LENOVO P1 G1 (X1 EXTREME) 0.8V MAIN POWER RAIL [LPM-1]

[reeluff]

No shorts on any main power rails.

All power mosfets over the whole board test good.

Have 20v into first DC in mosfet, when power button is pressed main power rail is at 0.8V. Not sure if it is the EC or BQ creating that 0.8V or?

The 2 input DC mosfets are not driving voltage, 2nd mosfet is N channel and no 25V on the gate. Again is this coming from the EC or BQ chip.

BQ chip has 25.6V at ACDRV.

Thankyou

[reeluff]

All finished, fixed beautifully.

[mcplslg123]

Its not possible that you have ACDRV at 25V on charging chip and not on the gate of dc-in mosfets. maybe you're looking at wrong mosfets.

[reeluff]

Definitely have 25.6V at ACDRV, and looking at the correct DCin mosfets.

Turns out that this ACDRV pin is floating - not connected to anything.

Mosfet drive must be coming from the EC chip, this board actually has 2 - specifically the TE1 chip at B2 / DCIN_DRV

[mon2]

The gate voltage has been boosted so it is the charger IC that will be supplying this line via ACDRV. Often there will be a 4k series resistor between ACDRV and the gate pin(s) of the DCin mosfets. Which exact charger IC is onboard? Would you have a link to the schematic? Probably the popular BQ (BenchMarq / TI part).

ACDRV voltage = adapter_in voltage + REGN voltage = 19v + 6v = 25 volts to enable the N-channel mosfet(s).

[reeluff]

BQ24780

I used green marker to highlight areas of interest. It definitely is not the BQ driving DCin mosfets from what I can understand, but the TE1 chip - which is a lenovo thinkengine chip right next to the main EC chip. This TE1 looks like a seperate EC lenovo thinkengine power management chip.

I have already replaced the TE1 chip, with no change. The chip was an aliexpress order, unsure if a good chip or not or whether this chip has to be programmed / can be programmed from the main EC however I am getting the exact same response from the board so will assume both new and old chips are as good/ bad as each other and operating the same.

Under a thermal camera, when I have the 0.8V main power rail - it is only this TE1 chip that is doing anything on the board.

[mcplslg123]

Now thats a completely new theory invented by you. To the best of my knowledge and experience, DC-IN mosfets are driven by charging controller thru ACDRV pin

[piernov]

It's not always the case. Sometimes you have a combination of P and N-channel DC-in MOSFETs, and the P-channel is not driven by ACDRV. And sometimes you even have both MOSFETs as P-channel, none driven by ACDRV.
And Lenovo likes to do weird things on their ThinkPad lineup. As far as I understand it, the secondary ThinkEngine (which is mostly a power sequencing chip but not much info about it) is powered directly before the DC-in MOSFETs and generates its own 3.3V rail.

[reeluff]

Correct Piernov, the power switch is getting the 3.3v from this thinkengine power chip.

[mcplslg123]

I've not seen a p channel mosfet(as dc-in) in BQ24780 design so far. Yes,P channel mosfet are used as dc-in on many boards but then the charging ic is like BQ24737(no acdrv) or ISL. Whereever, i've seen a BQ series charging IC with ACDRV function,its Usually a N channel. May be i've not come across such design.

[piernov]

Yes I meant with bq24780, it seems that it has replaced both the bq24735 (N-channel) and bq24737 (P-channel) designs for non-USB-C charging.
An example can be found on the Quanta X1BD board where 1st MOSFET is P-channel, driven separately, and 2nd MOSFET is N-channel driven by the charger IC's ACDRV output. I have no idea why they make it so convoluted though.

[reeluff]

This P1 board does indeed have 1st.P + 2nd.N channel DCin mosfet pair, neither of which are driven by the BQ24780.
Not a completely new theory invented by me, again i will upload the schemtatic pages of interest highlighted in green.
You can clearly see that the BQ24780 chip, ACDRV pin is floating - going nowhere as indicated by the big X - not driving any mosfet.
You will also see, the 2 DCin mosfets, first being a P channel, 2nd being an N channel. The N channel gate is being driven from the TE1 chip (thinkengine power sequencing EC) on the DCIN_DRV line.

[mcplslg123]

Yes,with the schematic pics,its clear now.Quite a weird design to diagnose.Sorry for my comments.

[reeluff]

All finished, fixed beautifully.