Jump to content

SE/30 logicboard recreation


Recommended Posts

  • 68020
20 hours ago, mcpublic said:

Kudos! Where and how did you find the parts? Cannibalizing pin-grid-array chips is hard, and doesn’t the se/30 have some CPLDs? I spotted a VLSI Technologies (VTI) package marking.  They made ASICs, which I’m guessing here is a custom chip. 

Therein lies the struggle.  If we can figure out what the custom ASICs do the entire board could be modernized as Bolle has already made the schematics, and did a beautiful job I must say.  The task can be done, it just requires the right skills and tools, neither of which I do possess.  We've reached a point where almost thing can be done.  New CRTs could be made, metal frames can be made, power supplies can be made, etc. 

Link to post
Share on other sites
  • Replies 140
  • Created
  • Last Reply

Top Posters In This Topic

22 hours ago, maceffects said:

New CRTs could be made


I'm dreaming of this too. I can't help but wonder what a "modern" CRT would look like after 15 years of technological advances since the last ones were designed. IIRC there are just a couple of places in the world still making CRTs for industrial/military purposes. 

Link to post
Share on other sites

@Bolle The Apple Sound Chip is one analog part that was found, in the heyday of its development inside Apple, to be sensitive to PCB variations.  So if the sound doesn't have terrible noise problems, chances are the board is pretty much golden in signal quality handling.

 

Quote

Sure enough, some months later I received a call on a Sunday evening that the Mac factory in Fremont had shut down because of a problem with the sound chip, and that there would be an emergency meeting at 8 AM in Fremont the following morning (with Debi Coleman in attendance). It turned out that the factory engineering team had cost reduced the motherboard, eliminating a couple of layers. This resulted in a significant reduction in the quality of the power planes feeding the sound chip, resulting in faulty behavior. This just goes to show the pitfalls of “designing in isolation” when dealing with larger more complex systems incorporating chips, boards, busses, and power supplies.

 

http://www.byrdsight.com/apple-macintosh/

 

EDIT: Careful re-reading and I noted the commercial version of the ASC is all digital and uses an external DAC.

Edited by quorten
ASC is all digital
Link to post
Share on other sites
  • 68kMLA Supporter
On 3/24/2021 at 5:55 AM, quorten said:

In other notes, did you figure out what is supposed to be on the missing Page 9 of the SE/30 schematics?


Just test points I guess. There's everything that's needed on the other pages. Also the Apple schematics are 100% correct - I just copied everything exactly like it's on there.

 

 

On a side note: I had the modernized schematics open next to them and encountered quite a few errors and mislabeled signals but unfortunately stopped taking notes at some point because there were quite a lot of them and I started to get confused with what to put down in Eagle.

Link to post
Share on other sites

not to be too offtopic here but in regarding to powermac_g4, I could think that maybe due to shrinking electronics and mechanisms it might be possible to reduce the necessary depth of the entire yoke assembly a little bit aside to a slight higher efficiency of the circuits overall. anything else I couldn't be sure about since I'm a little naive to the actual deep girth behind building monitors/tvs

 

of course just as a nitpick I would hope that a modern remake would not ever use pigtails, I'll prefer to see the vga port right on rear just like that (thats been the one ongoing issue with pigtails, they only come in one length which I've found to always be rather bit too long pretty much)

Link to post
Share on other sites
  • 68kMLA Supporter
On 3/24/2021 at 7:39 AM, Bolle said:

On a side note: I had the modernized schematics open next to them and encountered quite a few errors and mislabeled signals but unfortunately stopped taking notes at some point because there were quite a lot of them and I started to get confused with what to put down in Eagle.

Please share you’re notes about the errors you found and I’ll fix them. Quite a few issues were fixed after they were first published. How recently were you using them?

Link to post
Share on other sites

So Hackaday wrote up a piece on this which got posted in one of my Vintage Mac FB groups -- I initially mistook this for the SE board project which I've been following for some time.

 

Major congrats are in order -- I'm so excited to see this! I think we have a couple somewhat dead SE's in storage that I'd love to hack into a working SE/30 someday. 

Link to post
Share on other sites

Holy Mackerel!!! (Or is that Holy Mac-erel!!?)

 

Anyway, this is a fantastic feat of re-engineering. I can't wait to see all you do with it!

 

Imagine a 1:4 scale SE/30 sitting on your desk, with USB ports for modern mouse and keyboard, and an HDMI or Display Port for video out. Similar to the mini C-64's and other minis of that ilk, but not be an emulator. :) :) :) How cool would that be?!!!?

Edited by Tom2112
Link to post
Share on other sites
On 3/23/2021 at 5:38 PM, maceffects said:

Therein lies the struggle.  If we can figure out what the custom ASICs do the entire board could be modernized as Bolle has already made the schematics, and did a beautiful job I must say.  The task can be done, it just requires the right skills and tools, neither of which I do possess.  We've reached a point where almost thing can be done.  New CRTs could be made, metal frames can be made, power supplies can be made, etc. 

 

New CRTs?  This is wonderful!  Where can we read more?  I'm hoping for a source of new vector displays.

Link to post
Share on other sites
  • 68kMLA Supporter
On 3/23/2021 at 8:38 PM, maceffects said:

Therein lies the struggle.  If we can figure out what the custom ASICs do the entire board could be modernized as Bolle has already made the schematics, and did a beautiful job I must say.  The task can be done, it just requires the right skills and tools, neither of which I do possess.  We've reached a point where almost thing can be done.  New CRTs could be made, metal frames can be made, power supplies can be made, etc. 

 New CRTs??

Link to post
Share on other sites
On 3/16/2021 at 2:51 PM, Bolle said:

but I don't see how I would go about the GLUE (yet at least)

 

My general thoughts/experience from the BBU work-in-progress.  Since you know you have accurate schematics, that will greatly facilitate the process.  It's then mostly just a matter of reading Guide to the Macintosh Family Hardware and coding the hardware descriptions.  There will likely be only a few things you'll have questions on that would need to be answered with a logic analyzer, so you don't need to worry about probing all the pins of the GLUE at once.

 

I was trying to think through how much overlap there is between the BBU and the GLUE and I guess because of the use of dedicated video and sound processing, the different CPU, built-in FPU, and 32-bit memory architecture... I think really the only things that can be borrowed readily are the serial communications and SCSI logic.  SE/30 supports real SCSI DMA, though, so that's one more feature to add over the SE BBU.

Link to post
Share on other sites
On 3/23/2021 at 8:41 PM, PowerMac_G4 said:

I'm dreaming of this too. I can't help but wonder what a "modern" CRT would look like after 15 years of technological advances since the last ones were designed. IIRC there are just a couple of places in the world still making CRTs for industrial/military purposes.

 

You're limited by physics.  The beam has to be painted onto the whole screen, ergo it has to be at a distance where it can do that.  Samsung briefly made "super slim" CRTs, but that term is only relative to other CRTs.  They weren't thin by any comparison, not even to early plasma screens.

 

I think the best solution would be a combination LCD + glass.  It could either be just an LCD behind glass, or, projected onto the glass.  If you used old iPad 2 LCDs, the resolution of the LCD would be high enough that you wouldn't even notice the interpolation for non-native resolutions.

Link to post
Share on other sites
  • 68kMLA Supporter

Agreed, I would love to see someone make appropriately-shaped glass panels behind which one could easily mount an LCD.  I think @ants started this work a couple years ago then got busy with other things.

Link to post
Share on other sites

Another note.  I noticed the "factory test" card edge connector was omitted and the free space used for other purposes.  In particular that connects with power, the serial communications, and some of the VIA ports.

 

I'd imagine for our uses, though, a much smaller connector could be fitted, so as to leave the space available for the other uses.

Edited by quorten
Reword
Link to post
Share on other sites
On 3/26/2021 at 9:59 AM, Crutch said:

Agreed, I would love to see someone make appropriately-shaped glass panels behind which one could easily mount an LCD.  I think @ants started this work a couple years ago then got busy with other things.

 

Yeah I did manage to make a polyurethane panel in the shape of the classic CRT. I learned a lot and would love to have another go at it when time permits. Next time round I would include mounting tabs so that the panel could easily be screwed onto the case.

 

127888300_IMG_20190524_1818342-3024x2268.thumb.jpg.ae7485deafaff30b1f3945cdd3cb6a61.jpg.58ef385ed47f0b7bd58249125629d004.jpg

 

IMG_20190525_170759.thumb.jpg.aac5ea667466c4b0426c9f0fa4319cff.jpg.e11cd6c270df2f7eccdf0d6a318ea9b4.jpg

Link to post
Share on other sites
  • 68kMLA Supporter

Did run them without noticing any further issues, so I guess it's fine to have them here...

 

Remember, those aren't dumps of the actual chips but code I rewrote to match the behavior of the originals, so use at your own risk.

If something goes horribly wrong harm to the analog board is possible.

 

Jedec files are for standard 16V8 devices - either real GAL16V8, PALCE16V8 or ATF16V8.

 

I used ATF16V8B-15 for the 5 DIP chips and an ATF16V8C-7 for the single PLCC chip.

Be aware that ATF16V8C behave differently because they don't have internal pullups. I have not tested them except for UH7.

UE6_16v8.JED UE7_16v8.JED UG6_16v8.JED UG7_16v8.JED UH7_16v8.JED UI6_16v8.jed

Link to post
Share on other sites
On 3/29/2021 at 10:24 AM, Bolle said:

Did run them without noticing any further issues, so I guess it's fine to have them here...

 

Remember, those aren't dumps of the actual chips but code I rewrote to match the behavior of the originals, so use at your own risk.

If something goes horribly wrong harm to the analog board is possible.

 

Jedec files are for standard 16V8 devices - either real GAL16V8, PALCE16V8 or ATF16V8.

 

I used ATF16V8B-15 for the 5 DIP chips and an ATF16V8C-7 for the single PLCC chip.

Be aware that ATF16V8C behave differently because they don't have internal pullups. I have not tested them except for UH7.

UE6_16v8.JED 1.68 kB · 4 downloads UE7_16v8.JED 2.18 kB · 1 download UG6_16v8.JED 1.38 kB · 1 download UG7_16v8.JED 1.48 kB · 1 download UH7_16v8.JED 989 B · 1 download UI6_16v8.jed 1.05 kB · 1 download

 

Oooo will have to try that. Need the PCBs first tho. :-P

 

maybe if the logic is ok, shoving it all into a single CPLD for a newer/revised board design? idk. 

Link to post
Share on other sites

@Bolle Thoughts on future release of your recreated equations for the PALs?  I'm just thinking about the possibility of running the logic cluster under simulation like I've done with the Mac 128K PALs... I'm "almost done" sorting those out, but not quite.

Link to post
Share on other sites
On 3/16/2021 at 1:25 AM, Bolle said:

Just like I promised when teasing everyone in the SE logicboard thread here we go with the build-log/information exchange/everything thread on the SE/30 logicboard replacement PCB.

 

The plan to redraw the available SE/30 schematics in Eagle has been in my head now for quite some time. It wasn't the task of having to redraw everything that kept me off, but I genuinely hate creating new devices in Eagle, so there goes that.

Luckily @GeekDot kept bugging me about it and thankfully took over the job of creating Eagle devices for all the custom Apple ICs on the SE/30 board.

This got me started and things quickly went down from there... a month ago I began copying the Apple schematics to Eagle. With most of the devices available this went through in a breeze.

 

Only two days later I was already taking measurements on one of my SE/30 boards to arrange components and get them as close as possible to the original

 

TxJW1xX.png

 

Another week in (not doing much else than drawing blue and red lines all day long) and still somewhat surprised that things went through so quickly I was basically done with a first revision:

 

1.thumb.png.026106ea747eb07b0a3460ebf988904e.png

 

2.thumb.png.66a8a6ad87e6804a23b443bf86950d7b.png

 

3.thumb.png.13b072700d2fbab9b47f9d6ecdfd4258.png

 

4.thumb.png.730132abf6ae20148e805a4b837dcdc0.png

 

Most components found their place in exactly the same spot as on the original board - unless it was useful to move them around a little bit. The RTC for example to make some space for a CR2032 battery holder.

The battery is the only component I replaced with something that's not on the original. (yes I know, 72pin RAM, a single 16bit VRAM IC, accelerators, network cards, grayscale, and a PPC built right in - no)

Goal for me was to set a baseline and verify the Apple schematics are even correct and would result in a running board before thinking about modernization efforts.

 

One major advantage of re-doing the whole routing of the board is that everything can be made to fit into 4 layers, instead of the 6 layers the original board had. This is cutting costs dramatically.

 

So on it goes with ordering boards...

 

It took two weeks to turn this:

 

render.thumb.jpg.b0ac8870041843112ee274901746dae3.jpg

 

into that:

 

IMG_6170.thumb.jpg.bdceca554225bdac85fd3af5f7b5ef43.jpg

 

and a few hours later to find myself here:

 

IMG_6180.thumb.jpeg.fad3c7baa7705c2cba6a08ad2ced16d9.jpegIMG_6183.thumb.jpeg.9d27b1bc484d02c4e399472a29da1260.jpeg

 

Only a few components missing to get her powered up for the first time - I went without any RAM to see if it actually does anything at all before bothering to solder in SIMM sockets.

So a few more bits soldered down, a deep breath and the flip of a switch later:

 

IMG_6184.thumb.jpeg.ad6338dfa8255be049509e9d5f0435c8.jpeg

 

SUCCESS! It turns on at least and I got video. I seriously didn't expect it to even do that.

No chimes of any sort though - would have expected the slow chimes of death because there's no RAM.

At least the screen image is what is to be expected if there's no RAM installed which was a good thing.

 

I took a break here, got some food for dinner, took a shower and while doing the latter it struck me... to get sound you need the headphone port to be present, because of the switch that's in there and connects the internal speaker to the amplifier.

So back to the workbench... two little jumpers on the headphone ports solder pads and there we go - slow death chime was there indicating RAM failure.

 

It was time to go to bed at that point but I couldn't have slept anyways. I got a donor board, moved over the RAM muxes, control signal buffer and missing address line resistor arrays, installed 4 SIMM sockets.

I put in some RAM, hooked the board up again and couldn't believe my eyes when I was looking at that flashing question mark. :?:

 

Well past bedtime a SCSI and floppy port went onto the board, back on the testbed with FloppyEmu and a hard drive connected...

 

IMG_6190.thumb.jpeg.4891fa69d719af988fba626d033ca2d9.jpeg

 

I sh*t you not the damn thing was booting from floppy - first board revision, rushed through copying the schematics, board layout, routing, putting it together... and it just boots. :evil:

Not a single jumper wire (spoiler: yet!)

 

Notice how there's no HD showing up? This is where rushing through things strikes back. :tongue:

I copy-pasted the data lines to the SCSI IC from the nearby SCC - not noticing that they were sorted different in the original schematics. D24-31 on one and D31-24 on the other IC.

End result: the data lines to the SCSI controller were reversed. Nothing that can't be fixed though.

Popped the SCSI controller off again, cut the data lines and installed jumper wires.

 

IMG_6192.thumb.jpeg.933ee8a50d95e5f93401b93e61cdeded.jpeg

 

IMG_6200.thumb.jpeg.4282c0d7bcc40d0e3f5bd05fee1ca5e2.jpeg

 

 

And that's how I ended up here:

 

IMG_6194.jpg.6d972b16018ee6a876b95e5ffd5affcd.jpg

 

 

The board now has a few hours of runtime on it and seems to be rock solid so far.

Couldn't notice any differences from an original.

 

 

Left to do:

 

-test the serial ports (haven't done that yet because I totally never use serial ever)

-test external Floppy and SCSI ports

-chassis fitment

-runtime runtime runtime

 

 

The SCSI blooper has been corrected already for the upcoming rev 2 of the board.

A few SMD pads will be added here and there for the fuses, SCSI Term power diode - those were through hole only for simplicity.

 

 

And a few more images to drool over: https://imgur.com/a/fhTrPZ2

Thai is such amazing news!!! Incredible effort!! Please put me down for two boards!!! Wow! Just wow!!!

Link to post
Share on other sites
On 3/25/2021 at 12:08 PM, olePigeon said:

 

You're limited by physics.  The beam has to be painted onto the whole screen, ergo it has to be at a distance where it can do that.  Samsung briefly made "super slim" CRTs, but that term is only relative to other CRTs.  They weren't thin by any comparison, not even to early plasma screens.

 

I think the best solution would be a combination LCD + glass.  It could either be just an LCD behind glass, or, projected onto the glass.  If you used old iPad 2 LCDs, the resolution of the LCD would be high enough that you wouldn't even notice the interpolation for non-native resolutions.

 

It's quite a shame that SED / FED was effectively abandoned. Not only would it prove a great replacement for vintage CRTs, it might be a superior display technology overall.

Link to post
Share on other sites
20 hours ago, CC_333 said:

Can someone explain what these acronyms mean?

 

c

 

Surface-conduction electron-emitter display / field emission display. Uses phosphors like a CRT and possesses the same advantages (no backlight, great contrast, negligible motion blur, response time) in a flat panel. Also abandoned. :/

Link to post
Share on other sites

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...