Another IIci ROM hack

[olePigeon]

I can't wait!

[olePigeon]

It occurred to me that if you guys end up designing a SIMM programming board that'll work with other models, most notably the 9700 prototype, then that could also be used for the Apple Network Server, correct? From what I understand, you can stick a 9600 ROM in an ANS so you can run Mac OS on it, and not just A/UX, but you lose access to several pieces of hardware (most notably the SCSI controller.)

Would be interesting in the future to not only make clean 9700 ROMs available to 9700 owners, but perhaps enable hardware support on the ANS when running Mac OS.

[Trash80toHP_Mini]

Page 14 and counting!

By the cards arrive and you troubleshoot them, maybe I'll have the spare $$$$$$ to order one!

[dougg3]

Making a programmer board compatible with everything would definitely be more difficult. For starters I would have to add a way to connect to the DIMMs that most Power Macs used, and routing that will be weirder, I think. Luckily, though, we wouldn't be constrained by board area and thickness like I am with the SIMM itself. We'll see! Right now I only have plans to make my first programmer board work with the older II-series SIMM. Even if someone else steps up to the plate and designs something awesome, I'm going to make my own for the experience

[olePigeon]

Sign me up for one.

[Trash80toHP_Mini]

I'm wondering if the SE/30 SIMM might boot the IIsi?

[Trash80toHP_Mini]

GAAAAAAAAAAAAAAAAAAAAAAAA!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

< bangs head repeatedly on surface of 21" CRT face . . . . . . quiets down >

I finally found a set of four of the most decrepit, three chip 30 pin SIMMs and quickly installed them in my IIsi . . .
. . . and they're still only those *&^*&%^&^%&^$ 1MB sticks! I've been trying, for a couple of weeks now, to locate a set of ye olde keychain SIMMs, you know, those 256k POS SIMMs that NOBODY wanted in their Macs if they could possibly help it and weren't worth the price of their PCBs after about 1990!

All this for a simple test of the IIsi w/IIfx ROM SIMM and matching RAM allocated to both freakin' banks that the IIfx ROM seems to wanna' to see to boot!

Well, it's off to the Trading Post I go!

[dougg3]

Update time! I just received my tracking number for the PCBs today from Seeed Studio, and lo and behold, they were out for delivery. So with an order placed on Monday, Oct. 17, that was a 9-day turnaround with the UPS shipping. UPS claims it was shipped on October 24th, so it only took 2 days to arrive here (well, except for time zone differences, so I don't know for sure). Anyway, very nice! Seeed must not have been feeling quite as generous, since I only received 10 this time instead of 12, but since I only paid for 10, you won't hear any complaining from me!

They arrived on the same day as my hardcopy of the Steve Jobs biography from Amazon. Coincidence, or is it a sign from above???

Here are the revision 2 PCBs:

Looks like they turned out nice! I love olePigeon's jolly roger graphic -- with the LEDs it's going to look AWESOME. I wonder what the PCB manufacturers thought when they saw that? The vias are, in fact, big enough for the header now:

You'll notice the second pin from the left is shorter -- I believe this is from me pulling it in and out several times to test stuff out. The header has to go out from the back side with the graphic and capacitors, because if it goes out the front, the IIci's cache slot is in the way when you try to insert it. If other motherboards have similar clearance issues on the back side, I may have to abandon the programming header idea. But I'm hoping to make it work with 64-pin SIMM sockets anyway.

I haven't done any electrical testing, so I may do that just to confirm Seeed Studio's testing, and build one soon! Unfortunately I didn't expect the PCBs to arrive so quickly, so the resistors and LEDs are still en route. Still, I should be able to build one and leave the LED circuit out just to test the functionality. And who knows, maybe the extra parts will arrive soon anyway? We'll see when I get around to building one!

[Trash80toHP_Mini]

VERY, VERY !!!!!!!!!!!!!!!!!!!!!

I can't wait to test one with just about every 68030 ROM image under the sun!
My IIfx ROM SIMM in IIsi saga continues . . .

[dougg3]

Unfortunately, at least one of the SIMM PCBs is not testing OK -- I'm finding several instances of adjacent pins being electrically connected together. It may have happened after I increased the size of the vias above the SIMM contacts to make room for headers to fit. (This was one of the untested boards--one of the tested boards I also looked at does not have this problem). I'll have to try to physically find the problem, and hopefully it's fixable with an X-acto knife! I couldn't find anything after a quick inspection but I'll look at it more thoroughly tomorrow. This means I'm going to have to test every (un-marked) board before I assemble it.

[olePigeon]

[Butthead]This is the coolest thing I have ever seen.[/Butthead]

Oh man, I can't wait.

So the holes are there for possible future programming via a USB board? [Edit: Nevermind, just read it]

[olePigeon]

But I'm hoping to make it work with 64-pin SIMM sockets anyway.

White with metal clips?



Or black with single tension clip?



And how many?

[tt]

And how many?

Cool, I found there's a local shop that also has these, so it looks like there are plenty to go around.

Metal clips are probably better.

[olePigeon]

Certainly less stress on the board to do the 45° and snap in place than to push down until the white clip locks.

[dougg3]

I like the metal clip idea too. So now we have both tt and olePigeon with sources for the 64-pin sockets. Hopefully the ones you both have access to are compatible with each other!!! This is excellent news! I'm excited to make a USB programmer board now.

Now, I will get to figuring out why some of the nets on the untested SIMM are shorted together!

Edit: Upon further inspection, here's what's going on with the shorted stuff:

Definitely fixable with the X-acto knife, but annoying. So it's NOT my vias being bigger causing problems after all. Disappointing especially after my first flawless experience with Seeed Studio, but I can deal with it (especially for the price). The traces look fine, and this only appears to be a problem with the contacts on the bottom of the board; the contacts on the top are fine.

[bigmessowires]

Hmm, why is the section of the board under the edge connector pads a darker shade of red than the rest? Does it not have soldermask on it or something?

[Trash80toHP_Mini]

Definitely fixable with the X-acto knife, but annoying.

CrapShack has a MUCH better tool for that purpose, mine's an angled wedge blade and a bent point probe "connected" by an octagonal yellow plastic handle.

So it's NOT my vias being bigger causing problems after all. Disappointing especially after my first flawless experience with Seed Studio, but I can deal with it (especially for the price)

If you got the board yield you requested along with the "untested overrun," and they're fixable, I'd say your experience has remained flawless!

[dougg3]

Hmm, why is the section of the board under the edge connector pads a darker shade of red than the rest? Does it not have soldermask on it or something?

Yeah, there's no soldermask there. The connectors are so close together that my PCB software leaves it out of the entire area. It looks like I'm leaving 8 mils of clearance around pads and Seeed will let me go down to 0.1 mm (about 4 mils), so I might try 4 mils on the next batch.

Is it solder that is bridging between the pads in that picture? I guess I didn't pay attention...so in that case, the mask would have helped against the problem I'm seeing, right? I would assume they add the solder coating to the pads after everything else is finished with the boards?

Definitely fixable with the X-acto knife, but annoying.

CrapShack has a MUCH better tool for that purpose, mine's an angled wedge blade and a bent point probe "connected" by an octagonal yellow plastic handle.

I live in the middle of nowhere where we don't have Radio Shack anymore Interesting, what is the tool designed for originally?

If you got the board yield you requested along with the "untested overrun," and they're fixable, I'd say your experience has remained flawless!

Well, I ordered 10 boards and 50% electrical testing. I received 10 boards, with 5 of them marked as tested. One of the un-tested boards has that fixable problem and I haven't checked the others yet. I guess this would be a good reason to pay $10 for the 100% testing service. Anyway, not really a complaint on my part, it's not a huge deal. But I'm glad I checked the board *before* assembling it because otherwise I would have been freaking out trying to find shorts in my own soldering, which would be hard to see underneath the sockets

Can't wait to get the programmer board working, because then the chips can be soldered directly onto the SIMM! (Or I can put sockets on anyway if requested by the super hackers)

[olePigeon]

I vote keeping the sockets if it's not much more difficult. If you do the unimaginable and mess up one of the ROMs, it's much easier to replace.

[Arbee]

Hey dougg3,

Nice work I'm glad the MESS ASC code was able to help. Our IIci emulation plays music and game sounds great in both games and things like PlayerPRO so I'm pretty confident in both the wavetable and FIFO mode emulation for the original ASC. The stripped-down ASC inside the V8 ASIC (LC/LC II) has not-quite-compatible registers, and LC III's Sonora is different yet again. The Enhanced ASC (which Apple calls "Batman") removes the wavetable mode entirely and has real variable sample rate and volume control on the FIFOs.

[dougg3]

On second thought, there's only about 10 mils between the pads, so I don't think Seeed will be able to fit much solder mask in there anyway (only 2 mils). Probably not worth changing...

I vote keeping the sockets if it's not much more difficult. If you do the unimaginable and mess up one of the ROMs, it's much easier to replace.

Since I'm not a high-volume operation, I'll ask each person what they want The sockets add some cost to the board, but I see what you mean. For now I have to use sockets anyway, since I don't have the programmer board ready.

Hey dougg3,

Nice work I'm glad the MESS ASC code was able to help. Our IIci emulation plays music and game sounds great in both games and things like PlayerPRO so I'm pretty confident in both the wavetable and FIFO mode emulation for the original ASC. The stripped-down ASC inside the V8 ASIC (LC/LC II) has not-quite-compatible registers, and LC III's Sonora is different yet again. The Enhanced ASC (which Apple calls "Batman") removes the wavetable mode entirely and has real variable sample rate and volume control on the FIFOs.

Thanks Arbee! The ASC emulation in MESS is fantastic! Awesome work! Without MESS there's no way I would have figured it all out. How did you figure out what all the bits do? Was it basically a matter of disassembling the OS/ROM and looking at how the registers are accessed?


SIMM update: The resistors and LEDs have arrived. So I have all the parts needed to build some rev. 2 SIMMs. I may just have to put some resistors and LEDs on an otherwise unpopulated board to test out the glowing eyes circuit with a bench power supply

[dougg3]

Just in time for Halloween.

It's aliiiiive....it's aliiiiiiiiive...

The LED circuit definitely works. My camera doesn't do well with the huge difference in brightness between the rest of the board and the LEDs, but it DEFINITELY works! Now to assemble the rest of the board. More soon...

[dougg3]

Well, I'm fighting one hell of a battle assembling this SIMM. I *really* can't wait to ditch the sockets.

I'm finding annoying shorts between pins. It looks like all the solder paste isn't necessarily flowing nicely where it's supposed to go, or something. Every time I fix a short, I check for other shorts and don't find any, so I wash the board. Then, I come back after it's dry, and there's ANOTHER short! The worst part is that there are 4 possible sockets the short can be on and I can't see any of them, so it's basically "try to get some flux in that location on each socket, heat up the pins, and cross your fingers". Repeat, until the short goes away.

The shorts are gone according to my multimeter -- except for one I CANNOT for the life of me fix between VCC and write enable, but it doesn't matter right now because that would only prevent programming it from the external board -- when I put it in the IIci those pins are connected together anyway.

But it just won't boot. I got it to boot once with a set of smaller EEPROMs inside, but now it won't work at all. I'm thinking there may be a loose pin somewhere, but I'm too frustrated to keep working on it today.

On the bright side (no pun intended), the LEDs work great! They are looking good and they are extremely bright. Maybe too bright!

[Trash80toHP_Mini]

Fabulous progress, doug, it's Miller time!

[dougg3]

Thanks for sure!

I actually went back to it tonight and fixed that last short by adding more solder, which seemed to get the existing solder flowing. I also fixed a loose pin. At this point the SIMM tests perfect electrically but still won't boot.

I think I will start fresh with another PCB, and go back to cutting out the bottom plate for simpler soldering. I'll also leave out the LED parts until I know the SIMM can boot my IIci.