MBP 15" A1398 2014 820-3662-A FANS on full speed & slow machine

[rigway]

Hi Guys,

I've got 2.2GHz i7 Logic board for MBP 15" A1398 Mid-2014 MacBook Pro. Received it with no past history. Did some microsurgery to move iSight adapter to it as it came to me with ripped pads for it and no connector. Have been waiting for a heatsink to arrive to start fully testing when assembled..

Everything seems working: booting / charging, however FANS run on full blast (MAX FAN RPMS)

Thought to check what ASD can point out and it gives no specific clue. It goes to certain point and then machine goes to sleep and does not come back on. Adding a pic in attachments.

Later have been searching around previous threads here and someone mentioned HardwareMonitor software, tried it to check - no issues with visible output TEMPS.. also in attachment.

Any suggestion where to start checking? Got another board from 2014 for spare components - so anything I can replace is available.

Your help is greatly appreciated.

[rigway]

Re: MBP 15" A1398 2014 820-3662-A FANS on full speed & slow machine

Right, it's still a learning process for me. Now looking back on what I've shared here, can see myself that there is not much investigation provided by me to ask right questions. Apologies for that. Improving ; )

Later on I've tried ASD EFI version instead of OS ASD, which apparently is more specific on sensors. Result provided in pics.

EFI ASD reported faults with:
Sensor - Temperature (TH1H) -- Right Fin Stack Proximity - Sensor is reading below the low limit.
Sensor - Temperature (TH2H) -- Left Fin Stack Proximity - Sensor is reading below the low limit.

Further I've checked Hardware Output values:
TH1H Reading: -124°C
TH2H Reading: -124°C

Further I will put down my thought process and would ask community to say if I'm on the right track or troubleshooting? Much appreciated if that's possible.

Here is a piece of schematic for Left, Right Fin Stack Sensor


First I thought if reported (registered due to absence of the actual reading) temp is -124°C - maybe the actual sensor is bad. Replaced three leg little component Q5803 from another Mid-2014 doner board. No result, still same faults on EFI ASD.

Correct me if I'm wrong: As I understand Q5803 reports temperature to U5860, which is "working" with those readings and reports it to SMC, which then sends those "panic" request to the system to get slow and turn on FANS on full RMPS since there is an issue, i.e. TH1H/TH2H sensor is not between 10-60°C. Further I thought maybe this U5860 is not doing its job and could have become faulty - swapped U5860 - same ASD EFI sensors error results.

Further I took a perfectly good working Mid-2015 15" MBP with integrated GPU board (they are pretty much similar) and thought to check few points for continuity to compare. On a good working board I was able to get multimeter beep signal from R5350 to the point which meant to be testing connection with a point underneath the SMC chip (52).

On this Mid-2014 board which I'm working on - there is NO continuity.. Does this mean that SMC either not receiving this data from Fin Stack Sensor Circuit? Could that mean that it needs reballing or is faulty with this specific point?


At this stage I have to admit I've never tried reballing SMC and before getting stencil and everything else required - wanted to check with people who are knowledgeable or may have experienced something similar with MBP boards 2014-2015? Maybe I'm jumping ahead with conclusion, which is "guided" by my noob experience of proper electrical circuit diagnostic approach, which I'm lacking..

Many thanks to anyone, who can spend a bit of their time to reply. Greatly appreciated in advance!

[reformatt]

Re: MBP 15" A1398 2014 820-3662-A FANS on full speed & slow machine

Readings like that indicate the SMC is not getting any data. Provided you have all the correct inputs to U5860, then next step is checking the SMBUS lines. There are two pullups for the clock and data lines, so these both should be sitting around 3V or so. Note this SMBUS also connects to connector J4002 for the camera etc.

As you point out, check continuity to the SMC for both lines. Its more likely to be a track/via issue than the SMC itself. There is a test point next to the SMC for SMBUS_SMC_0_S0_SDA, so check that first. Additonally, you can also check the SMBUS_SMC_0_S0_SDA line to J4002, and the test pad that is above it.

MAC's often use via's under the pad of a component, so corrosion often creates open circuits. Sometimes when you go looking you will track it to a component and once you take it off, you realise the pad is competely gone or it's open circuit to the via underneath.

[rigway]

Re: MBP 15" A1398 2014 820-3662-A FANS on full speed & slow machine

Update and Gratitude! Thanks a lot to you Reformatt!
If I can be thankful to you somehow - just drop me a private message. Honestly saved me from hassle.

Dipped me into my own mistakes ; ))
It's all sorted basically. When you mentioned that J4002 uses same data line/ track I thought to double check what are the pins related to that. Basically due to MBP "panic mode" FANS full blast etc it was preventing me from using/checking camera so I wasn't even aware if I've done my micro-surgery correctly via testing it. It just displayed message - "Allow machine to cool down.."

Anyway.. the issue was - since all pads were ripped off, I've prepared points to attach to after that track. Thought OK it goes there so must do just that : )

Later I've restored the track itself to those two data line points behind the pads AND soldered pins onto those. Surprise, surprise TH1H and TH2H started reporting temps to SMC!

Like I said before.. it's all learning process to me, I like "playing" with microscope and I guess learning curve is necessary.

However, It's just bloody insane! Think Apple I guess.. if your webcam connection is ruined - full machine becomes unusable. Basically It needed fixing anyway just to get it running OK apparently anyway Initially I just thought to do it just to have it all working.

Appreciate your help. Many thanks again. You are my Savior! : ) Otherwise I would have gone totally in a wrong direction with my proposed "troubleshooting".

[reformatt]

Re: MBP 15" A1398 2014 820-3662-A FANS on full speed & slow machine

No problem. When doing board rework, it's important to buzz out all connections points with a meter. Often there are via's under pads etc. FlexBV shows things a bit differently to standard boardview so makes this stuff easier to buzz out and troubleshoot.