Reverse Engineering the Macintosh SE PCB & Custom Chips for 1:1 reproduction

[Kai Robinson]

OK - back to what's going on - I'm reading the guide to macintosh hardware and i've come across this:

 

When power is first applied to the Macintosh SE computer, the following sequence of
events takes place:

1. The Sony sound IC monitors the voltage levels on the board and asserts the /RESET
signal until 0.25 second after the voltage has stabilized.

2. The /RESET signal causes the CPU, the BBU, and all of the internal devices to come to a
known initial state.

3. The Sony sound IC deasserts the /RESET signal and the CPU looks at the first four
words in memory (starting at location $00 0000) to get the Reset vector. When the
BBU is reset, it starts up with the ROM overlay address map, which puts ROM at
location $00 0000. In the ROM overlay address map, an address to a location in the
range $40 0000 through $43 FFFF is also decoded by the BBU as an address to ROM.

4. The CPU goes to the memory address pointed to by the Reset vector and begins to
execute the code it finds there (the Reset handler).

5. One of the first instructions in the Reset handler is a jump to a location in the range
normally assigned to ROM ($40 0000 through $43 E800). The first time the BBU
receives an address in this range, it switches to the normal address map. In this address
map, RAM is located at $00 0000 through $3F FFFF and ROM is located at $40 0000
through $43 FFFF.

6. The Reset handler carries out the startup procedure described in Inside Macintosh




I wonder if there's a voltage issue somewhere that means the SND isn't releasing the /RESET pin...

 

[Kai Robinson]

More musings from cycling through that massive tome....for the floppy control logic in the GLU - it can be replicated with a single 74LS125 and a 74LS04...that'll sort the inverted WR signal and the ENBL1/ENBL2 signals. The spare inverter and buffer can be connected to make the SCSI IRQ's pass to the BBU...then it's 

Now just need to sort out the clock buffering - presumably a non-inverting buffer gate would suffice...? The Mac Plus feeds the oscillator signal into a single 74F04 (U4D) - question is, why inverted, rather than non-inverted...?

 

[Kai Robinson]

OK. So....some KiCad and i've done a thing. Made a GLU replacement with discrete logic using tiny VQFN parts. Will get JLCPCB to do a batch of 5, with assembly. Should be fairly cheap - the VQFN's are about 50p each.

 

[Phipli]

Good work Kai

 

[maceffects]

@Kai Robinson great work on the GLU!

 

[Kai Robinson]

I know we had reverse engineered the PAL, I just fancied a challenge to see if I could do it in discrete logic... 

 

[maceffects]

For those not on this Facebook group. Here is the guy who designed the Mac SE logic board. 

 

[quorten]

@Kai Robinson I'm pretty sure the only reason why the Mac Plus inverts the clock signal is because that must have appeared to be the most convenient chip-saving layout.

Good to know it is the Sony SND chips that's responsible for generating the power-on-RESET signal.  I believe the interesting thing about the RESET signal to the BBU is that it is not actually a chip-level RESET but rather primarily used to reset the overlay bit and a small amount of select other information.  The point here is that when you press and hold down the programmer's RESET button, the screen contents do turn to garbage, so the BBU itself must have its own internal power-on-RESET for the DRAM control and screen scanning state.

 

[Kai Robinson]

For those not on this Facebook group. Here is the guy who designed the Mac SE logic board. 

View attachment 40676

Can you direct him to this thread so we can pick information from his brain?

 

[Phipli]

Can you direct him to this thread so we can pick information from his brain?

Several of us did

I think he might be worried he's forgotten everything Hopefully he joins in anyway because any insights will likely be extremely helpful.

Or contacts!

 

[Kai Robinson]

I'd enquire myself, but i eschewed facebook completely (deleted in 2018). 

 

[CircuitBored]

So uhh... This happened:

https://hackaday.com/2021/02/08/recreating-the-mac-se-logic-board/

I'm a little confused and a little irritated about it. The article has basic errors and I wasn't even consulted about it before it was posted. 

Bit rude, imo. 

Did anyone here submit it to them? 

This is astonishing - the nerve of some people! It seems that a user named toru173 sent them the link but the 68kmla account by that name was only made six days ago... Hackaday is a fairly washed-up outfit, I don't think they even have a real human vetting the articles they are sent. 

 

[Kai Robinson]

This is astonishing - the nerve of some people! It seems that a user named toru173 sent them the link but the 68kmla account by that name was only made six days ago... Hackaday is a fairly washed-up outfit, I don't think they even have a real human vetting the articles they are sent. 

toru173 reached out to me - please don't blame them for this - it's easy to get swept up in excitement/possibilities of the project. 

 

[CircuitBored]

toru173 reached out to me - please don't blame them for this - it's easy to get swept up in excitement/possibilities of the project. 

Oh no, I didn't mean for any animosity. It sure would be sweet if this article ended up getting some more wind in the sails of this project. 

 

[Newtboy]

Oh no, I didn't mean for any animosity. It sure would be sweet if this article ended up getting some more wind in the sails of this project. 

Got me dusting off the Classic!

 

[Kai Robinson]

With so many months worth of content just gone - i'm not going to waste my time trying to fill in the blanks.

Archive.org has a backup of this thread upto May, i suggest anyone seeing this and wanting to find out what happened next, can view it there, or see the abridged version at the Phoenix Engineering Wiki: https://wiki.phoenix-engineering.org/wiki/Macintosh_SE_Reloaded

Anyone that ordered a board from me and is still waiting, or has issues, please send me an email at kai(underscore)robinson[at]hotmail|dot|com - i won't be relying on forum messages again.

 

[Kai Robinson]

@Cory5412 can you please lock this thread.