A2159 CPUs are trash, might be a failing CPU. Of course, messing with ROM will create unusual issues, but since you mentioned restore issues beforehand, it's possibly a failing CPU. T2 could be suspect, but it typically won't be affected by safe/non-safe boot conditions (as its work is largely 'done' once power is achieved).
Try reflowing the CPU and the RAM and see if conditions change. Sometimes you get lucky on these.
ugamazing
Senior Member
Last Activity: 06-05-2025, 09:05 AM
Joined: 06-26-2013
Location: georgia
-
Glad you got it sorted! Very common issue a few years back as many parts vendors did not specify or differentiate 2020/M1 vs 2022/M2. Lot of headaches for retail guys.Leave a comment:
-
To confirm...you're using a proper 2022 M2 A2338 trackpad/cable and NOT a 2020 A2338 M1 trackpad/cable, correct? Check part numbers to confirm. Big issue with parts vendors.Leave a comment:
-
You can find good chips there, but you'll also find bad ones. Best to source from a parts board if you can; the 02020 boards can have the controllers replaced individually if you pull from the same model/etc. There are plenty of good sellers on eBay for chips as well!Leave a comment:
-
Hey Mon, I sent Piernov a PM a week ago and haven't heard back. I wonder if it didn't go through?Leave a comment:
-
The important number will be the part number, i.e. "339S00467" etc. They MUST MATCH or the board won't power/boot/restore/etc. I believe on the 01700 2.3/1TB boards, they are largely the same but different RAM configurations will use different chips. For boards like the 00850, they had TONS of different T2 chips--maybe 10 or so. Super critical to match part numbers or you'll go crazy.Last edited by ugamazing; 03-06-2025, 10:10 AM.Leave a comment:
-
Yes that's a supported chip, should work fine once programmed and properly positioned....Leave a comment:
-
I took his "glowing on one side" comment as a T2 short, as that's not a usual thermal print for that chip (and if it's rails are cycling but it *still* heats up quickly enough to catch via thermal?). But yes if ppbus is stable then T2 may actually be fine, but damn what's left? Bad CPU/GPU on these rarely prevents DFU/restore, and he's replaced SSD/T2/PMU/ACEs. Maybe a continuity fault? Maybe start checking your rails on all areas of the board just to be sure, maybe it was bent at some point?...Leave a comment:
-
Make sure you check for SSD damage/death--super common on these and will give you 20v/low amps with trackpad click. Make sure the trackpad clicks fully for 30 seconds and make sure board isn't actually cycling on ~21 second intervals (a very common indication of SSD issues). If SSDs are toast, you may not be getting the data from them, so check carefully.Leave a comment:
-
With T2 cycling and glowing, it's probably the culprit in this case. Just wanted to confirm the SSD wasn't the culprit, but now you've confirmed: Bad T2. You can replace it, but you'll need an unlocked donor etc. Also need to make sure you replace with identical model T2 (confirm part number--NOT just pulled from another 01700 as they're not all identical)....Leave a comment:
-
No, you cannot install nands that are of a different model than the ones native to the board. You will need to check and confirm the part-numbers of the nands on the 01700 board, then source those. They must be compatible with the controller in the T2.Leave a comment:
-
Yes if you have a full proper set for the board you're installing on, and you write the correct dump to each nand and place properly, yes it will work. But you MUST confirm the same generation-nand chips between each device. Compare part numbers and go from there.Leave a comment:
-
Yes! We've replaced maybe 10 more of these 01041/01700 WiFi cards since this post; process not bad at all once you do it once. We pull from dead boards and re-ball to save a bit on part cost.Leave a comment:
-
Check SSD voltage and inspect for liquid to the power supplies. Common failure on these.Leave a comment:
-
I have some new good information for this issue, for any others:
We took four 820-02016 boards that were afflicted by this same issue--no history liquid/drop, but with a sudden exclamatory symbol without the ability to boot to anything (IR/OS/etc). We decided to just wholesale figure this out, so we did the following:
1. Reflowed the CPUs on ALL FOUR boards; this RESOLVED the issue (for ~36 hours so far) on TWO of the boards.
2. On the two boards that the issue persisted on, we replaced the SSDs. NEITHER BOARD resolved after this.
3. We took one of the CPUs...Leave a comment:
-
Apple M1 Boards & Exclamatory Error?
We've now created a stack of 10 boards that have the same issue, which we suspect to be CPU-related. The issue is very simple--the board will not boot to anything, only showing the exclamatory sign inside a circle. No amount of DFU/restore, wipe/erase, etc, will resolve the issue. No boards have physical damage. We went ahead and replaced the SSD nands on two boards (one 02020, one 02016) and the issue remained exactly the same. We are considering replacing the CPU on one of the boards using a locked CPU, just so we can see if it resolves the issue.
Has anyone else run into this... -
T2 rails stable and not cycling? Make sure you check for a solid 30 seconds. Same with PMU. Is the .25a stable and consistent as well? 12.3v screams T2 or PMU, so you're on the right track. Check diode of the T2 rails closely, one might be slightly higher or lower. Also of course confirm CPU isn't shorted to ground--common on the 01958.Leave a comment:
-
Heads up about the 820-01958 boards specifically:
We have found several cases of the CPU controller (U7100) being improperly soldered. So far we've ONLY found this issue to affect the 820-01958. Symptoms are almost always power and DFUs OK but no video or chime ever. Reflow U7100 and make sure you're not dealing with that issue before you mess with T2.
Also, something else worth noting: If you're DESPERATE for the data, you can swap the nands and T2 to a working board to retrieve the data. We've done it a hundred times.Leave a comment:
-
Update on this one: I actually noticed we had used an incompatible ISL (9900); I replaced with the correct one (9901) and the symptoms changed slightly; now PPBUS_AON cycles from ~11v to 12v, over and over. Not sure if this matters with regards to diagnosing this specific issue, but wanted to include it.Leave a comment:
-
820-02918 PPBUS Cycling, Minor Liquid.
Working on an 820-02918 board that had a minor liquid spill, but it was directly to the ISL/U5200. Essentially the entire area had to be replaced due to corrosion. Got it all cleaned up and everything looks good and checks out, but we're hitting a cycling PPBUS issue now. PPBUS_AON cycles quickly from ~9v-12v, indefinitely.
Knowing this is typically a sign of a possible short on a lower rail, we checked all around and found no obvious damage or shorts. We have VERY LIMITED parts for this model (exactly 1 single U5200 chip), but I suspect it will end up being a continuity issue or...
No activity results to display
Leave a comment: