A1708 820-00840 charging issue

[twonk9098]

Re: A1708 820-00840 charging issue

Strange how I got different readings from pin 2 of R7138 and R7139, I know it's only slight but I doubt there should be that much difference. Presume its my probe contact point.

[mon2]

Re: A1708 820-00840 charging issue

CPU_VIDALERT_L is at 0.1v

being a logic active LOW signal, this signal is reporting that there is a FAULT.

Review each of the current sense lines from the CPU power rail:

CPUVR_ISNS1_P
CPUVR_ISNS1_N

CPUVR_ISNS2_P
CPUVR_ISNS2_N

and be sure they reach the op-amp through the resistors in the schematic @ U5600.

That is, U5600 should output some value to report that all is well / or not.

It is possible that the rail is turning on briefly and then shutting down due to a missing or defective current sense component which is the loop of parts discussed (U7100 creates the PWM to drive the voltage rail for the CPU).

What is the history of this board ? Any liquid damage ?

The indicators seem to hint that it should all work but you are drawing very low current. You have 20V so the ACE controllers and local ROM to the ACE are good.

Need to grab some groceries...bbl.

[twonk9098]

Re: A1708 820-00840 charging issue

The history of the board is that it came in for not charging or turning on.
The ammeter showed 5v low amps.
The CD3215's seemed fine however I had to replace the PI3USB32324XEAE USB mux.
This got me to the stage I am now.
There is no signs of any liquid damage and no signs of previous work on the board.
I think the charger maybe causing this issue as I replaced the logic board a few weeks back for exactly the same issue.

I will test those lines shortly.

Many thanks for your help.

[twonk9098]

Re: A1708 820-00840 charging issue

I have taken a look at the schematic as suggested however the current sense components do not exist on this board. I have two of these boards and both do not have any of the components. I have attached images to show the areas where the components should be.

CPU_VIDALERT_L is low however I cannot understand what would generate this especially without the current sense components.

Thanks

[mon2]

Re: A1708 820-00840 charging issue

Remove all power. Meter in resistance mode.

What is the resistance to ground of PPVCCSA_S0_CPU ?

This rail = PPVCCSA_S0G.

This rail is also controlled by U7100 and there is a current sense circuit shown in the schematic. Perhaps this leg of the circuit is triggering the alert.

[twonk9098]

Re: A1708 820-00840 charging issue

I get 13.7 and 13.1 ohms either side of R0804.
I also measured CPU_VCCSASENSE_N at C7121 and I get 2.1 ohms to ground. This sounds very low even for sense circuit? Any idea if that is to low?

Thanks

[twonk9098]

Re: A1708 820-00840 charging issue

Hi,

Does anybody know if CPU_VIDALERT_L is sent to the CPU or if the CPU sends it to U7100?

Thanks

[mon2]

Re: A1708 820-00840 charging issue

OUTPUT from U7100.

[twonk9098]

Re: A1708 820-00840 charging issue

Thanks Mon2,

But if it's an output from 7100 why would there need to be PP1V_S3 injected into the circuit? Is that for sleep?

Thanks

[twonk9098]

Re: A1708 820-00840 charging issue

Think I've just found the answer to that. S3 is a sleep mode therefore the CPU needs a little power to be awake in sleep mode.
So my next question is what voltage should U7100 output on the CPU_VIDALERT_L rail under normal conditions?

[twonk9098]

Re: A1708 820-00840 charging issue

Does the screen have to be connected for CPU_VIDALERT_L to be high?3

[twonk9098]

Re: A1708 820-00840 charging issue

Just had another A1708 through the door so was able to test CPU_VIDALERT. It still does go high even without a screen attached.

[twonk9098]

Re: A1708 820-00840 charging issue

I've also tested the SAVR_FCCM line from U7100. It's supposed to output 5v according to a working board however I get 0v. Anyone have an idea what U7100 requires to output this line?

I also compared the following lines which are different. Measurements were taken without a HDD

CPUVR_ISUM_P is 0.703v on a good board and 0v on my board
CPUVR_ISUMN_R is 0.701v on a good board and 0v on my board
CPUVR_ISEN1 is 0.691v on a good board and 0v on my board
CPUVR_ISEN2 is 0.688v on a good board and 0v on my board.

All other voltages on U7100 are correct except ALERT_L which is also 0v.

[twonk9098]

Re: A1708 820-00840 charging issue

Not sure if this helps anyone however the attached pdf contains all the voltages and diode readings for the U7100 on the 820-00840 board.
Please note, the measurements were taken from a fully working board which was removed from the machine and no HDD attached.

[Pedro147]

Re: A1708 820-00840 charging issue

Hi. See attached.

The part is @ U3000 = Diodes / Pericom PI3USB32324XEAE .

The BV file that shows the U3000 = 820-00840-01

Hi mon2, any chance of sharing the sch/brd files that has that USB driver included please? Nobody else that I know has it. On my copy there isn't a U3000 at all, boardview or schematic. Thanks

[mon2]

Re: A1708 820-00840 charging issue

Hi. Let me check my database and will update you. Almost the start of our day here..

[bennX76]

hi mon2
I have the same question like pedro147.
any chance of sharing the sch/brd files that has that USB driver included?
the PI3USB on my board was too hot, so I desolderd it. the macbook works fine without this chip. charging the battery works. just usb/thunderbold doesn't work. i ordered a new PI3USB from ali-express. but i would like to check if there are any other defective parts around the usb multiplexer. the reason for the defective board was a defective charging cable! this cable killed 2 macbooks

thanks
benno

[mon2]

Will recommend to pick up fresh new usb-c paddle boards for this model. They are not that expensive and will be with clean contacts. Do inspect the usb-c parts on the logic board to be sure there is no damage from the shorted usb cable. A good idea to check the usb-c lines including CCx pins in diode mode.

no power -> meter in diode mode. Red meter probe, yes red to ground. Black on the pin to test.

if you read a low voltage like 0v then the line is shorted. Often schematics will offer diode readings to assist in troubleshooting.