A2141 820-01700 Water Damage L8504 Damaged, U3100/U3200 power cycles UB300/UB400 5V .25A

[abajor]

Like title says A2141 came in with Water Damage, customer desperately wants the data.

The actual water ingress doesn't appear too bad it sat for a day but was very dusty internally. I have no idea why L8504 was arced there is no apparent moisture there or around the LVDS cable; however, the LVDS to screen screw studs had popped off at some point and somebody tried securing it with kapton tape.

I had to create a micron wire pad jumper to reconnect the south side of the L8504 since the primary pad was blasted down to the vias.

That said after cleaning up the board only the right side USB would draw 5V and .25 amps, the Left side power cycled.

Inspecting the PCB with the USB power connected to the Right ports shows the only device on the laptop getting warm is the T2 Chip under the black shielding tape.

There doesn't appear to be a short on PP2V5 or PPBUS_G3H.

I swapped U3100/U3200 from an 820-01700 which fails DFU recovery. I am able to get the U3100 Port to behave the same as the right hand ports, the U3200 Port reads 5V 0 amps.

Its possible I didn't get enough solder on all of the pads. I don't have a stencil for the USB control chips so I just tried re applying solder with flux, I noticed some of the pads don't appear to want to take a dome of solder possibly not heating up enough They stay fairly flat compared to the others? I tried putting a layer of solder paste on the cleaned pads, flowing with hot air tool then removing the excess ball that formed with an iron.

Are there any tricks for half assed rebelling of the CD3217?

I am keeping my fingers crossed that this thing isn't in DFU, as always any tips or suggestions greatly appreciated.

[PITERPENY]

You are aware that replacing 3217 is not the same as 3215 and it is not recommended to replace one of the group?! Each 3217 chip has its own firmware according to the installation location (that's what the legends say))

[reformatt]

250ma sounds like the T2 not completing its boot. And you need all CD3217's to communicate to the T2 for it to fully boot. If it's water damaged, I would first reflow or preferably reball each CD3217, rather than replace. Usually it's crap under a BGA on the wings of these boards, rather than a chip failure.

I don't think your reball method would put enough solder on each pad for it to pull into place properly when you solder it on to the board. My preference is balls over paste for reballing. If you don't have a stencil, apply a thin coat of flux to the chip, and place a 0.3 or 0.35 mm solder ball on each pad manually and use low air to flow into place. This takes a bit of practice. You can also use a small heating plate used for removing glue off an IC (like the MaAnt SL-1 heating table for example) to hold the IC into place and pre-heat the chip/balls and then you don't need much air from your hot air tool.

[abajor]

You are aware that replacing 3217 is not the same as 3215 and it is not recommended to replace one of the group?! Each 3217 chip has its own firmware according to the installation location (that's what the legends say))

I literally swapped identical CD3217 chips from same location from a donor board, I was pretty explicit about that.

[abajor]

250ma sounds like the T2 not completing its boot. And you need all CD3217's to communicate to the T2 for it to fully boot. If it's water damaged, I would first reflow or preferably reball each CD3217, rather than replace. Usually it's crap under a BGA on the wings of these boards, rather than a chip failure.

I don't think your reball method would put enough solder on each pad for it to pull into place properly when you solder it on to the board. My preference is balls over paste for reballing. If you don't have a stencil, apply a thin coat of flux to the chip, and place a 0.3 or 0.35 mm solder ball on each pad manually and use low air to flow into place. This takes a bit of practice. You can also use a small heating plate used for removing glue off an IC (like the MaAnt SL-1 heating table for example) to hold the IC into place and pre-heat the chip/balls and then you don't need much air from your hot air tool.

That's what I was afraid of I'll try to get some solder balls, and hand place them with sticky flux. I agree there isn't close to enough solder on most of the pads.
Not sure what you mean by removing glue off the IC like underfill? There's no underfill on the CD3217 as far as I can tell.

Thanks.

[abajor]

250ma sounds like the T2 not completing its boot. And you need all CD3217's to communicate to the T2 for it to fully boot. If it's water damaged, I would first reflow or preferably reball each CD3217, rather than replace. Usually it's crap under a BGA on the wings of these boards, rather than a chip failure.

I don't think your reball method would put enough solder on each pad for it to pull into place properly when you solder it on to the board. My preference is balls over paste for reballing. If you don't have a stencil, apply a thin coat of flux to the chip, and place a 0.3 or 0.35 mm solder ball on each pad manually and use low air to flow into place. This takes a bit of practice. You can also use a small heating plate used for removing glue off an IC (like the MaAnt SL-1 heating table for example) to hold the IC into place and pre-heat the chip/balls and then you don't need much air from your hot air tool.

When trying to buy the balls, they claim "high quality tin" Are they only tin or a eutectic alloy? I do also see good old lead tin. I might just go for that.

[reformatt]

That's what I was afraid of I'll try to get some solder balls, and hand place them with sticky flux. I agree there isn't close to enough solder on most of the pads.
Not sure what you mean by removing glue off the IC like underfill? There's no underfill on the CD3217 as far as I can tell.

Thanks.

The platform is used for that, but can also be used to hold a chip and pre-heat it to make it easier for putting balls on with hot air. That's what I meant.

[abajor]

The platform is used for that, but can also be used to hold a chip and pre-heat it to make it easier for putting balls on with hot air. That's what I meant.

I think I should get a pre-heater haven't investigated those yet.

[abajor]

The platform is used for that, but can also be used to hold a chip and pre-heat it to make it easier for putting balls on with hot air. That's what I meant.

I tried flowing a little tacky flux first to make a thin layer then stuck the balls down, but it's impossible to keep them from moving once the flux liquifies. So that was an exciting lesson learned.

I ordered one of the BGA stencil kits https://www.ebay.com/itm/36340538008...iew_item&gQT=2

Its not clear whether the stencil is thick enough to just use solder paste or if I need to use balls.

I'm a little confused about the "tin" balls I got. Do they use pure tin for repair? The melting point is a bit higher than lead free?

Are there any other surface prep recommendations besides removing the old tin/solder and cleaning the old flux off?

Thanks!

[reformatt]

0.12mm stencils are for paste. These take some practice with as a lot depends on how dry your paste is, and the heat/air flow you use. What will happen is the flux will tend to boil when you heat it and the solder will pop out of the hole. You also need to hold the stencil down flat with tweezers usually (because it will warp when heat is applied, and the paste will form a flat mess across the chip). The magnetic platforms are intended to hold the stencil flat for you but not sure how successful that is.

Direct heating stencils for balls are much thicker and the good ones are made from steel and don't warp. These are way easier if you have a platform kit like the DS-908, but the newer IC direct heating stencils are a bit bigger and not magnetic.

For the others where I must use paste I use this platform and the So Fix MAC stencils.

https://www.aliexpress.com/item/1005...450c1802iGfw1W

Just use the Mechanic 63/37 solder paste, lead free is a PITA. Use the dry stuff at the top or dry it out before using on a stencil. For balls, I only use leaded as well. The Chinese all call solder tin, no matter the alloy.

The small IC's are almost impossible to do unless you use a trick I got off Paul S (ex Rossman repair) by using a sheet of glass over the top of the stencil. Keeps the stencil flat and results in a more even heating effect. The flux will start dispersing outwards from the IC under the glass rather than boiling. Makes it so much simpler if you struggle with paste like I do. The perfect glass is the top of an old A1466 trackpad. Pull it apart and scrape off the silver backing so it's clear. Then use the rough side facing the stencil and IC. Practice with heat/air flow on some spare IC's first to see what works with your setup.

[PITERPENY]

this is a fairly simple chip in reball and it is not worth buying a whole set for this. AMAOE MAC9 stencil would suit you quite well for $5-9. Any mechanic paste 183 degrees.

[abajor]

So this is weird I'm not sure what to make of it. On this board PP2v5_nand_ssd0 was reading as a short.
There was a little corrosion schmutz around U9580 I removed U9580 still saw a short on PP2V5.
Decided I would do voltage injection on the rail. I thought I saw a current limit flash on my psu out the corner of my eye but the rail wasn't drawing any current.

Now I go back to measure resistance without U9580 and it is reading 5M ohm across PP2V5.

I'm assuming I just blasted a little corrosion off somewhere? Got to love catastrophic liquid damage.