• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

Reverse Engineering the Macintosh LC III Logicboard

mmu_man

Well-known member
LOL! Possibly they later on patched the layout to bypass them but manufacturing still had the pick&place configured to put them. Next time I open mine I should check if they're here.
What part of the circuit are they on?
 

max1zzz

Well-known member
they are connected to pin 24 of the ext SCSI connector and pin 4 of the serial ports to ground, after this I found C40 (again by the serial ports) and C70 (Above the PDS connector) are the same. These look like they are linking the audio / IO ground planes to the normal ground plane, I guess apple orignally did this with caps then changed the design at a later date
 

360alaska

Well-known member
If you read the thread: This is actually common when dealing with EMC. The zero ohm resistor has some parasitic inductance, so the capacitor acts like a better path to ground at high frequencies. Many designs will have options for a few different ground coupling layouts and then test to work out what they need to achieve compliance.
 

cheesestraws

Well-known member
I did read the thread! What I don't understand is why coupling high frequencies effectively between the grounds is a good thing in this specific case. I'm being stupid rather than illiterate :). It's not meant as a passive-aggressive swipe, you may well be right, but I'd really like someone to explain to me why, because grounding is black magic and I don't understand it.
 

mmu_man

Well-known member
Are you just doing the PCB directly or the schematics also?
Could be interesting to be able to patch it. At least to add a switch/jumper for overclocking :)
 

max1zzz

Well-known member
I'm just doing the PCB, BOMARC schematics are already available for this model

I have been considering drawing up schematics for the LC though from my board layout, as no schematics are available for that one
 

mmu_man

Well-known member
Well, schematics available as ugly bitmap scans and schematics in a usable format is different but yeah…
 

max1zzz

Well-known member
The BOMARC schematics are modern scans and are 100% readable, I see no reason to re-engineer the wheel here.

The apple SE, SE/30 and PLUS schematics floating around the place are a different matter though.
 

mmu_man

Well-known member
Yeah but modifying the PCB without the schematics would be harder without at least a netlist for checks.
 

Trash80toHP_Mini

NIGHT STALKER
Well, schematics available as ugly bitmap scans and schematics in a usable format is different but yeah…
max is doing three boards at once, that's above and beyond in my book. Check out the Bomarc schematics as he suggested, translating those into schematic capture format for PCB design apps would be a great contribution to the overall effort.
 

mmu_man

Well-known member
max is doing three boards at once, that's above and beyond in my book. Check out the Bomarc schematics as he suggested, translating those into schematic capture format for PCB design apps would be a great contribution to the overall effort.
Well I would maybe if I didn't have so many projects already. And then I'd do it in KiCAD anyway so it probably wouldn't match with the PCB which would be a shame.
 

mogs

Active member
What I don't understand is why coupling high frequencies effectively between the grounds is a good thing in this specific case.

The designers get to choose exactly how these two grounds are bonded, maintaining control over the flow of return currents at all frequencies. Otherwise you're leaving it to chance how high frequency return currents make their way back to the power supply.

I've actually experienced this problem on a board I worked on, the designer used two ground planes bonded with inductors (they said to prevent high frequency coupling). This resulted in significant and measurable noise on the second plane. I ended up just bridging them together with solder and the noise went away. A capacitor might have achieved a similar result. At the end of the day I don't think the designer was right to do that without providing an alternate path for the high frequency return current.

It's also pretty common to place capacitors to bridge slots and notches in ground planes, to prevent high frequency return currents flowing around the slot (which has a tendency to emit RF).
 

cheesestraws

Well-known member
I've actually experienced this problem on a board I worked on, the designer used two ground planes bonded with inductors (they said to prevent high frequency coupling). This resulted in significant and measurable noise on the second plane. I ended up just bridging them together with solder and the noise went away

Thanks, this explanation was really helpful. One of these days I might understand this stuff...
 

max1zzz

Well-known member
It's getting there slowly, I have almost finished the copper areas on the top and bottom after that I only need to do inner planes and board outline.

Should be ready to send to production in a week or two :)
 

CC_333

Well-known member
Ah, OK!

You haven't been saying much about it lately. I figured you were still working on it quietly, but I was curious about your progress.

If you're up for it, I eventually wouldn't mind a clone of the backlit Mac Portable logic board made. I even have a broken donor for you to analyze, if so desired!

c
 

max1zzz

Well-known member
Indeed, I have spent the last week or two just drawing little coloured boxes on the board so not much to report!

Ohh that would be intresting! Do you know if the backlit portable's board is compatible with the non backlit's chassis? I only have a non-backlit portable so testing would have to be done with that
 
Top