[HN Gopher] Restoring an Apple Silicon Developer Transition Kit ...
___________________________________________________________________
Restoring an Apple Silicon Developer Transition Kit "DTK" Mac Mini
[video]
Author : theogravity
Score : 81 points
Date : 2024-05-21 00:14 UTC (11 hours ago)
(HTM) web link (www.youtube.com)
(TXT) w3m dump (www.youtube.com)
| shortformblog wrote:
| dosdude1 is perhaps the most interesting hardware hacker out
| there today. The willingness to do something as wild as he does
| here--literally using a file on the board to fix an otherwise
| device-breaking crack--is incredible. Happy he gets to experiment
| with this machine.
| sitzkrieg wrote:
| if you want to see this on a daily basis checkout "electronics
| repair school" on yt, silly name but extremely skilled repairs.
| this guy does stuff that used to take me 4+ hr a 30min yt while
| drinking coffee and narrating
| SpecialistK wrote:
| Not to mention how useful his software tools are as well! The
| MacOS patchers (and now work on OpenCore (Legacy)) have saved
| thousands of machines from an early retirement.
| shepherdjerred wrote:
| I didn't expect to watch that entire video, but it was so
| interesting. Part of me is jealous of how much knowledge and
| skill the author has.
| nxobject wrote:
| Especially what it takes to reprogram and re-pair NANDs and
| Macs together. It's absurd that NANDs not only encode board SN,
| but MAC ID as well.
| oneplane wrote:
| It is pretty interesting how this has evolved over the
| decades. Most hardware has some sort of one-way identity that
| can be stored in a ROM or sometimes just on the mainboard
| (like adding/removing some zero-ohm resistors). But because
| the information has gotten so much more involved over time we
| now have specific areas on almost all hardware that store
| device-specific calibration data but also code for network
| cards to be able to function at all (pretty much all Intel
| PCs have a GbE firmware embedded in the BIOS or UEFI ROM for
| example).
|
| There used to be separate chips for everything as well, but
| that became too error-prone and limiting, so instead of
| having 10 ROMs of varying kinds we ended up with 3 in most
| cases (embedded boot ROM, device ROM and storage ROM), which
| then got slimmed down even more to the point where there is a
| SiP or SoC ROM and either NAND or DFU has to do everything
| else.
|
| Depending on the vendor, some have been adding some 'extra'
| ROMs (usually Flash EEPROMs) for embedded controllers because
| they can't trust or modify the hardware delivered by their
| partners (AMD/Intel) as much as they like, so they do it
| externally (HP SureStart for example). But because those
| controllers themselves also have a boot ROM and an embedded
| config ROM you now end up adding 3 ROMs back into the
| completed product. We even had a short period where the Intel
| ME in the chipset had one as well, so you ended up with at
| least 5 boot ROMs (CPU, CH, GbE, USB, EC) and 2 separate
| combined ROMs (one of which was the combined traditional BIOS
| ROM, the other purely the EC), just to start up a random
| desktop PC.
| 1oooqooq wrote:
| it's almost like the purple shipping binary blobs dislike
| being inconvenienced by other people's binary blobs while
| they ship their binary blobs. shocking.
| rootsudo wrote:
| What was most interesting is that WI-FI is binded to the Mac
| Serial Number, meaning that the mac address even though it is
| able to be manipulated in OSX, gives out so much information
| about who you are and allows very easy:
|
| Mac Address (the wifi serial hw serial number) > Mac Serial
| address > who you are.
|
| It's interesting.
| sitzkrieg wrote:
| nation states love this one weird trick
| Hamuko wrote:
| Doesn't the operating system use a randomised MAC address
| when connecting to Wi-Fi though?
| 1oooqooq wrote:
| for the AP, yes. apple don't want google nest knowing who
| you are.
|
| but the real mac is used in lots more places. even for
| "random" uuid generation.
| Wowfunhappy wrote:
| Can anyone explain to a hardware idiot how he fixed the short
| (where Apple cut the board)? I don't understand why scraping off
| layers like that made it go away. Why can the board still work
| after you've taken out a chunk of it (as long as you take out
| _enough_ chunks)?
| RodgerTheGreat wrote:
| The damage to the board was caused by a tool like a pair of
| tinsnips, crushing the layers of the board together while
| cutting. The tiny layers of copper are folded together at the
| edge of the cut, causing shorts. By sanding away the board at a
| much shallower angle than the original cut it is possible to
| expose an edge along which no conductive layers are bridged.
|
| As for how the board still works when that area is cut out,
| probably just a degree of luck that none of the traces that
| were severed were essential?
| Wowfunhappy wrote:
| Ah, thank you! So it wasn't actually the cut that "broke" it,
| it was the fact it was crushed.
| simondotau wrote:
| Based on appearances, the board has a tight cluster of
| components laid out similarly to a contemporaneous iPad. The
| rest of it seems like a vast empty wasteland, lightly
| populated with debug terminals, ports, and compatibility
| shims which make it work inside a Mac mini chassis.
|
| Chances are the breakage caused a power rail to short and the
| "repair" fixed the short while leaving one or two debug
| points inoperable.
| runjake wrote:
| I believe Wowfunhappy is referring to this section of the
| video: https://www.youtube.com/watch?v=reQq8fx4D0Q&t=422s
| jwagenet wrote:
| When you clip the board, the conductive layers are also being
| smooshed together and touch despite the non-conductive layers,
| which seems to persist during the enlargement of the cut. I
| imagine the goal of the flat filing is to remove more material
| from the top layers an "terrace" the layer lines until they
| don't touch.
|
| As far as the repair is concerned, it seems like the cut only
| goes through continuous ground or voltage planes. So in a
| sense, it is lucky no important traces or logic lines are
| harmed in the process.
| nxobject wrote:
| I'm surprised that it wasn't any closer to any other non-
| power/ground traces.
| gorbypark wrote:
| It's like an imaginary (horribly made) club sandwich!
|
| From the bottom up, there's a layer of bread, then a layer of
| ham, then another layer of bread, then a layer of cheese, then
| the top piece of bread. The ham and cheese are your conductive
| layers, one carrying current the other the ground.
|
| When you first make the sandwich, everything is good because
| the ham and cheese are not touching each other. However, when
| you cut it in half, the bread, ham and cheese all smoosh
| together at the cut point and the ham and cheese electrically
| short out. Now, if you are super careful, you can probably go
| in with a super sharp knife, cut it again very carefully, and
| make it so the ham and cheese don't actually ever touch.
|
| Apple (or whoever) purposely put a cut in the sandwich knowing
| the ham and cheese would touch, making the sandwich inedible.
| This YouTuber comes along with a really sharp knife and cuts
| the sandwich again making sure the ham and cheese never touch.
|
| It just so happens when Apple made the cut, they didn't cut
| anything important, other than "smooshing" the layers (ie: they
| didn't cut through any traces or capacitors, etc).
| Lammy wrote:
| A lot of people got XBOX360 XDK boards in this same way back in
| The Day.
___________________________________________________________________
(page generated 2024-05-21 12:02 UTC)