Macintosh SE/30 logicboard recreation (thread revival)

[trag]

in terms of replaceing the se/30's 8 ram slot with 72pin simms, its certainly doable, however while I haven't looked at the differences its likely some addition logic (like a PAL/GAL etc) would be needed to deal with stuff like RAS/CAS etc..

No additional logic needed if one sticks to certain 72 pin SIMMs. Specifically, one could only use Single-Banked 72-pin SIMMs. Almost all 32 MB 72 pin SIMMs (which is a fairly common size) are 2-bank. They're actually built out of two banks of 16 MB.

If you rewired the SE/30 to use 72 pin SIMMs, and then installed a normal pair of 32MB SIMMs, the SE/30 would only see 16MB on each SIMM.

For 128 MB, a pair of 64 MB SIMMs would work fine. Those are almost always single banked, although there have been a few two-bank ones.

 

[trag]

if you hunt around you can still find 30pim sockets from time to time, but finding 64pin rom sockets its probably not even possible these days, last time i got some of those was around 4 years ago and while i bought about 5 of them, i kind of regret i didnt but like 10-20 of them.

More than 9000 in stock. Less than $2 each.

64 pin SIMM Socekt

Click on teh "with center post" link to avoid the QTY 6 minimum.

 

[Lyokian159]

Nice to see another fully socketed board!
Did you already check the clock and reset signals?

Well, I got a scope and checked the signals. Clock was good, but lo and behold, the RESET line has a 3MHz oscillation on it! I went one chip at a time until I figured out is was the 68030 itself that was putting this on the line! Once the CPU came out the RESET line was a stable ~5v.

Has anyone seen anything like this?
I am assuming the '030 is toast unless there is some micro bridge I am not seeing under magnification.

 

[Jockelill]

Well, I got a scope and checked the signals. Clock was good, but lo and behold, the RESET line has a 3MHz oscillation on it! I went one chip at a time until I figured out is was the 68030 itself that was putting this on the line! Once the CPU came out the RESET line was a stable ~5v.

Has anyone seen anything like this?
I am assuming the '030 is toast unless there is some micro bridge I am not seeing under magnification.

Does the signal change of you push the reset button (or bridge it)? Maybe there is a bridge somewhere in that circuit

 

[robin-fo]

Check all your socket pins and make sure there is no bridge (e.g. some stuck solder wick)

 

[Lyokian159]

Check all your socket pins and make sure there is no bridge (e.g. some stuck solder wick)

Yeah, I used a new socket so no wick was involved in extracting the old socket, I am going to re-flow all the pins to the new socket to rule out any bridge from my soldering the unit in.

I will also tone out pin F12, reset, to everything around it to see if I can hunt it down.

 

[Lyokian159]

Yeah, I used a new socket so no wick was involved in extracting the old socket, I am going to re-flow all the pins to the new socket to rule out any bridge from my soldering the unit in.

I will also tone out pin F12, reset, to everything around it to see if I can hunt it down.

So I have tried a known good 68030. Same behavior. (And this 68030 works too so that's good.)

I have reflowed the CPU socket, verifying there are no bridges between F12 and any surrounding pins.

I swapped the SONY ICs around. No difference.

I am seeing the same behavior. 3MHz oscillation when the CPU is in. Stable 5v when it is removed.

The reset switch does indeed pull the line to ground and black out the display.

Any thoughts would be greatly appreciated!

 

[Lyokian159]

So I have tried a known good 68030. Same behavior. (And this 68030 works too so that's good.)

I have reflowed the CPU socket, verifying there are no bridges between F12 and any surrounding pins.

I swapped the SONY ICs around. No difference.

I am seeing the same behavior. 3MHz oscillation when the CPU is in. Stable 5v when it is removed.

The reset switch does indeed pull the line to ground and black out the display.

Any thoughts would be greatly appreciated!

Well, I finally fixer her! The reset line was shorted to the address bus on A0. Micro brige under the socket for UE10, pin 40 to 41!

Thanks to everyone who lent assistance!

 

[avadondragon]

Did anyone notice PBCWay has the Mac SE/30 motherboard listed as their PCB Project of the week?

 

[Daniël]

Did anyone notice PBCWay has the Mac SE/30 motherboard listed as their PCB Project of the week?

Who the hell is "Nucci from the United States of America", and why did they feel the need to steal the credits from Bolle?

 

[avadondragon]

Good question. It was posted just a couple of days ago. I figured I'd draw attention to it.
Here's a direct link to the project:
https://www.pcbway.com/project/shareproject/Apple_Macintosh_SE_30_Motherboard_78143494.html

 

[Daniël]

Good question. It was posted just a couple of days ago. I figured I'd draw attention to it.
Here's a direct link to the project:
https://www.pcbway.com/project/shareproject/Apple_Macintosh_SE_30_Motherboard_78143494.html

Oof, apparently PCBWay hands over 10% of the profits to the supposed author if people purchase through that project page.

 

[Bolle]

Oof, apparently PCBWay hands over 10% of the profits to the supposed author if people purchase through that project page.

Honestly, I wonder why it took someone so long until that happened. Have stuff available for free for everyone to use and things like that WILL happen.
I can live with it

It's fun how they just copied the text from github and it says JLCPCB all over it

 

[croissantking]

Well, I finally fixer her! The reset line was shorted to the address bus on A0. Micro brige under the socket for UE10, pin 40 to 41!

Thanks to everyone who lent assistance!

I am really interested to know how you troubleshooted this successfully.

When you saw the 3MHz oscillation on the reset line, how did you know this meant something was wrong?

How did you locate the micro bridge, did you have to desolder the whole socket?

 

[Lyokian159]

I am really interested to know how you troubleshooted this successfully.

When you saw the 3MHz oscillation on the reset line, how did you know this meant something was wrong?

How did you locate the micro bridge, did you have to desolder the whole socket?

Well, So we know 2 things for sure. 1: The reset line is pulled high naturally through RP7. 2: The SONY chip, and the reset switch should be the only items able to pull the line to ground. So if I am seeing anything but a stable voltage, something is dragging the line down.
The fact that it was a steady frequency eluded me until later.

The line was only getting pulled low when the CPU was in the socket. This meant one of two things, either the CPU had a fatal flaw, or something was bridged to a line that was only actioned by the CPU.

I replaced the CPU with a known good one, so the CPU was working. I then decided to brush tone out every CPU pin to F12 which was the reset line. As i dragged up column 2 I heard a beep. Narrowed down to pin A2 which is Address 0.

Now these two pins are way to far away from each other to be bridged at the CPU socket. So we look at the schematics, and look for a chip where A0 and RESET are close to each other. UE10, Pins 40 and 41 are A0 and RESET.

I wasn't able to visually see a bridge due to the plastic of the socket. So I fluxed and reflowed the two pins, and boom, no more short.

The fact that it was oscillating was because, and i could be wrong, the system was in a feedback loop. RESET wants to be high, and is active Low. So the CPU tries to reset its state, at boot when the sound chip pulls the reset line low, which means that A0 should be low, which causes the RESET line to ground, which causes the CPU to reset, and so on and so on.

I really enjoyed this build and a Huge thanks to Bolle for your valuable work!

 

[croissantking]

Well, So we know 2 things for sure. 1: The reset line is pulled high naturally through RP7. 2: The SONY chip, and the reset switch should be the only items able to pull the line to ground. So if I am seeing anything but a stable voltage, something is dragging the line down.
The fact that it was a steady frequency eluded me until later.

The line was only getting pulled low when the CPU was in the socket. This meant one of two things, either the CPU had a fatal flaw, or something was bridged to a line that was only actioned by the CPU.

I replaced the CPU with a known good one, so the CPU was working. I then decided to brush tone out every CPU pin to F12 which was the reset line. As i dragged up column 2 I heard a beep. Narrowed down to pin A2 which is Address 0.

Now these two pins are way to far away from each other to be bridged at the CPU socket. So we look at the schematics, and look for a chip where A0 and RESET are close to each other. UE10, Pins 40 and 41 are A0 and RESET.

I wasn't able to visually see a bridge due to the plastic of the socket. So I fluxed and reflowed the two pins, and boom, no more short.

The fact that it was oscillating was because, and i could be wrong, the system was in a feedback loop. RESET wants to be high, and is active Low. So the CPU tries to reset its state, at boot when the sound chip pulls the reset line low, which means that A0 should be low, which causes the RESET line to ground, which causes the CPU to reset, and so on and so on.

I really enjoyed this build and a Huge thanks to Bolle for your valuable work!

Awesome detective skills there, brilliant.

 

[cheesestraws]

Oh well done. That's good troubleshooting, that is.

 

[Jockelill]

Has anyone been able to use the PLCC for the rom instead of the SIMM? No matter what we do, we can’t get it to boot. With a simm in place it boots right away.

If anyone has had success with these, I’d be very grateful for some rom files so I can compare. Byte swap has been done as per instructions.

 

[Cam]

So I have a potentially daft question, but I'm going to ask it anyway.

I've been slowly transfering parts from a dead SE/30 to a Reloaded board and managed to score a second battery bombed SE/30 board with the thought that it would be good in case I baff'd any parts in the transfer. The new dead board wasn't in horrible shape. Yes the battery leaked, and yes there is damaged to the board, but not as much as the first donor board.

So my question? How many broken traces are worth attempting to repairing before you consider going to a full Reloaded rebuild?

 

[WillJac]

So I have a potentially daft question, but I'm going to ask it anyway.

I've been slowly transfering parts from a dead SE/30 to a Reloaded board and managed to score a second battery bombed SE/30 board with the thought that it would be good in case I baff'd any parts in the transfer. The new dead board wasn't in horrible shape. Yes the battery leaked, and yes there is damaged to the board, but not as much as the first donor board.

So my question? How many broken traces are worth attempting to repairing before you consider going to a full Reloaded rebuild?

For me, traces are not so bad but as soon as vias are gone both ends missing and no copper inside, it’s a done deal for a transfer. Simple trace repairs I would say 15 or less would try fix the board 16 or more do a transfer. Be sure you move the items needed from the back also.