ThinkPad T14s Gen 1 (NM-C891) no power, I could use some help with the troubleshooting

[mrfree2ita]

I'm trying to revive a Thinkpad T14s Gen1 that seems to have issue with the power. Plugging in the USB-C charger the LED on the charging port doesn't lid and nothing happens if I press the power button.

Test conditions: board is outside the laptop, battery disconnected.

I made a number of checks already and found out that the ThinkEngine (page 94) does not set important signals such as M_ON that enable power rails such as 3v and 5v. After I verified that on such rails are not shorted, I also injected 3v into M_ON and the board comes to life (i.e. charging port LED gets orange, 3v and 5v are present). Naturally, it was just a test. ThinkEngine is responsible for a number of other signals needed for the laptop to work.

I'm now trying to trace down the issue starting from the IC charger. Here the measures I took on the IC charger PU0201 (page 102):

PIN

Measure

Label

PIN

Measure

Label

PIN

Measure

Label

PIN

Measure

Label

1

20v

VBUS

9

0v

IBAT

17

1.09v

COMP2

25

6.03v

BTST2

2

20v

ACN

10

0v

PSYS

18

3.5v

CELL_BATPRES

26

6.03v

LODRV2

3

20v

ACP

11

0v

#PROCHOT

19

9.24v

SRN

27

GND

4

3.27v

CHRG_OK

12

0v

SDA

20

9.24v

SRP

28

6.03v

REGIN

5

0v

EMZ_OTG

13

0v

SCL

21

3.25v

#BATDRV

29

6.03

LODRV1

6

2.18v

ILIM_HIZ

14

0v

CMPIN

22

9.24v

VSYS

30

6.03

BTST1

7

6v

VDDA

15

n/c

23

0v

SW2

31

0v

HIDRV1

8

0v

IDAPT

16

1.2v

COMP1

24

0v

HIDRV2

32

0v

SW1

I'm not an expert of such boards, but looking at the schematic (that I attach to this post for convenience) that missing PSYS seems rather important to me

The IC charger talks to the EC via SMbus. I also sniffed the signals on pins 12 and 13 with my logic analyzer and I found that the EC (MEC1663) sends something when I plug the charger in. So the EC doesn't seem to be completely dead. In fact, if I understand correctly, the IC charger acts like slave on the bus and just waits for commands.

Resistance to ground on PSYS is around 8k ohm, so in line with what I can see on the schematic.

I could use some help on the the next steps, guys

Is PSYS supposed to be there since the very beginning, is it something that comes up later on in the power on sequence? Do the values I measured look strange to you?

Many thanks in advance!

[SMDFlea]

Schematic & boardview -> https://www.badcaps.net/forum/troubl...atic-boardview

Reminder: rules and organization of this section --> https://www.badcaps.net/forum/troubl...f-this-section

[mon2]

Missing the voltage on SCL & SDA lines is of concern since these are open-drain lines and in their parked state, the lines will be floating so the pull-up voltage should have ~3v on each line but are at 0v.

With no power to the board, confirm the resistance to ground of the pull-up voltage rail of VCC3M (created by PU0501). What is the resistance? Also of concern since this rail is also mated with the PCH.

[mrfree2ita]

Missing the voltage on SCL & SDA lines is of concern since these are open-drain lines and in their parked state, the lines will be floating so the pull-up voltage should have ~3v on each line but are at 0v.

Thanks for pointing this out. Indeed, the VCC3M rail is absent at the moment. PU0501 is one of those that wait for the enable signal M_ON from the ThinkEngine. In fact, if I inject 3v on the M_ON line the VCC3M rail comes up regularly.

With no power to the board, confirm the resistance to ground of the pull-up voltage rail of VCC3M (created by PU0501). What is the resistance? Also of concern since this rail is also mated with the PCH.

Resistance to ground on the inductor PL0501 starts around 900kohm when I start measuring and then it goes down until it stabilizes at 40kohm

[mon2]

Confirm the power rails of UTE1B (ThinkEngine-3 block). Is this component being powered?

[anhbanxoi]

Plugging in the USB-C charger the LED on the charging port doesn't lid and nothing happens if I press the power button.

Try to plug usbc cable to the thunderbolt port nearby. The laptop may turn on

[mrfree2ita]

Confirm the power rails of UTE1B (ThinkEngine-3 block). Is this component being powered?

Here the measurements on the UTE1B (starting top-left corner in the schematic and moving anticlockwise). In the table, I use mV when I measured just millivolts:

PIN

Measure

Label

PIN

Measure

Label

PIN

Measure

Label

PIN

Measure

Label

44

6.6v

BAT_VOLT

55

3.3v

#SHUTDOWNIN

1

0v

M_PGS

36

0v

RD1_DRV

43

18.7v

#3SW_OFF

54

mV

TH_DTCT

33

0v

5B_DRV

35

0v

RD0_DRV

29

mV

RD0_ON

2

GND

34

0v

3B_DRV

30

mV raising

CP10OUT

27

0v

RD1_ON

37

GND

42

0v

RD6_DRV

32

mV

VDD10

26

mV

RD2_ON

46

GND

41

0v

RD5_DRV

47

3.3v

VCC3SW

25

0v

RD3_ON

57

GND

40

0v

RD4_DRV

28

0v

5B

24

0v

RD4_ON

53

3.3v

#PWRSHUTDOWN

39

0v

RD3_DRV

31

0v

VCC5M

23

0v

RD5_ON

56

0v

B_PGS

38

0v

RD2_DRV

45

19.8v

VREGIN20

Try to plug usbc cable to the thunderbolt port nearby. The laptop may turn on

I just tried. Unfortunately, no luck. Same result

[mrfree2ita]

I didn't manage to find much about the ThinkEngine chip (marked BD4179MWV (on my board) or TC62D518FTG).

Do the values I shared look good/healthy to you?

[reformatt]

There is no power sequence in the schematic unfortunately. There is however a sequence diagram in the Thinkpad X270 NM-B601 schematic, and looks to use the same ThinkEngine chip, so you should be able to use that as reference. Basic design looks identical to me.

The power management is shared between the ThinkEngine and MEC chips. Given that EC_RESET comes from the ThinkEngine chip, seems to me that he is the first guy active in the power up sequence. You are supposed to have M_ON directly after RTCRST. Are RTCRST and SRTCRST both 3V? Share the values for UTE1A might help.

Have you checked your PCH power rail for a short BTW? A lot of the signals relating to this initial power on do require involvement from the PCH. Power management IC may have damaged the PCH, it's always a possibility to rule out.

[mrfree2ita]

Thanks for your input reformatt, very interesting points.

Here the measurements on the UTE1A (starting top-left corner in the schematic and moving anticlockwise). In the table, I use mV when I measured just a few millivolts:

PIN	Measure	Label		PIN	Measure	Label		PIN	Measure	Label
22	0v	EXTPWR#		21	3.3v	PWRSW#		4	0v	SPISS#
19	mV	SLP_S3#		20	0v	PME#		5	0v	SPIMOSI
18	mV	SLP_S4#		9	0v	PGPIO0		6	0v	SPICLK
17	mV	SLP_S5#		10	3.3v	PGPIO1		8	0v	ECRST#
16	mV	SLP_LAN#		11	mV	PGPIO2		48	0v	CPUON
15	mV	SLP_M#		12	mV	PGPIO3		49	0v	BON
14	mV	SUSPWRACK		13	mV	PGPIO4		50	0v	AON
7	mV	PLTRST#		3	300mV	SPIMISO		51	0v	LANON
								52	0v	MON

I can measure 3.14v on both RTCRST and SRTCRST.


As you suggested, I also measured the resistance to ground on all the power rails indicated on page 16 (PCH POWER):

Label	Measure		Label	Measure
VCC3M_PCH	40Kohm		VCCDSW_1P05	5.7Kohm
VCC3_SUS_PRIM	26Kohm		RTCVCC	747Kohm
VCC1R8_SUS_PRIM	53Kohm		VCCRTCEXT	414Kohm
VCC1R05_SUS_PRIM	51Kohm		VCC1R05_SUS_XTAL	52Kohm
VCCPRIM_CORE	51Kohm		VCC1R05_SUS_AMP	51Kohm
			VCCDPHY_1P24	39Kohm

[reformatt]

To me, pin 44 (BAT_VOLT) seems low and could potentially be why M_ON is not being asserted. If VREGIN20 is indeed 20V, then I'd expect roughly 12.8V at pin 44, not 6.6V. This is based on the voltage divider R9504 (1M), R9506 (1.8M). R9509 is 0 ohms according to the schematic. That is of course unless the version of the schematic and the board don't line up.

Everything else to me looks fine. So check out this voltage divider network, R9505 and C9505.

[mrfree2ita]

Everything else to me looks fine. So check out this voltage divider network, R9505 and C9505.

Well, that's interesting.

R9505 and R9509 are both 0ohm resistors and I get consistent values while checking them.

R9504 and R9506 are... funky

Every time I measure them after a try (i.e. I plug in the charger and measure voltage on pin 44) I get different values:

For R9504 (nominal 1Mohm 1%) I've observed values like 780k, 804k, 921k and around 970k in different tries.
For R9506 (nominal 1.82Mohm 1%) I often get something around 1M, but I've also seen 1.6Mohm and even 1.7Mohm once.
C9505 is there and since it's in parallel with R9506 I get the same values.

Moreover, currently I seem to consistently get 12v on pin 44. However, I'm pretty sure of the 6.6v I previously reported for it and I found some older notes I took a few days ago that say 7.7v on pin 44.

I'm confused

What would make more sense to try at this point? I was thinking of removing C9505 from the board and see if it affects the resistors values, or even try injecting 12,8v directly on pin 44 (do I risk to damage something in this case?) to see if M_ON comes to life.

PS: the schematic matches with both the board view I'm using and what I observed on the PCB, so I'd consider it accurate.
PPS: I took measurements directly on the components. The resistance increases during the measurement. The value I report here were taken when the values stabilized.

[reformatt]

Either ground the top of the divider network to remove any residual charge (no power connected of course) or remove the resistors for an out of circuit measurement. Given they are pretty small, I move the component with hot air and keep one end soldered to one pad. Makes them stop flying off (measuring these components on the bench with probes is akin to changing Tom Thumbs nappies when he was a baby).

If both resistors check out and the volts are inconsistent, the next step would be replacing that ThinkEngine IC. Other than a donor board, the only sources I can see for them are off Aliexpress. SUHMS is normally a pretty reputable supplier and don't send you fake IC's.

[mrfree2ita]

Both resistors' values are spot on when not in circuit.

Out of curiosity, I also tried injecting 12.8v (300 mA max) on pin 44 of the ThinkEngine, but I didn't notice any change (i.e. M_ON didn't go high). Moreover, my PSU didn't report any current draw.

Many thanks again for your interest and support. I'll see if I can find the chip from a reputable source since I don't have a donor board at the moment.

Do you know if a new ThinkEngine (BD4179MWV) would need to be programmed?

[reformatt]

No I don't. There is a SPI interface between the MEC and TE chip though. It's possible that any programming is uploaded from the MEC at power on, but that would mean if the MEC has an issue then the TE won't work. We don't have any data sheets so it's just a guess.

[mrfree2ita]

"It's alive!"

I know that it has been a while, but I finally replaced the ThinkEngine chip and the board shows some signs of life!

The power led comes on (orange), and the white led under the power button blinks 3 times (then it stays off). My USB-C power meter shows that the board takes 30mA for a few seconds, then goes back to 0mA and keeps cycling between the two values.

Any idea what the 3 blinks might indicate? What would be the next logic step in the troubleshooting?

[mrfree2ita]

Some additional pieces of information... The power button led blinks only on the moment I plug-in the AC charger.

On AC plugged-in (both with or without battery), the following power rails are showing stable and correct voltage:
VCC5M_PD
VCC3M
VCC5M
VCC1R05_SUS
VCC3_SUS

while:
VCC1R2AP stays at 0v
RSMRST is also low

so it seems to me that accordingly to the power up sequence you suggested before it's kind of stuck on (or just before) the A_ON signal.

Moreover, on battery only (i.e. AC unplugged), when I press the power button the power rails go high for a split second and then discharge towards 0v

[mcplslg123]

Think Engine chip is programmable?

[mrfree2ita]

Think Engine chip is programmable?

I couldn't find any information about this for the ThinkEngine-3. According to the schematics, it can be one of the following:

Vendor	P/N	LCFC P/N
ROHM	BD4179MWV	SA000090J00
TOSHIBA	TC62D518FTG	SA00008E500

Based on my understanding, on this board the EC role is implemented by the MEC1663 (which is programmable) and the ThinkEngine-3. The two chips communicate via SPI.

The EC doesn't seem to be stone dead to me, considering that it should be the one responsible for the blinking pattern on the power led for instance.

Considering that when I got the board the BIOS chip was missing, I bought a new one and flashed a version I found here on the forum. But with or without the BIOS chip on, the blinking pattern doesn't change.

Any idea what I could inspect? I was thinking for instance to sniff the SPI communication with my logic analyzer to see if there's any useful pattern there...

[mon2]

To where are the LEDs that blink bonded? Is it the EC or the TE? If the TE, then for sure there is firmware already present inside to offer the blink pattern.