Dell Latitude 5320 USB-C port issue

[D3D_Tray]

Hey guys,

so my half-year out of warranty Dell Latitiude 5320 (19817-1 SOUTH PEAK 13 10L) w/ i5-1145G7 and 16GB just started to act weirdly...

The first/upper USB-C/TB port suddenly didn't accepted the charger every time and sometimes only a different one... or with a different cable but so far, the self-diagnostics didn't noticed anything so I suspected a dirty port.

Today however... it stopped completely to accept a charger, the second/bottom one however does and after a half an hour, I notice something getting hot under there... blazing hot!

I traced the issue down w/ alcohol (too poor for thermal camera) and it's a 5pin IC with the marking "53B7" (see picture)

Diode readings to GND from top to bottom;
1L = 14Ohm
2L = 0Ohm
3L = 1310Ohm
1R = 14Ohm
2R = 1840Ohm

Now I'm asking you for help... may that be the real issue or is something more beyond that?

Best regards
~Daniel

[mon2]

Suspecting this to be a SOT23-5 packaged TVS ESD diode array. Flux and carefully remove the part and test the PCB pads again (without his part on the board).

Also, when you are in DIODE mode, these are voltage readings.

When you in RESISTANCE mode, they will be in ohms.

When in DIODE mode, be sure the RED meter probe is to ground. Black meter probe on the pin to test.

[hobostove]

Looks like a switch/mosfet. It may have gone short. Remove it like mon said, and test resistance between each of IN, OUT, and GND pins. Should be 6 readings. Also test the board pads resistance of IN and OUT to GND. Let us know what you find.




Chip Datasheet: https://cdn.badcaps-static.com/pdfs/...41e690a0ad.pdf

[D3D_Tray]

Suspecting this to be a SOT23-5 packaged TVS ESD diode array. Flux and carefully remove the part and test the PCB pads again (without his part on the board).

Also, when you are in DIODE mode, these are voltage readings.

When you in RESISTANCE mode, they will be in ohms.

When in DIODE mode, be sure the RED meter probe is to ground. Black meter probe on the pin to test.

Ahhh okay, TY

1L = 500 kOhm initial, raising to infinit kOhm but slowly
2L = 1, 1 Ohm
3L = 55,9 kOhm
1R = 2,4 kOhm initial, raising up to 4,3 kOhm
2R = 104 kOhm

[mon2]

TY for the correction.

Pin # 1 is much healthier now after this part was removed. This USB load switch is defective. Needs to be replaced. Study the pinout and compare to other drop in parts from Texas Instruments / Diodes Inc. / Microchip, etc. There are many similar vendors that can drop into this same space. This is a current limited load switch and they just die after a transient. Wimpy by design. You can define the peak current support by the part you will purchase.

Before ordering, share the full details on what you intend to purchase.

[D3D_Tray]

Looks like a switch/mosfet. It may have gone short. Remove it like mon said, and test resistance between each of IN, OUT, and GND pins. Should be 6 readings. Also test the board pads resistance of IN and OUT to GND. Let us know what you find.




Chip Datasheet: https://cdn.badcaps-static.com/pdfs/...41e690a0ad.pdf

Uhm, now I'm bit confused but on the chip itself, I got these readings;

IN - OUT = 2 Ohm
IN - GND = 14,5 Ohm
OUT - GND = 14,5 Ohm
(used the G524 images as pinout)

On board I got;
IN - OUT = 16 kOhm initial, raising to infinite kOhm slowly
IN - GND = 2 kOhm initial, raising to 4,6 kOhm
OUT - GND = 500 kOhm initial, raising to infinite kOhm slowly

[hobostove]

Yeah, switch is definitely dead. The 5V in has only 14.5 ohms to jump to ground. No wonder is was getting so hot. Almost 2 watts going through that tiny chip!

Thermal Resistance Junction to Case, (θJC)
SOT-23-5. . .60°C/W (Means it was about 120C above room temperature) 😅

Board is good. Time to order a chip

[mon2]

TBH the original part vendor is crap. Most of these switches do die under such transients but my money would be on a TI or Diodes Inc. or Microchip part before a China only vendor.

[D3D_Tray]

I'm honest with you... I have right now no idea what I am looking at, the datasheet says something about 5V power limiter but... I had issues with charging the notebook at 20V (or whatever PD negotiated) and ...

*brain melting* So I need a replacement USB load switch for 5V and 2.5A (that probably is meant to prevent over-current over 5V for all ports
https://www.mouser.de/c/semiconducto...=Load%20Switch

[hobostove]

It's possible it's just a coincidence. One way to find out is you can power it up with that chip taken out. You just won't have 5V on that USB-A port, so it won't work. But everything else will work fine

[D3D_Tray]

Hmm... Both USB-C/TB ports takes power and PD in except the first/upper port only on one side, I tried with several cable different cable and my Lenovo USB-C brick (don't have the Dell original one)

Maybe USB-C port is a bit loose? Loose and somehow bricked the IC by accident? (At least no heater inside anymore)

I use a J7-c in between to check the voltages and something caught my eyes, I have one cable with a small display for Watt reading and if PD is enabled or not, no matter how I flip it, it keeps blinking on the first/upper port but not on the second/bottom port (on the picture I've sent it's upside down ofc ^^)

[hobostove]

The switch chip may have been broken for a while, and you just noticed it because you were actively looking for a problem. Don't think it's related to the USB-C issue.

The fact that it got worse overtime and didn't just suddenly stop makes me think the port is dirty or damaged. Try to get a good look at the pins. You'll need a magnifying glass or microscope, etc.

What does the J7-C tester say on that port? Does its screen come on?

[D3D_Tray]

The J7-C test only powers up on one side, the other side it acts like being not connected.
Are there any testpoint on the board to test the USB-C pins? I have a breakout board for USB-C so I could check for continuity on each pin (the manual way)

I also realized this morning that only the USB-A port on the same side doesn't have 5V, the other side still functions fine :o