Dead Dell Latitude 5420 not receiving 20 volts from USB-C charger

[BTubbs200]

Hello,

Dell Latitude 5420 P137G
Board: LA-K491P REV A01

I have received this laptop which was described as "does not power on or accept charge." Battery shows no sign of charging when charger is plugged in, nor does the board show any signs of life when battery is unplugged and charger is plugged in. No light from the LED at all or any other indications of activity.

Board seems to have no shorts, all power rails are fine and show high resistance.

When charger is plugged in, voltage is present in 3.3V rail but absent in others. Testing for 20 volts at +20V_LDO_INPUT and +20V_TBTB_VBUS shows between 4.5-5 volts instead. Does this mean the system is failing to negotiate with the USB-C to go from 5 to 20 volts? And if so, what would be my next steps in the diagnostic process? I am using a brand new OEM Dell 65w charger so it shouldn't be a problem with the cable.

[mon2]

Does this mean the system is failing to negotiate with the USB-C to go from 5 to 20 volts?

Yes. Suggest to pick up a low cost USB-C power meter that will display the voltage and current draw on the USB-C port. If you deal with Mobile Sentrix (Virginia), they have a new unit from Qianli that we recently purchased which can monitor the USB-C or Lightning port.

iBridge A3 Tail Plug Comprehensive Analysis Tester (Qianli)

Diagnosing and Repairing iPad Air 5 Charging Failure with iBridge A3 Tool – Professional Guide - YouTube

Does this same laptop offer the older style barrel / coax power adapter connector for charging? If yes, do test that as well.

On USB-C ports, there will be a power delivery controller which operates as a stand alone controller but may have either internal or external flash for firmware to function as noted. If all is well, the logic board's PD controller will request the PD contracts (volts & amps combination) from the power adapter after which the logic board will select the optimal choice to power the board. This is often the 20V PD contract. There are some affordable PD traffic logic analyzers including from ST Electronics and others. This tool allows you to view the bus traffic between the logic board and the power adapter. Not required at this stage.

Review if the USB C port is clean of debris / burn marks. Do test by flipping the connector by 180 degrees on the USB-C port. It is possible the firmware for the PD controller is corrupted.

[BTubbs200]

Review if the USB C port is clean of debris / burn marks. Do test by flipping the connector by 180 degrees on the USB-C port. It is possible the firmware for the PD controller is corrupted.

The ports look fine, I forgot to mention there's actually two of them and they're both showing the same symptoms. Flipping the connector 180 degrees resulted in no change on one of the ports, but on the other it resulted in 0 volts supplied to the system.

I do need to pick up a USB-C power meter, I will look into getting one shortly so I can proceed with some more detailed information.

Thank you

[mon2]

USB C ports communicate using a single wire communication line using biphase mark coding (BMC) which is the CC1 and CC2 lines on the USB-C connector.

Locate these pins on each of your USB C ports. Remove all power. Meter in DIODE mode. Red meter probe to ground, yes ground. Black meter probe to the point to test.

Test in DIODE mode the following pins of interest:

TBT_A_CC1
TBT_A_CC2

TBT_B_CC1
TBT_B_CC2

DT21 ; test each diode
DT22

We are testing each CCx pin to see if the short-to-Vbus transient protection device is damaged or not. The port with the 0v is of concern and could be a show stopper. Which port is that? Locate the short-to-Vbus local device for that port and inspect for visual damage. Most likely, the diode mode values will be off for that port.

Post each measurement. Remember, no power to the board in DIODE mode testing.

[BTubbs200]

Test in DIODE mode the following pins of interest:

TBT_A_CC1
TBT_A_CC2

TBT_B_CC1
TBT_B_CC2

DT21 ; test each diode
DT22

CC1/CC2 pins
TBT_A_CC1: 744mV
TBT_A_CC2: 743mV

TBT_B_CC1: 744mV
TBT_B_CC2: 742mV

DT11
CC1: 613mV
CC2: 602mV

DT22
CC1: 614mV
CC2: 602mV

The port of concern is JBUSBC1. No visual damage but the components are tiny and I don't have an appropriate magnifier to see in detail.

[bk304]

Did you get anywhere with this? I have a similar problem with a 5420.

[BTubbs200]

Did you get anywhere with this? I have a similar problem with a 5420.

I haven't. Spent a long time studying the board but couldn't find anything that pointed toward being the cause of the problem.

I tried replacing the USB-PD chip thinking maybe it was failing to negotiate voltage properly but I'm pretty certain I botched it. Absolutely tiny 50-pin BGA made of silicon, it was a pain to work with. Didn't fix the issue.

Only thing I haven't tried at this point is reprogramming the BIOS

[mon2]

When the 20V power delivery contract is not reached on a laptop (any brand / model),

a) check the contacts of the USB connectors themselves. Seen at least a few times where the contacts were scorched / dirty / contaminated. IPA (alcohol - the Costco pharmacy stuff is great) cleaning works well here. Use with Q-tips.

b) confirm the power adapter is worthy enough to power the unit. While the original OEM version is not mandatory to use, be sure that 3rd party units from Amazon for example, are able to support at least 100W on USB C + the mating cable is 100W or higher rated. Only such cables will offer the internal e-marker IC inside the cable connector hoods to permit passing of the higher contracts from the cable A to B side. In the absence of this cable, the PD (power delivery) contracts will remain at low current such as 3A or under and will not be suitable to power most portable devices. May be ok for phone / iPAD charging.

c) The integrity of the USB connector is important so a quick test tool like mentioned above (tail tester) is gold. You can learn quickly if there are any contact disconnects or perhaps damage to the mating PD controller.

d) The actual power delivery controller is a stand alone device and vendors vary. They will require local firmware to boot and offer the support power delivery contracts for this logic board. A PD logic analyzer is gold to have here and are now relatively affordable (many are < $100 USD). This tool can monitor the data traffic to/from the logic board and the power adapter. Then you can view if there is a fault by studying the traffic.

e) The PD controller firmware is a suspect and so is the controller. Confirm the voltage rail is present on each PD controller (much like the CD3215 20V article by piernov and others on internet). CD3215 is a Texas Instruments pseudo custom controller made for Foxconn / Apple for macbooks / iPads. CD3217 and other variations followed.

f) The above shared diode mode readings look ok but there could be other faults on the same controller (power rails, etc.). Review the schematic and inspect each.

Suggest to start with a USB-C meter (above tail tester should be ample) - share what voltage & current draw you see on each port. Being USB-C, the power is reversible to rotate by 180 degrees and test again. Same or different result?

[bk304]

On my 5420, plugging the original charger into USBC-1 shuts down the charger. Using USBC-2, it only goes to 5V. I am using the LA-K371P schematic because the LA-K491P schematic is not available easily. SCRIBD has it but there a too many hoops to go through to get it. The LA-K371P schematic is close but not exactly the same.

After some poking around I found that the 5V rail is not not working. I disconnected the solder jumper PJP4 and USBC-1 no longer shuts down the charger but still only went to 5V on the input. Measuring resistance on both sides of the open PJP4 showed a short on the output side. I have not isolated where the short is. The next step will be to put an external supply on the 5V path and see what gets hot.

[mon2]

Check you PM for details on the true schematic.

See below. What is the voltage to ground on this USB load switch?

[bk304]

I don't want to hijack this thread, but could BTubbs200 check to see if there is a short in their board between ground and pin 5 of UI3?
Thanks for the partial schematic.

[mon2]

What is the exact resistance to ground on pin #5 of UI3 on your board?

What about pin #1 of UI3 on your board?

[bk304]

Between pins 2 (GND) and 5 the resistance is .087 Ohms. Pin 1 shows about 3.5 MOhms to gnd.

[mon2]

Is your multimeter auto-ranging or manual type for the scale selection? If manual, can you post a pic of the meter face (scale) being used?

[bk304]

My meter does both autoranging and manual. I am using a Keysight 34461A and nulling out the probe resistance. I can send a photo of the meter face tomorrow.

[bk304]

Here is the photo. I don't think it says much becuase anywhere I measure +5_ALW shows similary results. This measurement was done on UI3. I am not sure what to do next. This circuit has so many protections that I can't tell what is shutting down what.

[mon2]

OMG - you truly have a nice measurement tool here. This is a dead short. Would you have an adjustable power supply?

Suggest to dial it up to 1 volt and inject with the highest current your power supply can support onto this shorted line. Be sure you are not injecting onto the ground plane of the pcb but otherwise the shorted part will be smoked out or will crack under pressure.

You can pour IPA (alcohol) onto the part or parts you suspect to be shorted. If the IPA evaporates quickly, that part is shorted. Share details on which part is the guilty one.

[bk304]

I should have done this first. I measured the resistance to ground of the 20V bus on the USBC connectors. The 20V buss connection on USBC1 measured about 9 Ohms to ground. The 20V buss connection on USBC2 measured about 229k Ohms to ground. Using a bench supply, I applied 6V to the 20V buss on USBC1 and turned up the current until it reached 6V @ 0.9A. The TPS65994AD USB PD controller chip got blistering hot as shown in the thermal image. I guess that is the problem.

[BTubbs200]

I don't want to hijack this thread, but could BTubbs200 check to see if there is a short in their board between ground and pin 5 of UI3?
Thanks for the partial schematic.

My attempt to replace the PD controller has put me in a worse spot than before, and now the charger is supplying no voltage to the board and my CC lines are absent, so I'm not sure I can help with this query. I did check, though, by probing CT44 which feeds into pin 5. Couldn't get a reading, multimeter said OL.

[bk304]

I will try and replace mine too. I will let you know what happens. One odd thing; I tried to read the BIOS chip and I couldn't. My programmer acts like the BIOS chip is dead. I may have to try it off board.