A2251 820-0194 CD3217B12 (U3100_W) donors/replacements

[abajor]

I have an A2251 820-01949 which had a failed TPS62180.
I was fortunate enough to find the video on "Exploding USB-C" https://youtu.be/CAGnMR1aVuE?si=O1oYvAyNMeAr4dv4 what was shared here on the topic.

I was wondering if there are any cross compatible models that have been discovered?
We have an apple ID locked A2179 Mac AIR it's CD3217B2 has different lithography and looks slightly gray but I'm assuming it's the same chip in principal, but I am also aware that for some inexplicable reason mac flashes these chips firmware depending on their location on the board.

This makes me suspect that I might not be able to just find a suitable replacement on Aliexpress, and I will likely need to buy an A2251 donor board.

Also I did pull the CD3217B2 from the A2179, and I'm having a lot of trouble trying to reball it with flux and leaded solder, some of the pads don't appear to want to take a dome of solder and just stay flat and gold.
Any idea why this maybe? I'm haven't had much luck getting USB chips to work when I ball the pads on the PCB instead of the chip, and they tend to float around and out of position much easier this way.

[mon2]

https://youtu.be/7cNg_ifibCQ?feature=shared

https://youtu.be/CAGnMR1aVuE?feature=shared

[reformatt]

I have an A2251 820-01949 which had a failed TPS62180.

Also I did pull the CD3217B2 from the A2179, and I'm having a lot of trouble trying to reball it with flux and leaded solder, some of the pads don't appear to want to take a dome of solder and just stay flat and gold.
Any idea why this maybe? I'm haven't had much luck getting USB chips to work when I ball the pads on the PCB instead of the chip, and they tend to float around and out of position much easier this way.

I did have this a lot with SMC's off old boards, especially if you wick off the old solder first. Some pads would not take, even though they weren't oxidised.

So what I do now is I don't wick at all. I drop my iron temp to 270 degrees C, use flux and leaded solder on the pads for the chip to see which ones take and which ones don't. On the ones that don't, first check to see if it's oxidised and scratch it up a little if so. Then put a small dab of solder paste on it and then heat with hot air at low air flow. The paste should form a small ball and stick to the pad. Do this for each pad till all are uniform and run the iron over it again with a dollop of solder and flux. At this point you should have the same amount of solder on each pad. Clean it, and you can now put it in your stencil and re-ball. I use balls not paste and a direct heat stencil. The extra solder on the pad doesn't make that much difference but you can go slightly smaller on the ball size if it's a concern. I use the DS-908 kit for my reballs.

[abajor]

https://youtu.be/7cNg_ifibCQ?feature=shared

https://youtu.be/CAGnMR1aVuE?feature=shared

I think I beat you to the punch on that second video in my initial post.

Good news is after removing TPS62180 there is no short, bad news is the original U3100_W gets very hot as the laptop starts to power up. So something is damaged there. Otherwise it's pulling 20V and appears to be cycling between .6A and 1.7A

[abajor]

I did have this a lot with SMC's off old boards, especially if you wick off the old solder first. Some pads would not take, even though they weren't oxidised.

So what I do now is I don't wick at all. I drop my iron temp to 270 degrees C, use flux and leaded solder on the pads for the chip to see which ones take and which ones don't. On the ones that don't, first check to see if it's oxidised and scratch it up a little if so. Then put a small dab of solder paste on it and then heat with hot air at low air flow. The paste should form a small ball and stick to the pad. Do this for each pad till all are uniform and run the iron over it again with a dollop of solder and flux. At this point you should have the same amount of solder on each pad. Clean it, and you can now put it in your stencil and re-ball. I use balls not paste and a direct heat stencil. The extra solder on the pad doesn't make that much difference but you can go slightly smaller on the ball size if it's a concern. I use the DS-908 kit for my reballs.

I'll have to give that a try. Haven't bought any reballing kits. Right now I just have some MG-chemicals lead free solder paste. I normally just use that for assembling new boards with their own solder masks.

[abajor]

I got optimistic for a second but PP2V5_NAND_SSD0 is shorted to ground. RIP, how has mac not been majorly sued over this?

[mon2]

That was the take away conclusion from the linked videos. Both are worthy of a view. IMHO it will only be a matter of time before another class action lawsuit is triggered from these faults.

[abajor]

That was the take away conclusion from the linked videos. Both are worthy of a view. IMHO it will only be a matter of time before another class action lawsuit is triggered from these faults.

Don't tell that to the Mac Enjoyers
https://www.reddit.com/r/MacOS/comme...es_chips_m123/

[mon2]

OMG - just read the entire thread. Some are just clueless. They will only believe it once their macbook is fried.

Reminds me of a time in our company when we thought about writing custom software for corporations. Paid some fee to the Canadian government to enroll in the OHIP (doctor) automated billing program. After reviewing the details and before throwing real money at it, decided to pickup the phone and dial up local doctors for their feedback on the pending software. Most were receptive and requested to be signed up. Then one kook of a senile (female) doctor shredded my privates like a sabretooth tiger. Yelled to say 'donot ever phone her and harass my staff with such lies about how the billing will be automated by a COMPUTER. It will never happen'. Whack job would not allow for one piece of free advice. Needless to say, it is all now automated billing.

To add to:

a) us wanting to build LED brake lights when it was not even being discussed - feedback was; it will never sell. The automotive industry is controlled by the mafia.
b) us wanting to build car logging device to gauge when the driver is driving too fast - feedback was; it will never sell. Which nutcase will allow themselves to be monitored? Umm..insurance companies.

Can say something positive that at least in one case, I ignored the negativity and proceeded forward. Can retire from that ride of our widgets.

We should be thankful that Louis and later iBoff posted the linked videos. Those that are not, can continue to kiss Apple booty.

[reformatt]

I got optimistic for a second but PP2V5_NAND_SSD0 is shorted to ground. RIP, how has mac not been majorly sued over this?

I have been lucky on a few occasions to have a shorted cap on that rail, rather than the converter failing. Usual failure pattern is PPBUS_G3H is shorted to ground via the converter and one of the NAND's. But if the short is only present on 2.5V rail, then it's usually just a cap on the edge of the board.

Unfortunately, Apple is big enough and has such a cult following from their user base that they often aren't held to account. In fact, they have slowly educated their user base into believing the lifespan of their products is expected to be no longer than a few years. I often get customers wondering if they should even attempt a repair a device that is only a few years old. So the motherboard is now seen by customers as a consumable part like a battery and not repairable. That's clever social engineering. So why should Apple care about bad engineering.

[abajor]

I have been lucky on a few occasions to have a shorted cap on that rail, rather than the converter failing. Usual failure pattern is PPBUS_G3H is shorted to ground via the converter and one of the NAND's. But if the short is only present on 2.5V rail, then it's usually just a cap on the edge of the board.

Unfortunately, Apple is big enough and has such a cult following from their user base that they often aren't held to account. In fact, they have slowly educated their user base into believing the lifespan of their products is expected to be no longer than a few years. I often get customers wondering if they should even attempt a repair a device that is only a few years old. So the motherboard is now seen by customers as a consumable part like a battery and not repairable. That's clever social engineering. So why should Apple care about bad engineering.

I suppose I can try to see what heats up when I inject 1V into the PP2V5_NAND_SSD0 rail. If the NAND Glows thats all she wrote?

Will that also cause the issue with CD3217B2 at U3100_W heat up? both the original and the chip from the A2179 are heating up. I get 20V on the two USB circuits on the left side of the laptop, and fan spin with the original U3100_W but those USB jacks stay at 5V ~0.2A with PSU, it does the same with the donor CD3217B2 but no fan spin.

[abajor]

Doesn't appear to be a failed capacitor. The silver coating on the NANDs appear to be low emissivity and the bad one doesn't get very warm to the touch; however the board around the NAND starts to show on thermal. I put Kapton tape over all the NANDS and injected 1V into the NAND supply rail.
One of the NANDS began showing signs of heating. I'm surprised it's hardly enough to feel by touch, it's drawing about 1A at 1V.

I'm going to call that case closed on this poor MBP. I ordered the same model with an apple ID lockout. I think I'll try doing a NAND swap just to torture myself.

[reformatt]

Be sure to watch iBoff's video on NAND swaps for MAC's. There are a lot of gotcha's so this can save you a bit of hair pulling. T2 onboard RAM capacity (which is soldered on top of the T2) comes in 2 configs, which limit the total size of the SSD NAND's. You can also just swap the T2 from an unlocked board to locked one too to get rid of iCloud.

https://www.youtube.com/watch?v=yR7m...annel=iBoffRCC

[abajor]

Be sure to watch iBoff's video on NAND swaps for MAC's. There are a lot of gotcha's so this can save you a bit of hair pulling. T2 onboard RAM capacity (which is soldered on top of the T2) comes in 2 configs, which limit the total size of the SSD NAND's. You can also just swap the T2 from an unlocked board to locked one too to get rid of iCloud.

https://www.youtube.com/watch?v=yR7m...annel=iBoffRCC

Do the T2 chips need to match boards? Or only match the memory size? AKA the 1G and 2G versions? I have an A2141 and an A1990 I can probably pull the chips from.
I tried to remove a T2 chip from the A2251 with the bad NAND flash, and learned the hard way that it's Two chips sandwiched together, that shouldn't be a problem except the pads 738 and 776 popped off.

[PITERPENY]

this is an interesting question! I think that in each T2 the MacBook model and the device boot sequence program are specified! 1g is if the memory is up to 1 TB and 2g is more if as I understand it
but with similar boards on T2 you will need to experiment

[abajor]

this is an interesting question! I think that in each T2 the MacBook model and the device boot sequence program are specified! 1g is if the memory is up to 1 TB and 2g is more if as I understand it
but with similar boards on T2 you will need to experiment

According to the iBoff video they the 1G T2 chip only supports 512GB, the T2 sounds like it supports up to 4TB maybe even 8TB if you repurpose mac pro modules.

[reformatt]

Do the T2 chips need to match boards? Or only match the memory size? AKA the 1G and 2G versions? I have an A2141 and an A1990 I can probably pull the chips from.
I tried to remove a T2 chip from the A2251 with the bad NAND flash, and learned the hard way that it's Two chips sandwiched together, that shouldn't be a problem except the pads 738 and 776 popped off.

The A1989 and A1990 use the same stencil, however A2159 is different. If you move them between boards (and thus different NAND's) you still have to do a DFU restore.

Now not sure about the SOC ROM's when going between models. I haven't played too much with this to experiment, but maybe Piernov might have some info on it.