5V only on EL580 NM-B461

[theBread]

Hi 😄

I have a NM-B461 board that doesn't boot up.
Current observations:

• The board has no obvious damage, corrosion or previous tempering
• I measured the voltage on the TYPEC_VBUS20 and USBC_VBUS20 lines (which are the rails that get powered first according to the schematics) - the dock port measures 5V and the direct port 0V.
• USBC_VBUS20 (the rail from the direct port) measures 0.3V when powered via the dock port?! There is no short between those lines tho and there is no connection in the schematics.
• None of the power rails are shorted to ground
• Diode measures (red probe to ground, black to the CCx pins) measures .845 for each CCx pin
• The voltage on the CCx pins of the dock port flip as expected (depending on the orientation), but CC1 (and only CC1) of the USB-C port is high (between 1v and 1v6)
• None of the (currently theoretically active) components get noticably hot, neither does the power adapter

Page 3 and 97 of the schematics show a nice diagram which rails should be available and for what - VSYSTEM2 should be available when using the dock port, but its 0v because the input voltage for the ESD protection chip is 0.3v. The ESD chip itself does not short the input and output rails.

I currently suspect a faulty USB-C port, but I don't think that that's also the reason why the 20v don't get negotiated.

Thanks in advance 😄

(Sorry if this post should go in a somewhat similiar thread - didn't post there as the discussion moved towards the USB power and I wasn't sure if I'm supposed to necropost there)

[mon2]

Focus on one USB-C port from connector back to the motherboard.

Confirm the CCx pin is without short as you have done but continue to test the same line after the ESC / OVP circuits like @ U4701. Confirm that U4701 is being powered (VCC3_LDO_PD).

Study U4501 PD controller - check each power rail. Perhaps the firmware is corrupt or not loading for this PD controller which will then result in the power up default of 5v.

Test the TBT_xx lines that mate with this PD controller over the SPI interface.

Confirm that the flash device @ U4502 is being powered. The contents of this flash device may be corrupt leading to the PD controller firmware to be defective. Test the voltage on pin #8 of U4502.

[theBread]

Thanks for your quick answer

Some measurements so far:

• VCC3_LDO_PD is 3v3 as expected
• the CCx are not shorted (up to the point where they enter U4501), CC1 is 1v6
• PP_CABLEx is 0v as 5VTPC is created somewhere later on after VSYSTEM2
• U4502 is powered, DO pin is high - CS and the clock however are 0v (altough I suspect that the clock is simply not measurable that way)
• SPI_MISO is the only SPI pin that is high on U4501, LDO_1V8
• TYPEC_VBUS20 is 5.2V, USBC_VBUS20 (again) 0.3V

Is there a way to safely verify that the PD firmware is not loaded? My first though for that would be to listen on some of the SPI/I2C connections, but I doubt I could simply verify some handshake.
Also, the CS pin is low, which means that the output is theoretically there, but not selected/used?

Edit: Appended TPS65988 firmware dump - can't tell if its good or not or whether I can simply try the firmware of notebook for that chip

[mon2]

Vina appears to show a W25Q80 ROM file dump for this model. That may be worth a try but their site is not free. Always backup your original file 2 times; verify the contents and then erase and apply the next firmware for testing.

The clock not pulsing and stuck @ 0v is of concern. The clock is always driven by the SPI bus master so that will be the PD controller. Remove all power. Meter in resistance mode. What is the diode mode measurement of the SPI CLK pin? Red meter probe to ground; black meter probe to TBT_EE_CLK. We are testing if the TBT_EE_CLK pin on the PD controller is defective or not by testing the internal diode of that device.

To monitor SPI traffic, you require a passive bus analyzer. We use the Total Phase tools for I2C/SMBUS analyzing but there are many others including many clones of the Saleae devices. If you plan to do more work, review their original products with the USB 3.x front end and also one that offers varying voltage scales to monitor even low voltage devices. Otherwise, the low-cost clones on Aliexpress are strictly for 3v3 use only. We have other tools in the building, but they are also getting old and no longer supported for new interfaces.

Can you confirm if you have a stable +1.8VALW rail? This voltage is used by the mosfet level shifter that shifts the voltage for the SMBUS lines that are also present on the PD controller. See Q1001 & Q1002. Do you see any activity on these lines?

[theBread]

TBT_EE_CLK in diode mode measures .791 and is not shorted.

+1.8VALW is indeed a stable 0v - according to the schematics it depends on +5VALW (via PU801), which is 0v as well.

Btw the whole B+ rail (which itself depends on VSYSTEM2 according to schematics page 97) is 0v as well, which explains why +5VALW is 0v.

I think I'll have a look at those, but I'm still unsure as my repair knowledge is still pretty limited ^^'

[theBread]

Hi, small update - I just remembered that I have one of those usb c voltage/amperemeter so I decided to plug that in an check the ports for fun and games.
I plugged it in with just a little bit accompanying force and voila, 20v, fan spin and POST. My next steps will be to replace that port and then continue debugging the thunderbolt issue (which may be a bit harder because lots of "source rails" for other rails seem to be missing). I will post my results here just in case other people find this useful.

If someone happens to try this as well, please stay aware that these ports are highly delicate (or should be treated as such or else they may show these symptoms) and pure force won't fix anything.

[tomodachi]

Got same board on my bench with a destroyed U4502 firmware, could you please backup and share with me ?

[theBread]

Hey, yeah, that's the file above called TPS65988.zip. It should work as I've now discovered that my board boots when I use just a little force on the standard USB-C port - good luck

Edit: Read that you're search for thunderbolt firmware - not entirely sure, but that chip stores the PD controller firmware if I correctly understand the schematics