Old Macbook Pro 15 820-2101-A / A1226, missing PP1V5_S0

gregwtmtno

New member
Hi,

I'm a total novice at this stuff so please be patient. I took apart a working non-unibody MacBook Pro 15 inch (probably vintage at this point) to reapply the thermal paste and to blow out the dust. When I tried to turn it back on, I got a green light on the charger and the HD spins up, but no fan spin or display.

After investigating, I found that I have all S0 rails except for PP1V5_S0. As best I can tell from the schematic, the rail should be generated by U7600. I measured voltages around there, and things definitely look a bit off, but I'm not sure what they should be. (I’m not able to measure directly on the chip—instead I’m measuring at a different component, hopefully nearby, but sometimes on the other side of the board. Should I be trying to measure at the chip itself?)

For those measurements, I get:

U7600 voltages (clockwise)

Line LabelPin NumMeasured OnVoltage
PP5V_S510C7600, pin 14.96
P1V5S0_TON2R7619, pin 10.08
P1V5S0_VBST14R7615, pin 14.5
P1V5S0_DRVH13Q7620, pin ?12?
P1V5S0_LL12L7620, pin 10
P1V5S0_DRVL9Could not measure
GND7R7611, pin 20
GND15R7611, pin 20
GND8R7611, pin 20
P1V5S0_TRIP11R7605, pin 10.6
P1V5S0_VFB5R76111, pin 10
PP1V5_S03L2700, pin 10
P1V5P1V05S0_PGOOD6U7880, pin 10
PM_SLP_S3_DELAY_L1U7850, pin 43.27
PP5V_S5_P1V5S0_V5FILT4R7601, pin 24.9
Some of these values look suspicious to me, especially TRIP, TON, and VBST, but I don’t know what they should be.

Also, the problem seems to be intermittent to a point. After measuring these voltages, I was surprised to find that the laptop screen came on the next time I connected the power cord. However, I wasn’t able to get it to happen twice. I wonder if it has something to do with the heat of having the board on for a while as I was taking measurements.

Thanks very much for any guidance at all you can give me!



Greg
 

2informaticos

Administrator
Staff member
First of all, welcome to the forum!

Avoid problem measuring directly on the pins of QFN chips.
Very good idea to use surrounding components as test points.

Check/replace R/C7615.
Be sure R7610/11 are good.
Resolder all components and check traces continuity.
 
Last edited:

gregwtmtno

New member
So just to give an update, I was able to remove C7615 from the circuit and on my multimeter, it shows an increase in resistance until it hits open line. I think that means it is working ok. Should I replace it anyway? The resistors are still in circuit but I think they are measuring ok. Measuring across R7615 gave me .9 ohms which I think is close enough to 0. Measuring across R7610 and R7611 both read 5.6kohms which I think shows 2 working 10kohm resistors in parallel. Would appreciate any other ideas. Otherwise I'll start removing resistors (I'm just not very good ad soldering so I'm trying to minimize how much I have to do.)
 

2informaticos

Administrator
Staff member
No need to replace components if you know how to test them and are surely good.
Post resistance to ground at L7620; compare reading with similar board.

Remove U7600 and apply leaded solder on all its pads and board too; then resolder it back.
Do NOT apply power on the board without U7600!!!
Consider to change U7600, if symptom persists.
 

gregwtmtno

New member
On either side of L7620 I get 425 ohms to ground but I have no other board to compare with. I am working on following your other advice, but I need better equipment and supplies, which I've ordered!
 

gregwtmtno

New member
Hi, I measured the resistance to ground on an equivalent part to L7620 on a mid 2010 macbook pro logic board (L7630 on an 820-2850-A) and I only got 16 ohms resistance. 16 oms is pretty different than the 425 I measured on the 820-2101-A. Could this be a source of the problem?
 

2informaticos

Administrator
Staff member
Totally diferent platform.
Some PCHs can have low reading on that rail.

Post actual voltages of U7600.
Be sure to have good contact between pins and test points choosed to measure voltages.
 

gregwtmtno

New member
The situation has deteriorated, but I am not giving up yet! First of all, it now seems that PP1V5_S0 is shorted to ground. I remeasured on either side of L7620 and I get around 0.4 ohms. I plan to see if any capacitors on that circuit get hot to see if they should be replaced. Two more things: The green charger light no longer comes on reliably. Sometimes it works, sometimes it doesn't. Also, I let the "magic smoke" out of Q5315 when I was trying to bridge the power pads and my screwdriver slipped. Oops!

UPDATE: PP1V5_SO is only shorted to ground when the left IO board is connected at J3400. Could this be the problem? A short on the left IO board?

I've measured the actual voltages on U7600:

Line LabelPin NumVoltage
PP5V_S5104.96
P1V5S0_TON20.06
P1V5S0_VBST144.5
P1V5S0_DRVH130
P1V5S0_LL120
P1V5S0_DRVL90
GND70
GND15N/A
GND80
P1V5S0_TRIP110.6
P1V5S0_VFB50
PP1V5_S030
P1V5P1V05S0_PGOOD60
PM_SLP_S3_DELAY_L13.3
PP5V_S5_P1V5S0_V5FILT44.9
 
Last edited:

2informaticos

Administrator
Staff member
Is already explained on the forum, how to simulate the power button (in safe way).
I don't see why complicate with the tweezer.

If LIO board is not connected, you can't get Magsafe light; that is normal.

Do you get voltage on any coil now?
If yes, post them...
 

gregwtmtno

New member
Thank you. I will use that method to simulate the power button in the future. However, LIO is connected at J6900 and at J3400. I believe only J6900 should be needed for Magsafe light. But when J3400 is connected L7620 is shorted to ground, when J3400 is not connected, L7620 measures around 400 ohms to ground.
 

gregwtmtno

New member
Good news! It boots! The problem was that PP1V5_S0 was shorted within the j3400 cable itself. I noticed that the short was present even if the cable was only hooked up on one side, so I replaced it.

New question: The fans are running at full speed all the time. Is this an SMC problem maybe? This is the reason I took the computer apart in the first place. I thought the fans were blowing because of too much dust, but now I think it must be an SMC issue.
 

2informaticos

Administrator
Staff member
Run HWMonitor and check reported temps.
LIO board has temp sensor, check the new flex and board connectors.
 

gregwtmtno

New member
FYI, the entire first page of results on Google for HWMonitor are for other programs named HWMonitor (for Windows). I would refrain from giving this bad advice in the future.
 

gregwtmtno

New member
Thanks very much Piernov. I wish I had known that before I ordered new ones, but live and learn! Regardless, the burned MOSFET doesn't seem to have been essential, as the computer is now working well even without a replacement. The fan problem was caused by my failure to reconnect the GPU temperature sensor.

I'm happy to report that after reconnecting the GPU temperature sensor, the computer is now in good working order. Thank you both for your help and your patience with a beginner!
 
Top