V, CC2 = 0V

mike

0 Bob Ma over 2 years ago in reply to Michael Durian

TI__Expert 8980 points

Hi, Mike,

Do you have a cable connected? If not, CC1 and CC2 open value should be ~4.3V. Please check if CC1 and CC2 have a good connection.

Regards,

Bob

0 Michael Durian over 2 years ago in reply to Bob Ma

Prodigy 210 points

Hi Bob,

You are right about having the cable connected. One of the CC lines on the cable shows about 900 ohm to ground. This is the one that read 0V.

I have another cable where I read 4.8V on both CC lines, but I don't think the TPS25831 generated 5V with it either.

I hesitant to do more testing with a real device. I'm pretty sure I fried the charger on my iPad when I was using it for testing (luckily it was under warranty and I was able to get it replaced). What is the easiest way to convince the TPS25831 to generate 5V so I can test without using a real device? Or if that is not easy to do, can you think of any inexpensive device that I could use for testing.

Thanks,

mike

0 Bob Ma over 2 years ago in reply to Michael Durian

TI__Expert 8980 points

Hi, Mike,

It's weird that it damage your product. To test it without devices, you can apply Rd(5.1k resistor) from CC1 or CC2 to GND. Then there will be 5V on VBUS.

You can refer to below table in datasheet.

Regards,
Bob

0 Michael Durian over 2 years ago in reply to Bob Ma

Prodigy 210 points

Hi Bob,

Maybe it was something else that damaged the iPad. I did spill some coffee on it recently. The Apple tech didn't detect any signs of water damage, so I'm not sure.

I'm still seeing weirdness with the TPS25831. I got a new USB-C cable and cut it in two so I could access the wires easier. There was only one CC wire and the other CC position on the connector was not pulled up or down to anything. I don't read quite open loop between various pins, but the resistance I do see is a few Mohms or higher.

When I put a 5.1k resistor between the CC signal and ground, I still do not get 5V from the TPS25831. I get 0.48V. I think it is also strange that Vcc reads 5.5V. The datasheet says it should be between 4.75V and 5.25V.

With the cable plugged in (with the 5.1k resistor), I read 4.75V on one CC and 0.43V on the other.

mike

0 Bob Ma over 2 years ago in reply to Michael Durian

TI__Expert 8980 points

Hi, Mike,

You are correct, VCC should not be 5.5V. It probably related to NTC circuit. Can you measure NTC pin voltage?

Bob

0 Michael Durian over 2 years ago in reply to Bob Ma

Prodigy 210 points

Hi Bob,

The NTC pin voltage is 1.43V.

mike

0 Bob Ma over 2 years ago in reply to Michael Durian

TI__Expert 8980 points

Hi, Mike,

Can you request an EVM to do cross check. You input and schematic looks good, but the part isn't in correct condition, CC1/CC2/VCC voltages are not correct. Can you check IC connection and GND connection.

By the way, I didn't see you connect PGND to GND in schematic, but you should have connected it together, correct?

0 Michael Durian over 2 years ago in reply to Bob Ma

Prodigy 210 points

Hi Bob,

Yes, PGND and GND are connected through an inner layer ground plane.

I'm going to have a board x-rayed again and see if that shows any problems. It was x-rayed during assembly, but maybe they missed something.

We purchased the chips from DigiKey so I'm not too concerned with their authenticity, but that said. Does these markings look reasonable?

0 Bob Ma over 2 years ago in reply to Michael Durian

TI__Expert 8980 points

Hi, Mike,

The mark looks fine.

To check the connection, you can measure the ESD diode. There should be a diode can be measured from GND to each pin.

And you can request an EVM to do cross check.

Bob

0 Bob Ma over 2 years ago in reply to Bob Ma

TI__Expert 8980 points

Hi, Mike,

Any update on this? Do you have this resolved?

0 Michael Durian over 2 years ago in reply to Bob Ma

Prodigy 210 points

Hi Bob,

I just got the x-rays back this morning and everything looks normal. Do you have any thoughts on what might make the internal LDO overshoot to 5V? I think if I could identify why that is happening, I'd be able to figure out what I did wrong.

How does one request an EVM?