STUSB4500 Repeated Component Failure

[valentinescott03]

Good Evening,

I am a student doing a capstone project, and I have been having a massive headache with the STUSB4500 PD negotiator. I am using it to pull 15V at 3A from a power brick so I may power a handheld device downstream. We utilized a recommended layout from the datasheet itself so that we may program it ourselves. Attached below.

The good news is, it can sustain loads and power our project... but the bad news is these chips have up and died about as randomly as a carnival prize goldfish, attached are pictures of my layout and schematic for this section of the board. Apologies for any amateurish mistakes on schematic labelling and layouts.

The trouble is that we can program the chip, and it stays stable for a long while. But out of nowhere, three times now, we will go to plug the board in one day, and the chip will be dead. 

I believe the chip to be dead since it no longer tries to negotiate the preset 15V like we wanted, the 2V7 and 1V2 lines are seemingly dead, and holding reset low or high changes absolutely nothing (I only do this after the chip becomes unresponsive to see if anything wakes it). 

We have appx 3 weeks left, and I have one good chip left, with two more on the way. If there are any bodges or even board updates I can make to this layout, please alert me. 


I thank you all for your time and help!

[valentinescott03]

I would like to add, this behavior only seems to occur during plug in.... it has never outright died during use.

I have had a weird problem where the chip will shut off if I were to try and probe the board with a multimeter, but that hasn't happened in weeks, in fact it only happened once.

[Hamady]

Hi @valentinescott03 

We have a project where we need to use USB PD so we used the STUSB4500. Scematic from sparkfun devkit

I had 5 PCB and 3 of them have a IC fail . Not able to communicate in I2C and unable to negociate power.

And it was after multiple plug / unplug. And even one failed at first use . And unable to have 5V 

Also one big flaw of this chip is that if your device have a battery the IC doesn't have VCC and the I2C is grouded so all of my sensors was unable to work.

Its seems that this chip  have a big issue and major flaws 

Does the STUSB4531 is safe to use ? And can ST recommand a fix ?

Thanks

[valentinescott03]

I wouldn't know. I see no recommended schematics on the STUSB4531 datasheet anyways.... I went to check recommended layouts for the STUSB4500 only to find my layout seems to be fine when compared to sparkfun's board and the devboard for the STUSB4500... I really need help with this, my degree kinda depends on it...

[Didier HERROUIN]

Dear valentinescott03,

There is no known weakness of the STUSB4500 even though it has been deployed in millions of units.

I confirm that 2V7 and 1V2 being low put in evidence that the chipset does not operate. 

Your schematics are correct except capacitor C7 should be 100nF instead of 1µF.

As the STUSB4500 is supplied through VDD pin, which requires VBUS voltage, could you check the presence of VBUS (provided by the charger) ? If no VBUS is present, it could mean that the Source (charger) does not detect the Sink (STUSB4500) via CC pins as expected.

Best regards.

[Hamady]

Hi @Didier HERROUIN 

Very strange to hear that STUSB4500 does not have issues . In the ST forum there is a lot of topic talking about STUSB4500 not working after connection / deconnection.

Are they things in the schematic that the chip is sensible to ? 

We tested STUSB4531 with 300 cycles and didn't see issue but the 4500 have failed in custom PCB 4 times (out of 5).

Thanks 

[Didier HERROUIN]

Dear Hamady,

The issues in the topics you mentioned were due to different reasons, like schematics mistake, unexpected high current in the MOS transistor, misunderstanding of the USB standard or STUSB4500 itself, etc.. these are normal discussions in a forum. But none of them (until now !) has put in evidence a weakness in the chipset.

Coming back to the issue, in addition to VBUS presence checking, are you still able to communicate with the GUI, using I2C ?
Could it be a soldering or an assembly issue ?

Best regards,

[Hamady]

when the issue is occuring i'm unable to communicate in i2c and VBUS is at a weird voltage like 2v.

For assembly issue i don't think beacause the chip worked before and we heated the pcb with flux to melt solder. 

Thanks