Jump to content

Macintosh SE/30 Schematics (modernization effort)


Recommended Posts

8 hours ago, quorten said:

desertrout, those do look to be photos of the SE/30 logic board, the SE logic board is different and its layout is mentioned elsewhere in the discussion thread.  The SE logic board only has two internal layers for power and ground planes.

Thanks @quorten - sorry it was indeed a silly question, it became obvious once I looked at the photos. :-/

Link to post
Share on other sites
  • Replies 146
  • Created
  • Last Reply

Top Posters In This Topic

I'm new to meddling with the electronics side of vintage computers (did not touch a soldering iron until May).  I'm getting into the more component side of Macs now.

 

This project intrigues me.  A lot.  There's efforts for this on other vintage platforms, like Commodore 64, etc.  I don't exactly know how it works with regards to improvements, but would there be any way of improving upon the SE/30 design when remaking it?  As in, could a different type of RAM socket be integrated?  Support for a faster CPU (like, say, the 68030 40MHz from the IIfx)?

 

This is a great, promising project.  Maybe if it succeeds into the production of a board, we could have modern brand new Mac clones (still 68k though).  I would imagine it would be possible to build out a machine that used this board design and yet made the machine into a modern box, with color video output to a standard Apple/VGA monitor, etc.  Sort of like what they did with the Ultimate 64, which is a Commodore 64 reproduction, using FPGA, and modern HDMI video out, and USB connectors.  I would get behind a project that did something like that, financially.

Edited by MrFahrenheit
Fixed missing words
Link to post
Share on other sites
9 hours ago, MrFahrenheit said:

This project intrigues me.  A lot.  There's efforts for this on other vintage platforms, like Commodore 64, etc.  I don't exactly know how it works with regards to improvements, but would there be any way of improving upon the SE/30 design when remaking it?  As in, could a different type of RAM socket be integrated?  Support for a faster CPU (like, say, the 68030 40MHz from the IIfx)?

There has been some talk about it across the forums. Some relative easy ways to improve the board can be achieved. Theoretically, you could just drop a 40MHz 030 on a stock SE/30 board (if it is one of the socketed ones), but you'll find it'll still run at 16MHz. The SE/30's flaw is that it has a singular 32MHz clock crystal, that runs several components' clock signal. It gets divided on the board for the components using it, such as the FPU and CPU receiving a halved frequency, resulting in the 16MHz speed. Of course, if you were to replace it with, say, an 80MHz crystal, the CPU and FPU would now be running at the full 40MHz... but now your video circuitry runs out of whack, and you get an inoperable machine.

 

A simple solution is to do what some of the II-line machines did: Separate crystals for the components needing specific frequency. If you have the FPU and CPU on their own crystal, or heck, each on their own, you can theoretically install a 40MHz 68030 and 68882, and effortlessly run them at full speed with the rest of the system having their necessary clock speeds untampered with. The fastest models of 68030 were only in PGA form, which the SE/30 uses, so that can go up to the 50MHz mark. The 68882 on the SE/30 is in PLCC form, which gets you to the 40MHz mark. To get to 50MHz without overclocking on the FPU, you would need to modify the board to use an PGA FPU. However, the FPU doesn't need to run at the same speed of the CPU, it can run asynchronously at a higher or slower speed than the CPU.

 

The eight 30-pin SIMMs could theoretically be replaced with two 72-pin SIMMs, allowing for easier and cheaper upgrading to the maximum of 128MB. One 30-pin SIMM is 8 bits, so a bank of four is 32 bits, which is equal to one 72-pin SIMM.

 

I would personally also like to see the 68-pin ROM SIMM slot with socketed flash modules/chips. And you can go on, maybe cache could be added, maybe an onboard Ethernet solution, etc. And if the proprietary chips are ever reverse engineered, you can do quite a lot more with FPGAs, if authenticity isn't a concern, then the sky really would be the limit.

Link to post
Share on other sites

Interesting thread!

 

I have a fair amount of experience with PCB Design and I think it would be pretty interesting to see a remake of the SE/30's PCB, my biggest bottleneck would be to make all the individual parts and models to put on the board.
Routing traces is just a fun puzzle and some math

 

Might consider giving it a try if I can find a way to convert KiCad files to AutoCAD EAGLE

Link to post
Share on other sites
8 hours ago, smiba said:

Might consider giving it a try if I can find a way to convert KiCad files to AutoCAD EAGLE

There is only one way: redraw them.

That's the same reason I didn't start working on it yet. Can't be bothered to do this with KiCad :D

I have a lot of the Apple ICs in my personal Eagle library already and I am adding more every now and then when I find some time.

So eventually I will get this started.

 

On 9/2/2020 at 10:35 AM, MrFahrenheit said:

would there be any way of improving upon the SE/30 design when remaking it?  As in, could a different type of RAM socket be integrated?  Support for a faster CPU (like, say, the 68030 40MHz from the IIfx)?

Yes, I see several possibilities... implement a clone of the logic that's found on the DiiMo or PowerCache right into the logicboard to get 50MHz CPU/FPU + Cache or solder 128MB in high capacity RAM chips right to the board to get rid of the SIMM slots altogether.

Also what you'll definitely find on a new SE/30 logicboard (if I am doing one) will be four PLCC sockets for xxF040 ROMs to get rid of the ROM SIMM socket as well. That one gets damaged anyways when the battery goes off, so you won't be sourcing that one from your old logicboard.

Implementing ethernet right into the logicboard also wouldn't be hard and could be easily done.

 

Edited by Bolle
Link to post
Share on other sites

Well that's a bummer... it is after all just a schematic, you don't need a full conversion to start laying out a PCB, generating a netlist would do the trick for starting out, then it's just a matter of making sure the parts are linked to identifiers in your parts library.  If I understood correctly, I thought netlists are a fairly industry standard format due to their simplicity, but I'd vouch I could whip up a converter script if necessary on KiCad's generated netlists.  Then it should be possible for you to get running in Eagle.

Link to post
Share on other sites
15 hours ago, Bolle said:

There is only one way: redraw them.

That's the same reason I didn't start working on it yet. Can't be bothered to do this with KiCad :D

I have a lot of the Apple ICs in my personal Eagle library already and I am adding more every now and then when I find some time.

So eventually I will get this started.

 

Yes, I see several possibilities... implement a clone of the logic that's found on the DiiMo or PowerCache right into the logicboard to get 50MHz CPU/FPU + Cache or solder 128MB in high capacity RAM chips right to the board to get rid of the SIMM slots altogether.

Also what you'll definitely find on a new SE/30 logicboard (if I am doing one) will be four PLCC sockets for xxF040 ROMs to get rid of the ROM SIMM socket as well. That one gets damaged anyways when the battery goes off, so you won't be sourcing that one from your old logicboard.

Implementing ethernet right into the logicboard also wouldn't be hard and could be easily done.

 

That sounds really awesome!  I would totally buy a board that could do these things.

 

What about a standard motherboard, that could be used inside the standard SE/30 case, but also could be used inside a custom-made case, with external monitor only?  Maybe like mATX modifiable?  Powered with ATX power supply, mounted in a mATX case, etc.?

Link to post
Share on other sites

It should be quite possible to do a variation with digital video out, incl. color support. At a bare minimum we could probably integrate Toby, the Macintosh II's video card, for 8-bit 640x480, because that card is about as simple as possible and very well documented in the Designing Cards and Drivers for the Macintosh Family book, IIRC, but with some ambition maybe we could even clone the Micron Xceed 30 and run up the score with grayscale on the internal CRT. Of course my mouth is currently writing checks I definitely can't cash, but I'm also amazed at the amount of skill we've assembled here already. Thank you all very much for your efforts here and elsewhere on the forums.

Link to post
Share on other sites
On 9/2/2020 at 8:25 PM, Daniël Oosterhuis said:

The eight 30-pin SIMMs could theoretically be replaced with two 72-pin SIMMs, allowing for easier and cheaper upgrading to the maximum of 128MB. One 30-pin SIMM is 8 bits, so a bank of four is 32 bits, which is equal to one 72-pin SIMM.

With the price of a large DRAM (around $5) these days I don't know if you would even bother with SIMMS - a FPGA would be useful in any redesign (attach a 50MHz 68030 maintaining original mainboard timing, ethernet, generate all clocks from one source etc..) and could easily make a single chip look like 8 SIMMs. In case you prefer unmodified ROM and don't want to wait so long for RAM check on every boot, make it possible to configure the RAM amount. (And there's the board space issue solved too, a full 90 square centimetres freed of SIMM sockets.)

Link to post
Share on other sites

Micron Xceed has been cloned, mentioned elsewhere here on the forums.  I think Bolle was the one who created the clone, if I remember correctly.  EDIT: Looks like Bolle did an CPU/cache accelerator clone.  Bur looks like there's other mentions of it out and about.

Edited by quorten
Micron Xceed clone mentions
Link to post
Share on other sites
9 hours ago, aeberbach said:

With the price of a large DRAM (around $5) these days I don't know if you would even bother with SIMMS - a FPGA would be useful in any redesign (attach a 50MHz 68030 maintaining original mainboard timing, ethernet, generate all clocks from one source etc..) and could easily make a single chip look like 8 SIMMs. In case you prefer unmodified ROM and don't want to wait so long for RAM check on every boot, make it possible to configure the RAM amount. (And there's the board space issue solved too, a full 90 square centimetres freed of SIMM sockets.)

Well, an FPGA can also just emulate an 030 while you're at it ... make enough of these optimizations and at some point we might as well just run Mini vMac :) 

Link to post
Share on other sites
  • 2 weeks later...

Another error spotted while being a user of the SE/30 schematics.  Page 4, the RTC chip pin label "D" was erroneously transcribed to a zero.  I've fixed the error and uploaded corrections to my GitHub repository.

 

I haven't yet carefully eyed down the original schematics to verify there are no more transcription errors like these, that's still on the to-do list.

Edited by quorten
Clarification
Link to post
Share on other sites
On 8/23/2020 at 12:09 AM, davidg5678 said:

I too have some corrections. These are for the Sound Interface. (I verified these against Apple's Schematics. )

 

UB11 pin 8 is not connected to anything but it should go to ground.

 

UB10 and UB11 pin 6 are also floating. These are connected to a series of components that are missing from the schematics as well. The reset/programmer's switches are not present at all.

 

UB10 Pin 9 is not connected to anything but it should go to the ground.

 

These schematics have been very helpful in my troubleshooting, but I would greatly appreciate if you could go over the Sound Interface Schematics again and make these corrections. There might be some errors that I missed, so I think this section could use a little more work. Thanks!

The GND connections have been fixed, but would need more info to help with the second issue. I just copied what I saw in the original drawings. If you can provide more detail and another member can help to validate that it's correct, I'm happy to integrate it.

Link to post
Share on other sites
On 8/22/2020 at 11:54 PM, quorten said:

@elemenoh Looks like there's a few outstanding improvements over the past few weeks that you haven't incorporated into your GitHub version yet.  Just wondering, putting this out as an idea, would you rather prefer collaborators on your master repo as opposed to forks and pull requests?  It really depends on what you would want to do as a maintainer, which I'm guessing is still all very new to you.  But, point in hand, there are methods you can use to ease keeping the repo up-to-date without needing yourself to be the single gatekeeper to changes.

Sorry I hadn't been paying attention to the thread. It's great to see additional feedback!

 

I updated the master, but wasn't sure how to deal with merging your fork. Please PM me and we can work on details. I definitely do not want to hold up progress, so let me know what you think would be best to keep things up to date without it getting chaotic.

Link to post
Share on other sites
  • 1 month later...

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...