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

[JustG]

I am so glad I found this forum.  All of these projects amaze me and I love seeing an idea come to reality.

 

[LaPorta]

Perhaps we can find a US-based supplier to crank some out here as well. An incredible job!

 

[Kai Robinson]

After a full inspection of the boards, i can say the following: JLCPCB have done a fantastic job, for the money. The silkscreen on both layers is sharp and better quality than PCBWay for sure. Everything fits - although the holes in the prototype may be a *little* tight for reusing some parts - thankfully most of the parts that need to be reused are to be socketed anyway, so not really too much of an issue - although the SIMM socket pads were 1.4mm x 0.72mm, so i've made them and all the other DIP socketed parts a touch larger, at 1.6 x 0.8mm to make for a better fit, especially if reusing parts. The pads for the rear I/O, such as the ADB/Serial and the metal support bracket are a tight fit and i missed out excluding the pads on the underside from the solder mask, so i've already re-done the sockets from scratch to give a 0.25mm larger surface area (track width from 0.25mm to 0.50mm as they're 'drawn' and then infilled with an exclusion zone), as modern PCB's are very, very accurate and incredibly fine, compared to ones from the 1980's.

Now the biggest fear i had was hole alignment. Fear not - for all the holes line up bang on! Literally, 1:1 is what i was aiming for, and i assumed i may be out by about 2-3mm in some places, but from the measurements, 1mm difference at best - In addition, all the rear IO ports are PERFECTLY ALIGNED.

I'm going to start building these up on Monday/Tuesday next week as i have family commitments and a well earned barbecue to host. I need to get some parts from a US supplier of SIMM & PGA sockets - would someone mind being a middleman for that?

 

[pax]

Beautiful!

 

[Kai Robinson]

It's been a while since i last posted - just having time to myself has been nice. So - i'm missing the capacitors, a 53c80 SCSI chip and the SIMM sockets (my old ones may not fit in the holes) but here's how it looks built up so far...

Still need a middleman for the parts from phoenix electronics, if someone wants to volunteer?

 

[blusnowkitty]

This'll be a fun project and I might have to get in on a board or two. Both my SEs sit in the corner gathering dust because they have bum SIMM sockets. Stupid brittle plastic clips.

 

[Kai Robinson]

@blusnowkitty Well - you can get some new sockets from Phoenix Electronics - i linked them earlier in the thread - but you're welcome to sacrifice an SE to the cause

 

[techknight]

I love the fact that you put everything in sockets

 

[Kai Robinson]

I love the fact that you put everything in sockets

Well, makes it easier to swap out stuff - especially during the prototype stage - i even socketed the oscillator

 

[techknight]

True, plus it makes it easier to test chips that may be dead or not from other boards. 

 

[Torbar]

If you're still looking for a US based middleman for the parts from phoenix, shoot me a DM,  I can help out

 

[jimjimx]

Are these FS yet?  What’s the cost?

 

[Kai Robinson]

@Torbar - yes please, theres about $50 of parts i'd like, obviously willing to pay extra for your time and shipping, obviously. Need some PGA68 sockets too as well as a glut of the 30-pin SIMM sockets. 

@jimjimx - Not yet, this is the prototype stage, the prototype boards i just had made are in need of bodge wires for VCC and GND for the I/O ports, but that's been fixed, along with the silkscreen, in the latest board revision. If the prototypes are a success, i'll produce, and sell for about £30 (GBP) per board, to recoup development costs and board production costs. I'll do a limited run of boards at OSHPark for those who want the glass black finish and clear soldermask (so you can see bare copper on black), but otherwise will be a standard run of boards in either green, purple, blue or matte black....
 

 

[Kai Robinson]

OK - can someone else with an SE logic board take a photo of the text on the side of C19 - need to know if it's a 10nf Cap, or something larger - Also, is C67 (above the BBU, and to the right of the uppermost 74F257) populated at all or just blank pins?

 

[cheesestraws]

I can't get good enough light right now for a photo, if nobody has provided one by tomorrow I'll have another go with more light.

 

[Kai Robinson]

Bar the 50pin IDC, PRAM battery, the electrolytic capacitors (finding axial ones is a pain), one transistor, one resistor and the ferrite bead....we are t-minus 4 days to launch!

 

Getting the SIMM sockets back in was a pain due to the slightly bent pins, but after the requisite blood sacrifice was made both went in without much trouble. Blood for the blood god, I guess....?

 

[Kai Robinson]

Revised BOM here:

* Partlist from file: MacintoshSE.lay6" / 26/07/2020 00:32:29
*
* Name	Value
*

C1	100nF
C2	100nF
C3	100nF
C4	33uF 16v
C5	33 pF
C6	100nF
C7	100nF
C8	100nF
C9	100nF
C10	100nF
C11	33uF 16v
C12	100nF
C13	33uF 16v
C15	100nF
C16	33uF 16v
C17	100nF
C18	33pF
C19	10pF
C20	100nF
C21	100nF
C22	100nF
C23	100nF
C24	100nF
C25	100nF
C26	33uF 16v
C27	100nF
C28	100nF
C29	100nF
C30	100nF
C31	100nF
C32	33uF 16v
C33	100nF
C34	33uF 16v
C35	100nF
C36	1uF 16v
C37	33uF 16v
C38	33uF 16v
C39	100nF
C40	68pF
C41	100nF
C42	100nF
C43	100nF
C44	100nF
C45	100nF
C46	100nF
C47	100nF
C48	100nF
C49	100nF
C50	100nF
C51	100nF
C52	100nF
C53	100nF
C54	100nF
C55	100nF
C56	100nF
C57	100nF
C58	100nF
C59	100nF
C60	100nF
C61	100nF
C62	33uF 16v
C63	100nF
C64	100nF
C65	100nF
C66	100nF
C67	100nF
CR1	1N4150
CR2	1N4150
CR3	1N4150
J1	EAR				Kobiconn 161-3520-EX
J2	SERIAL			TE 5749268-1
J3	SERIAL			TE 5749268-1
J4	EXT SCSI		Amphenol L77SDB25S1ACH4R
J5	INT SCSI		Molex 87831-5019
J6	EXT FLOPPY
J7	UPPER DRIVE		Molex 87831-5920
J8	LOWER DRIVE		Molex 87831-5920
J9	ADB				TE 5749181-1
J10	ADB				TE 5749181-1
J11	Speaker
J12	MAIN			Molex 39-28-1143
J13	PDS				TE 5535032-4
J14	SIMM			PE Connectors HWS9229
J15	SIMM			PE Connectors HWS9229
J16	1x3 HEADER
L1	33uH
L2	33uH
L3	33uH
L4	Ferrite Bead
Q1	2N3904
R1	10k
R2	10
R3	47
R4	47
R5	560k
R6	10
R7	10
R8	1.2k
R9	47
R10	47
R11	1k
R12	150
R13	1k
R14	470
R15	1k
R16	3.3k			
R17	10k
R18	1k
R19	10
R20	75
R21	3.3k
R22	47
R23	47
R24	47
R25	47
R26	3.3k
R27	3.3k
R28	47
R29	47
R30	47
R31	10k
R32A	47
R33A	47
R34	10k
R35	150
R36	150
R37	39
R38	39
R39A	47
R40A	47
R41	3.3k
R42	39
R43	3.3k
R44	3.3k
R45	3.3k
R46	3.3k
R47	3.3k
R48	3.3k
RP1	39				4610X-102-390LF
RP2	39				4610X-102-390LF
U1C		SND
U2C		26LS30
U2E		74ACT257
U2F		74ACT257
U3C		26LS32
U4B		FILTER
U4C		26LS32
U4D		MC3488A
U4E		BBU
U5D		85C30 SCC
U6B		58C30 SCSI
U6D		ROM HIGH
U7D		ROM LOW
U7E		MC68000
U8B		FILTER
U8D		SWIM
U9D		GLU
U10C	ADB
U10D	16 Mhz Clock
U11A	FILTER
U11B	FILTER
U11B	FILTER
U11C	RTC
U11D	Rockwell 338-6523 TPI
U11F	74FCT245
U12F	74FCT245
Y1		32.768 KHz

 

[tt]

@Kai Robinson Wow this is such an interesting take on PCB repair. I have this strange urge of wanting to transplant components off an old working board to a new one, which I know makes no sense. If you need help with buying parts in the USA, let me know. I am interested in buying some simm sockets to repair some on a board but I'm not sure if they are 1-up or multi-up ATM.

 

[cheesestraws]

@Kai Robinson R25 on your BOM above you have marked as a 47Ω, but on the three boards I have here out of machines, it's 560Ω (see photo attached).  Mind sanity checking me to make sure I'm not going mad?

 

[Kai Robinson]

You are indeed correct, it is 560Ohm, and that was on the list i have noted down - I could have sworn i amended the BOM correctly above...