Macbook Pro 2018 dead with 5V 0A

[Jiayi]

Hi everyone, this is my first time trying to repair my 15" MBP 2018 A1990. I've been watching logic board repair videos for a long time so I wanted to experiment with reviving my own MBP, board number 820-01814.

This is all the information I gathered so far:

- All USB-C ports are dead with 5V 0A, none of them loops

- Weird fact: when charger is connected to the port corresponding to UB400 and I try to measure the volts at CB401, the charger is restarted. It happens only with that port and UB400. CB401 is not shorted. I don't know enough to explain this.

- PPBUS_G3H measured 0V and shorted when measuring continuity to ground at F7000. F7000 and F7001 fuses are okay though.

- PP3V3_G3H_RTC 0V, however C3100 not shorted. Measured around 440Ω when measuring for continuity

- PP3V3_G3H_RTC_X also 0V



Today I got a simple PSU from Amazon so I tried injecting 1V 4A into the board from F7000 using the probs. Volts dropped to 0.7V so around 3W as output. I don't have a termal camera, so I used a infrared thermometer with laser pointer. Found a hotter area around C7605. Used isopropyl alcohol around that area then. It was clearly visible that C7605 was getting hotter faster then the rest of the area, even hotter than C7607 next to it. However, when measuring for continuity, both C7605 and C7607 seemed okay, i.e. no continuity and resistance around 440Ω. So they're not shorted right?

I thought it might something else on the opposite side. PPBUS_HS_3v3GH_T is definitely shorted on that side. C7680,C7660,C7661,C7662 all shorted to ground. But they don't seem to get particularly hot (isopropyl not evaporating faster) as C7605 does.

What should I do next?

[mon2]

Hello and welcome to the forum. Post exact resistance measurements for a review.

450 ohms is low but not a short.

what is the exact resistance measurement at the fuses? Do note that often there are enable pins for the power rails. So it is possible that the current sense logic decided the rail is under excessive load so turn off the power rail. Review the resistance to ground and also observe the enable pin voltage to ground for the same power rail under review.

each type c connector is linked to a power delivery controller (Texas Instruments ACE nicknamed). Each is a standalone micro with firmware. Each ace controller must be functional to dial up the 20 volt power delivery contract from the power adapter.

be sure the type C connector itself is clean without cat puke, etc.

post the exact current draw on your type C meter. There must be some current draw in the mA range?

post a pic of your meter used to confirm it is a bidirectional type.

[Jiayi]

Hi, thank you for the reply.
- Fuses resistance are 0.7Ω
- Not sure if I read the board view correctly, for the enable voltage I measured PPDCIN_G3H_CHGR at C7029 and C7042. Measured 5.17V. Will attach board view of that page.
- Tried to measure resistance of R7061 and R7062 as suggested here and measured 1.2Ω at one side (CHGR_CSO_R_P) , 2.3Ω at the other side (CHGR_CSO_P). Was not able to measure resistance of R7021 and R7022, will retry later maybe.
- Type C connectors look good. I always tried to keep my MBP in good conditions, no water etc. Indeed board looks good everywhere, no evident damaged capacitor or corrosion area
- Current draw is exactly 0A. Will attach photo.

This is the meter I'm using: https://www.amazon.it/dp/B00TM0W8ZY

[mon2]

I think that you are measuring the resistance across the fuse F7000 and F7001? Instead, remove all power and measure the resistance to ground of F7000. One meter lead at F7000; other meter lead to ground. We know from the last post with the low resistance that your fuses are not blown.

[Jiayi]

Resistance to ground of both F7000 and F7001 are 0.6/0.8Ω, but also resistance across is the same 0.6Ω roughly. (I've set the multimeter to 200Ω)

[mon2]

Ok. Often this means a shorted cap downstream but let us confirm.

carefully flux and remove the 2 fuses, F7000 and F7001 off the board.

then without power, measure the resistance to ground of each side of the removed F7000 fuse pcb pads . Post each measurement.


reference reading:
https://old.repair.wiki/w/A1990_2018...ar_MacBook_Pro

[Jiayi]

Dumb question: so I got my flux today, but I was not able to desolder the fuses with the temperature set at 400°C. With the thermometer with the laser pointed at the fuses, I would get at most 110°C. Stopped when I was hitting it for 2mins straight and noticing the plastic wraps of the memory chips starting degrading. I also noticed the heat dissipation pipes getting hot as well. Should I had them removed before trying to desolder? So that heat doesn't get dissipated.

[mon2]

Suggest highly to practice the process on a donor / defective board from ebay, etc. You should protect nearby parts that can have a meltdown with aluminum foil and/or kapton tape. Hot air will deform plastic and nylon parts.

The air pressure should be just enough to melt the solder joints - go in tight circles. Too high pressure will melt outside parts from this region. Do view the many videos on YT on this topic.

The key issue here is that the PCB is multilayers - probably 8 layers. Inside the buried layer is a large ground plane of copper which acts like a heatsink so the heat you are applying is being dissipated by the hidden ground plane. It should be possible to still remove the part. Best again to practice on a do not care board to get a feel of the capabilities of your hot air tool and the process. You may have to dial up the heat to a higher value but do not burn the board (gauge with the color of the solder mask color) - with low heat air pressure. An underside heater is great and it would help to warm up the logic board.

[Jiayi]

I guess I don't have the skill. Tried with some microcontroller boards and was easily able to desolder capacitors, but I just can't get it moving at all with the MBP board.
I've first removed the heatsink pipes over the CPU to remove dissipation. Tried with first preheating the board at 80°C with the hot air gun, then max air pressure and 480°C, but the fuses just won't budge. Tried even with soldering iron at 360°C with flat tip. Won't budge either, just did some mess with the solder that I was fortunately able to clean away. Used flux in every try. I don't have an underside heater unfortunately.

At this stage I'm considering myself lucky that I didn't ruin the fuses despite by bad attempts. Is there any other step that I can do to investigate the short, that doesn't require desoldering the fuses? Like injecting 1V to see where the short is. I hope the shorted component will be easier to desolder, otherwise I guess I'll accept my defeat.

[reformatt]

Most common fault I see on A1990 are the caps getting corroded and shorting near the fan intakes. Or the SSD regulator going short and pumping 12V into the SSD's. First thing I do is check resistance to ground on L9080/9081, then check caps at the edges. Have seen bad trackpads too cause low impedance on PPBUS.

Check C9081, C9087 etc as these are the most common.

MAC boards have large ground planes so will suck up a lot of heat especially in the middle of the board. Using too small a nozzle is often a cause of inadequate heat delivery. You should be using 5mm at least and 400 is way too low on a board like this.

[piernov]

Make sure to use a good soldering station and a good hot air station. The cheap stuff just won't be enough on these boards.

[PITERPENY]

I shoot with a hot air gun at 420C (without heating) and a little flux. nozzle 5mm or 9mm. (3mm is useless without heating). air flow 30-50%.

[Jiayi]

Thanks to all of you for the tips. I was able to find the shorted cap indeed!
Using a 9mm nozzle, max air pressure + some flux was very effective to desolder any component other than those two fuses at the center of the board. But actually keeping them soldered was useful as I injected 1V and 3-5A and tried to find where it was hot. It was very difficult and I had to take some wrong attempts, but eventually I was able to notice a barely noticeable warmer temperature around C6503 near the right audio J6500. It was also indeed in the area where I initially noticed a warmer capacitor C7605 as stated in my first post! I wasn't really able to tell exactly which one was shorted despite the usage of alcool, so I desoldered at once the several capacitors in the area and one of them was shorted indeed! It's one of those with 22UF and 0805-1 as spec.

Re-testing F7000 and F7001 proves that PPBUS_G3H isn't shorted anymore.

Now I need to buy a replacement cap. I think I'll buy several copies of each that I desoldered, they basically have similar specs like 22UF. I just have to find where I can buy them online in Italy/EUrope. Then I hope I'll be able to solder back the poor UB400/CD3215 that I thought was shorted.

Thank you again for the support, I'll keep you updated if I'll be able make the MBP live another day!

[PITERPENY]

ADVICE! You shouldn't buy 1 capacitor or other small part, buy any non-working or blocked MacBook 2015-2023 board on a local marketplace for $10-30 and you will receive a huge set of spare parts for future repairs. but before buying, check the diagram so that you have the parts you need now.

[Jiayi]

Quick question: how do I choose which cap to pick as far as they seem all to have the specs specified by Apple?
E.g. https://www.mouser.it/c/passive-comp...%20dc=25%20VDC

The spec on the board schematics is 22UF 20% 25V X5R-CERM 0805-1.

[mon2]

Good. It is a bulk capacitor and there are others on the same power rail. Therefore this removed defective capacitor will not be missed.

[Jiayi]

I'll buy a donor board then as suggested. My biggest worry now is how to get the CD3215 IC back to the board. I need to either trust buying it online, already reballed, or learn to reball it myself right?

[PITERPENY]

[QUOTE=Jiayi;n3167362]I'll buy a donor board then as suggested. My biggest worry now is how to get the CD3215 IC back to the board. I need to either trust buying it online, already reballed, or learn to reball it myself right?[/QUOTE


If I understand, you are afraid that there is a chance to buy a board with 4 (minimum 2) non-working
CD3215A ? I think it's minimal. Of course, if you have the opportunity to order everything new and take 99% new! but if it's not just a capacitor and I'll have to wait months for a new one to be repaired, I'd buy a used board and throw in parts from it. and re-solder
The CD3215A is very simple, it is not a processor and temperatures at 420C will not damage it. I could be wrong and everything you do is at your own peril and risk. Maybe someone else can give some advice, but ordering one component at a time by mail is not normal)

[Jiayi]

Small update. I followed the tip and bought a cheap A1989 donor board (2018/2019 MBP 13") and got the donor caps. Now the board is taking 5.2V and 0.27/0.25A, so not completely dead anymore. It doesn't work yet because I have not yet restored the CD3215 chip.

But meanwhile I've tested that PPBUS_G3H now has 12.26V, not 12.6V, which is expected since one UB400/CD3215 is missing and thus T2 is not booting (AFAIK). I've also checked that PP3V3_G3H_RTC_X has 3.35V measured both in CB300 and CB400. So all seems good.

The remaining issue seems indeed to be the UB400 chip. Other than putting it back to its place, I've unfortunately knocked off some parts next to it during the removal. Now I'm more careful and I use some kapton tape to keep the surrouding in place and limit heat exposure, but the damage is done and I had to remove all the displaced parts. So my question is: do I have to put all of them back? Looks a annoying work due to their size. They are not marked as critical in the schematic but they seem to belong to the PP3V3_TBT_T_SX line, which is input of the UB000 thunderbolt controller. I'll attach some image.

The CD3215A is very simple, it is not a processor and temperatures at 420C will not damage it

That's good news, thanks for the info.

EDIT: with lots of patience, I've resoldered the tiny resistors and caps back the to board.

[PITERPENY]

Small update. I followed the tip and bought a cheap A1989 donor board (2018/2019 MBP 13") and got the donor caps. Now the board is taking 5.2V and 0.27/0.25A, so not completely dead anymore. It doesn't work yet because I have not yet restored the CD3215 chip.

But meanwhile I've tested that PPBUS_G3H now has 12.26V, not 12.6V, which is expected since one UB400/CD3215 is missing and thus T2 is not booting (AFAIK). I've also checked that PP3V3_G3H_RTC_X has 3.35V measured both in CB300 and CB400. So all seems good.

The remaining issue seems indeed to be the UB400 chip. Other than putting it back to its place, I've unfortunately knocked off some parts next to it during the removal. Now I'm more careful and I use some kapton tape to keep the surrouding in place and limit heat exposure, but the damage is done and I had to remove all the displaced parts. So my question is: do I have to put all of them back? Looks a annoying work due to their size. They are not marked as critical in the schematic but they seem to belong to the PP3V3_TBT_T_SX line, which is input of the UB000 thunderbolt controller. I'll attach some image.



That's good news, thanks for the info.

EDIT: with lots of patience, I've resoldered the tiny resistors and caps back the to board.

Having a donor board with a bunch of spare parts, I would return everything to its place, it's not that difficult, that's what I would do! and also look at the startup algorithm for this chip. https://logi.wiki/index.php/CD3215_Bootup_Sequence

[Jiayi]

I'm back to just say thank you to all of you. The Macbook is live again!
After putting back the knocked off caps and resistors and replacing the bad CD3215 with a working one (easier said than done), the macbook is finally back at 20V and booting up.

This link also helped me a lot with better understanding exactly if the CD3215 was bad (I succeeded only with the 3rd replacement I got from eBay) along with the above one for CD3215 Bootup:
- https://logi.wiki/index.php/CD3217_a...er_on_Sequence

On the journey I've ruined:
- the usb-c connector connected to the bad CD3215 due to continuously working on that port
- the right audio port. Desoldered and melted the plastic a bit when working on the near shorted cap
- the touch id connector: torn cable, very fragile
- the display board: one screw securing the bracket for the display board cable connector (https://www.ifixit.com/Guide/MacBook...122939#s237415) detached itself from the display board ruining some traces apparently, right when I was mounting back the board after it was revived. Ffs.

But they're minor problems compared to a bad CD3215 or not-powering board. At the end of this journey, I really learned a lot about the logic board and CD3215 booting. If anyone encounters similar issues, ping me! I'll be happy to help.