Jump to content

Reverse Engineering the Macintosh Classic PCB for 1:1 reproduction


Recommended Posts

8 hours ago, anthon said:

I understood the purpose of the resin, but not being a mechanical engineer person didn't understand how one would accomplish it. Thanks @Trash80toHP_Mini for explaining one possible way to do it. Sounds like it might be more suited to a CNC mill (for the leveling part) than a manual mill like I had in my head.

 

1 oz copper is 0.035 mm thick... that's some mighty precise leveling to pull off.

 

Mom's dad was a master machinist at GE during WWII and well into my youth. I'll bet he could have pulled this off using the manual tools in his basement machine shops. He made and fixed the machines used to fabricate the machines that produced the factory floor machines used roll out mass quantities of war machines.

 

.035mm = 1.38 mil, how much of that thick copper cladding did you want shaved off? [:)]

 

 

Edited by Trash80toHP_Mini
Link to post
Share on other sites
  • 3 weeks later...
  • 1 month later...

Oh wow. I just pulled my Classic from storage (no battery, I'm not stup...erm well, I shoulda maybe checked. The caps leaked and ate some traces. I might be a customer if I can't fix the damage.

 

Great work.

Link to post
Share on other sites
  • 68kMLA Supporter
19 hours ago, yuhong said:

I am surprised that people are focusing on the SE and Classic, when the Plus should be much easier.

 

There is barely any need for fresh Plus, 512K(e) or 128K boards. Those had the battery in a separate compartment attached to the analog board, thus battery corrosion practically never reaches the logic board. That, coupled with the axial through hole capacitors that are not known for leaking, means most of those boards are just fine, all these decades later. Unlike the SE, SE/30 and Classics, which have the battery on the board, which is prone to completely exploding all over the surrounding area with corrosive fluid.

Link to post
Share on other sites
  • 68kMLA Supporter

I suspect it's just more a situation of what machines Kai had. :) And that he had a FDHD/SuperDrive, not an original SE. Those Vartas on the original SE just don't explode. (At least not yet.)

Link to post
Share on other sites
On 4/5/2021 at 6:13 AM, Daniël Oosterhuis said:

 

There is barely any need for fresh Plus, 512K(e) or 128K boards. Those had the battery in a separate compartment attached to the analog board, thus battery corrosion practically never reaches the logic board. That, coupled with the axial through hole capacitors that are not known for leaking, means most of those boards are just fine, all these decades later. Unlike the SE, SE/30 and Classics, which have the battery on the board, which is prone to completely exploding all over the surrounding area with corrosive fluid.

On the other hand, the only custom chip on the Mac Plus board is the IWM.

Link to post
Share on other sites
  • 68kMLA Supporter
On 4/6/2021 at 5:22 AM, yuhong said:

On the other hand, the only custom chip on the Mac Plus board is the IWM.

 

But that's partly it, isn't it.  Part of the fun of it is the challenge.

Link to post
Share on other sites
  • 68kMLA Supporter

Well, to clarify - yes, the SE was a specific target, I didn't have one, saw the battery issues and realised that the SE is the ideal target machine: -

 

Original 68k, but without the issues of a lot of PAL's on it, still has SIMM support, has 1.4MB floppy support and DUAL ones at that, ADB support means that keyboards/mice are easier to source, has a PDS slot, and is upgradeable, and it's the last pure through hole board iirc. Everything the Plus has but...betterer.

Link to post
Share on other sites
  • 68kMLA Supporter
On 4/7/2021 at 2:35 AM, Kai Robinson said:

Original 68k, but without the issues of a lot of PAL's on it, still has SIMM support, has 1.4MB floppy support and DUAL ones at that, ADB support means that keyboards/mice are easier to source, has a PDS slot, and is upgradeable, and it's the last pure through hole board iirc. Everything the Plus has but...betterer.

 

Don't forget, it has internal SCSI as well. The Plus can be modded as such, but the SE is definitely easier to internally mount a SCSI2SD/MacSCSI/BlueSCSI/etc.

Link to post
Share on other sites
  • 3 weeks later...
  • 68kMLA Supporter

Okay - time to ramp up this project now that the SE has tapered off a bit!

 

I'd like anyone with a wrecked board to contribute, if they're up for it - what i mean is - if the board is truly bollocksed, and can't be patched - SAND IT! I'd love to know the inner two layers - i'm fairly certain that again, it's a 4-layer board. Bruce from Branchus Creations has taken a photo of a cross section and it looks very similar to the SE in that the middle layers are VCC and GND, but i'd love to confirm this. However, i actually don't have a sander and my hand dexterity has taken a hit recently, so hand sanding it for me is out of the question. 

 

Any volunteers? 

Link to post
Share on other sites
  • 68kMLA Supporter
22 hours ago, Kai Robinson said:

Any volunteers? 

I have some very dead classic boards that I could try sanding and scanning. How do you maintain an even sanding layer on the board? I'd imagine the layers are really thin, so I'd need to make sure that only one layer is removed at a time somehow.

Link to post
Share on other sites
  • 68kMLA Supporter

The disassembly of one of my dead Classic Logic board is underway! I have already removed all of the SMD components from the top of the PCB. I am now working on the thru-hole connectors. As I don't own a desoldering gun, I decided to use a heat gun to melt all of the pins at once from the back of the PCB while tugging on the connector with pliers. --This technique works, but it makes lots of nasty fumes (from melting plastic), and goes a little slower than I would like. In any event, I am fairly close to being able to scan and sand down this logic board for reverse engineering!

 

Here are pictures of the board before I desoldered everything. As you can tell, this poor Logic Board was beyond saving (just look at all of the destroyed VIAs in close proximity to the CPU!). About a year ago, I tinkered with it as a practice board for a while, but alas, this one is truly dead.

PXL_20210410_160442157.jpg

PXL_20210410_160451774.jpg

Link to post
Share on other sites
  • 68kMLA Supporter

Woohoo! The more the merrier! I've spent the last 8 hours inserting the split ground plane into the Sprint Layout, and toning out most of the major pins of stuff - pretty damn sure that everything in the middle layers is just ground and power, i can't see anything else wierd, except MAYBE /DACK on the BBU not being connecte to anything despite a trace and via being connected to....something. 

 

Here's where we stand:

 

 

macclassic-top-3.JPG

macclassic-bottom-3.JPG

Link to post
Share on other sites
  • 68kMLA Supporter
Posted (edited)

I've finished removing all of the components from my PCB, so I'm ready to start sanding down to the internal layers! I found an old electric sander to use, so hopefully it will make things less difficult. I'll have to do this outside with a mask/goggles on, as I bet there's going to be an absolute ton of fiberglass dust everywhere. I will keep everyone updated...

 

 

 

 

MacClassicLBScan1Compressed.thumb.jpg.16f30f830be4b9b828bb20bbbb69922b.jpgMacClassicLBScan2Compressed.thumb.jpg.c3f8a15574781d71ecb7b68047447946.jpg

Edited by davidg5678
Link to post
Share on other sites
  • 68kMLA Supporter

The area around C26 on the rear looks like it's taken a fair amount of heat! I'd start in that area on the back, just to make things easier! Also - remember that there is a split power plane across the rear I/O. 

Link to post
Share on other sites
On 5/3/2021 at 7:30 AM, Kai Robinson said:

I can't see anything else wierd, except MAYBE /DACK on the BBU not being connecte to anything despite a trace and via being connected to....something. 

 

You mean this pad/track/via combo (UI4p56 BBU)?

image.png.6084a27d33c84e6f79762cf333c7b86d.png

 

According to the schematics (and my paint bucket agrees) the track that goes up from the BBU pad finds it's way to UE8p11 (SCSI/DACK) which looks correct.

 

image.png.b42e35d1a7cf8193e7f980d5fafe3ac9.png

 

The via however appears to go nowhere. Maybe it's a relic of a revision when the track started going that way, but was later moved to exit the other end of the pad.

 

image.png.e38201d4cc264f075ee2352432442279.png

Link to post
Share on other sites
  • 68kMLA Supporter

Sanding progress is underway! Last week, I started sanding through the PCB. I found this information useful. I am using 60 grit and 220 grit sandpaper, as well as an electric sander. I started off wet-sanding by hand (in my very dirty utility sink), but progress was impossibly slow, so I switched to the electric sander after about an hour of minimal results.

PXL_20210503_190249729.thumb.jpg.dedd4b1a1dd40e132b84d7749652db29.jpg

 

After removing the silkscreen and solder mask, this is what the PCB looked like:

layer1sandClassicLB.thumb.jpeg.7f5378c9fd2b9bd544737fa3365f3ef1.jpeglayer1sandClassicLB2.thumb.jpeg.5b73d7119fc4f1e2cc8a2262f0a4e69e.jpeg

You can really see where I accidentally de-laminated parts of the PCB here!

 

On 5/2/2021 at 3:58 PM, Kai Robinson said:

The area around C26 on the rear looks like it's taken a fair amount of heat! I'd start in that area on the back, just to make things easier! Also - remember that there is a split power plane across the rear I/O. 

I accidentally burnt this section (and a few others) with hot air while I was de-soldering some components. Luckily, as I am sanding through the PCB anyway, I should be able to go right past the de-laminated layers.

 

It was at this time I switched to using an electric sander, which greatly accelerated progress, but also had the unfortunate side effect of making everything less even. I was able to remove the entire bottom layer of the PCB, and I went through all of the solder mask, traces, and silkscreen. I then hit a layer of fiberglass, followed by another layer of copper, followed by yet another layer of fiberglass. It was at this point, I temporarily stopped my sanding, as I really think I need to figure out a way to make it more even. I hope the images below should give a good idea of where the first ground plane goes, and it could be probably be matched up to the existing PCB model using the VIAs as an alignment guide. I have the PCB held down with very strong double-sided tape for now, so I can scan it, but I'd need to first break the tape. I'm happy to do this once I figure out how to sand evenly...

 

I believe there should still be a way to salvage this project so that we can gain some meaningful information from it, but I'd hate to sand too much away before we can learn where the copper planes are actually located. If anyone has ideas about how to sand more evenly, I'd love to hear them! Are some electric sanders more accurate than others? Would I be best off evening things out by hand? Please let me know what you think.

 

PXL_20210504_215514080.thumb.jpg.f6e4b848fb9bb2144e7e2e1cb0f9f3e6.jpgPXL_20210504_214636556.thumb.jpg.fe0da6207aa96d90ccab66a1b7f61e81.jpg

 

 

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...