Retina Pro 820-3476 Low PPBUS_G3H

[Flippers McCoy]

Hey everyone!

TL;DR: PPBUS_G3H is reading 12.6V when it should be reading 18.5V, I don't really know how to test it.

I've attached the schematics with some notes on page 51 and 52, I suspect that the bad component is somewhere there, but (as you'll gather from the notes I made), I don't even know where PPBUS_G3H gets 18.5V from, so I don't know where to go from here.

Any advice would be greatly appreciated, thanks!

[Flippers McCoy]

Re: Retina Pro 820-3476 Low PPBUS_G3H

Also, all the diagnostics I've been doing have been with the bare board on my desk running on the adapter (which, when plugged in, starts with a green light, then switches to orange)

[BlueMidnight]

Re: Retina Pro 820-3476 Low PPBUS_G3H

The inductor L7130 is part of a buck converter circuit. Buck converters produce a smaller output voltage from a bigger input voltage. They always consist of 2 switches of some sort (usually 2 MOSFETs), something to control the switches (in this case, U7100), an inductor, and one or more capacitors on the output. I suggest finding things to read or watch about how buck converters work. They are very common and important to understand.

The 12.6V you are seeing on PPBUS_G3H is exactly what it should be. It is PPDCIN_G3H_CHGR (the supply voltage for the buck converter, which comes directly from the AC adapter) that is supposed to be 18.5V.

Regarding D7005... This allows the PPVIN_G3H_P3V42G3H rail (which is the supply for the chip that produces the 3.42V rail) to come from either the AC adapter or the battery. The diodes prevent current from following in one of those paths and out the other.

The 5V/12.5V you are getting near C7126 is the voltage that controls Q7155. When it is 5V, that means Q7155 is turned on to allow the system to run from the battery or to allow the battery to charge.

PPBUS_G3H is the main power rail for board. If the AC adapter is connected, it will be stepped down to 12.6V from the AC adapter's 18.5V. That 12.6V is also high enough to charge the battery. If the AC adapter is disconnected, it will equal the battery voltage, which is less than 12V.

The problem with the board must lie somewhere farther down the line.

[Flippers McCoy]

Re: Retina Pro 820-3476 Low PPBUS_G3H

The inductor L7130 is part of a buck converter circuit. Buck converters produce a smaller output voltage from a bigger input voltage. They always consist of 2 switches of some sort (usually 2 MOSFETs), something to control the switches (in this case, U7100), an inductor, and one or more capacitors on the output. I suggest finding things to read or watch about how buck converters work. They are very common and important to understand.

The 12.6V you are seeing on PPBUS_G3H is exactly what it should be. It is PPDCIN_G3H_CHGR (the supply voltage for the buck converter, which comes directly from the AC adapter) that is supposed to be 18.5V.

Regarding D7005... This allows the PPVIN_G3H_P3V42G3H rail (which is the supply for the chip that produces the 3.42V rail) to come from either the AC adapter or the battery. The diodes prevent current from following in one of those paths and out the other.

The 5V/12.5V you are getting near C7126 is the voltage that controls Q7155. When it is 5V, that means Q7155 is turned on to allow the system to run from the battery or to allow the battery to charge.

PPBUS_G3H is the main power rail for board. If the AC adapter is connected, it will be stepped down to 12.6V from the AC adapter's 18.5V. That 12.6V is also high enough to charge the battery. If the AC adapter is disconnected, it will equal the battery voltage, which is less than 12V.

The problem with the board must lie somewhere farther down the line.

Ah, thanks for enlightening me!!
I'll check out some other rails and post updates

[Flippers McCoy]

Re: Retina Pro 820-3476 Low PPBUS_G3H

So I checked all the G3H and S5 rails, and here's what I'm getting. I put in bold the rails that I think are particularly relevant. I also checked one or two others that I've seen being important on Rossmann videos, so there's that too.
One thing I'm having trouble figuring out is that on the Power Aliases page, the schematics say that PPBUS_S5_HS_COMPUTING and PPBUS_S5_HS_OTHER5V should be 8.6V, but when I look at where they get power from in the schematics I see no reason why them being at 12.5V is a problem, or why it would ever not be that way. So why does the Power Aliases page say 8.6V?

PP5V_S5 is kinda freaking out, it goes from ~3V to ~11V and is just constantly fluctuating
PP3V3_S5 is also doing the same thing, but with different numbers
PPVCC_S0_CPU is a solid 0V
PPBUS_S5_HS_COMPUTING is reading 12.5V, the schematics say it should be reading 8.6V
PPBUS_S5_HS_OTHER5V is reading 12.5, the schematics say it should be reading 8.6V
PPDCIN_G3H_ISOL is reading 18.92V, which is nice
PPDCIN_G3H is reading 18.92, which is nice
PP3V42_G3H is reading 3.41, which is nice
PPVRTC_G3H is reading between 3.30V and 3.41V, which is nice

Any pointers on what to check next would be greatly appreciated!
(If needed, I can provide video of what I'm reading on PP5V_S5 and PP3V3_S5)

[piernov]

Re: Retina Pro 820-3476 Low PPBUS_G3H

Constantly fluctuating with random values sounds like you're not making good contact, either on the coil or on the ground point. If you're sure you're making good contact, then you may have an issue here (but it would be very uncommon).
PPBUS_G3H (and related rails) reads 12.6V on Macbook Pro, and 8.6V on Macbook Air. Sometimes the schematics have the incorrect value, but that's not an issue.

[Flippers McCoy]

Re: Retina Pro 820-3476 Low PPBUS_G3H

I double checked that I was making a good contact, and it was definitely not "user error." It's also not an issue with my meter, I can get solid readings elsewhere.
Any tips on where to check for a cause for that?

[mcplslg123]

Re: Retina Pro 820-3476 Low PPBUS_G3H

is F7140 ok?? Are you getting 12v on both sides??

[Th3_uN1Qu3]

Re: Retina Pro 820-3476 Low PPBUS_G3H

With the board disconnected from power, check for short circuits with your multimeter on S5, S3 and S0 3V and 5V rails. Report back.

I suspect the reason for your fluctuating PP3V3_S5 and PP5V_S5 voltages is a short further downstream.

[Flippers McCoy]

Re: Retina Pro 820-3476 Low PPBUS_G3H

is F7140 ok?? Are you getting 12v on both sides??

Looks like F7140 is fine, I get continuity, and 12.56V on both sides

With the board disconnected from power, check for short circuits with your multimeter on S5, S3 and S0 3V and 5V rails. Report back.

Thank you so much for the tip!

5V_S0 and S5 are fine, 3V_S0, S3, S4, and S5 are fine, none of those got continuity to ground.

PP5V_S4 got solid continuity to ground at several test points, looks like you were right!
How can I go about diagnosing where the short is, and how did you know to check this?

[Th3_uN1Qu3]

Re: Retina Pro 820-3476 Low PPBUS_G3H

Experience.

There are two ways to check for shorts. One is to have a really fancy multimeter with a very high resolution on DC resistance, and then you would scan the rail until you get to the component with the lowest resistance, the second, easier, faster and cheaper is to put some power into it and see what gets hot. Since you may have a short heading to a chipset or something, you will want to start with the voltage very low, say 0.7 volt, and increase slowly until you can feel something getting hot.

Do you have a bench power supply?

[Flippers McCoy]

Re: Retina Pro 820-3476 Low PPBUS_G3H

Aw shucks, it had to get to that at some point, didn't it!

I don't have one of my own at the moment, but I may know a guy. I will keep the thread updated with my current progress
I should be able to inject voltage anywhere on that 5V rail, right? If I see no effect, I'll just slowly increase the voltage no higher than 5V (although I would hope to see an effect before then).

Again, thank you so much for helping me out, I really appreciate it. I'm learning a ton!!

[Th3_uN1Qu3]

Re: Retina Pro 820-3476 Low PPBUS_G3H

In this particular case, you can use an ATX power supply and pull power from a 3.3V wire off of that. Since the short is on a 5 volt rail, this should be reasonably safe.

[Flippers McCoy]

Re: Retina Pro 820-3476 Low PPBUS_G3H

I'm working on making a bench PSU I can use now and in the future
Just to confirm, so I don't fry anything, all I really have to do is determine which wire is the 3.3V rail and touch that to a connection on the board that is part of the 5V rail, right? I don't have to worry about putting any ground wires anywhere?

[Flippers McCoy]

Re: Retina Pro 820-3476 Low PPBUS_G3H

I finished my bench PSU, used the 3.3V wire first, both on the adapter and off, nothing got hot to the touch or anything like that, same with the 5V. After that I tried testing around with my meter using 5V and everything read 5V, except for L4807, which was jumping all over the place. When I tried the same thing with the 3.3V wire it got a solid 3.3V on both sides.
Should I try swapping out just for the heck of it?

[Th3_uN1Qu3]

Re: Retina Pro 820-3476 Low PPBUS_G3H

You need to also connect the ground wire to the ground of the board (obviously), but it sounds like you already did that.

Yes, do replace L4807, it could be that it got damaged. These ferrite beads also act like fuses sometimes, maybe someone did something dumb like short the touchpad flex.

[Flippers McCoy]

Re: Retina Pro 820-3476 Low PPBUS_G3H

Sorry about the super late response, got a lot on my plate :/

Anyway, I replaced L4807, and there is no change. When I was mucking about with injecting power into PP5V_S4, at first I didn't attach my PSU ground to the board ground, then I realized how boneheaded that is and grounded it. Could this have caused more damage?
PP5V_S4 is still getting continuity to ground, so I'm going to try injecting power (properly) again with my slightly fancier PSU that I set up
One technique question: What is the best way to hold the PSU ground to the board, the PSU power to the board, AND feel around for any hot components? The only thought I had would be to use alligator clips to connect the ground.
Thanks a bunch

[Flippers McCoy]

Re: Retina Pro 820-3476 Low PPBUS_G3H

Ahha I think I know what's going on.
Looks like the short to ground on the board is messing with the PSU failsafe, so when I try to inject power into the rail, which is shorted to ground, my test PSU says "nah" and quits on me. Which would explain why on my meter it looks like the PSU is working fine, but when I connect 3.3V or 5V to the board nothing happens!
I may have to see if I can get my hands on a real variable PSU

[Flippers McCoy]

Re: Retina Pro 820-3476 Low PPBUS_G3H

I got my hands on a legit variable PSU, and it didn't really help much to be honest. I kept the voltage fairly low, between .8V and 2V, and started with a pretty low current before cranking it up to as high as the PSU would go. I couldn't feel anything on the board heating up at any point, and whenever I hooked up the variable PSU to the shorted rail the voltage reading on it would drop to 0.
Any tips on where I can go from here?
Thanks

[Flippers McCoy]

Re: Retina Pro 820-3476 Low PPBUS_G3H

Bump