This is incredible @trag !!!!!!
Could you P L E A S E send me one for testing? 🥹 My rom is already bad so this would be perfect!
Apple Network Server MacOS based ROMs found
- Thread starter eric
- Start date
Could you P L E A S E send me one for testing? 🥹 My rom is already bad so this would be perfect!
This is awesome! Thanks everyone for the effort! I will pull my two ANS out of storage in anticipation of further shenanigans!
Let's let ClassicHasClass do his testing first, to determine whether I made any fundamental errors. I would hate to start assembling these for folks only to find that every single one has some fatal problem.
I made four because that way, if they all fail, it's probably a procedure flaw. If one or two fail, that's probably just an assembly error.
When I was making Rev. C Beige G3 ROMs (over 100) my failure rate was about 1 in 20. But I'm older and my eyes and hands are not as good as they were 20 years ago...
There shouldn't be any problems. I've used this board to make ROMs for the 7200 through the PEx. But there can be odd things. For example, on all the X500/X600 and Power Computing ROMs I examined, pin 122 is tied to ground and it doesn't seem to do anything. It's not connected to anything on the DIMM. But on the Beige G3, if you leave pin 122 tied to GND, then the Beige G3 will not see more than 512MB of RAM. I guess it's signaling a register in the XPC106.
Hopefully there isn't weirdium like that in the ANS.
Anyway, after CHC has tested, and assuming good results, I'll put you at the top of the list. Sad for an ANS to be sitting ROMless.
I'll be sending out the ROMs I make untested, so I'll ask for patience if there are any individual failures. It's too big of a chore to dig my ANS out and get it going (probably have to recap the LB). Plus I'm not sure I left a CPU card in it. Those may be in a box in the attic, because the last time I really messed with it I was trying to get a Mac to ANS CPU adapter working.
Found in my notes from about 20 years ago when I was pinning out various ROM modules by hand, that Pin 28 of the ANS ROM is tied to GND through a 10K resistor.
I hope that isn't necessary for them to work because the modules I made don't do that...
I think I see a way to add a bodge, if necessary, but the modules would need to make another round trip.
Anyone with an ANS ROM handy want to take a close look at pin 28 on your module?
Never mind. Almost made it to bed and then remembered that there's a thread somewhere where someone replaced the chips on a 7100 module to make an ANS ROM. Dug out an old 7100 ROM and checked pin 28 and it is NC. So, assuming that report was a good one, not having pin 28 connected should not be an issue.
https://68kmla.org/bb/threads/mac-os-on-ans-modding-a-rom-dimm.51120/
Last edited:
Totally understandable!! I'm just so excited about this. I so much appreciate your work on this along with everyone who contributed!
A new thread describing conversion of a 7100 ROM DIMM to ANS 2.0 ROM DIMM with more pictures:
https://68kmla.org/bb/threads/my-attempt-at-an-ans-rom.52132/
Those conversions really shouldn't work, unless the 7100 ROM DIMM is supplying 5V on pin 33 for each of the chips. Perhaps it is. Now I wish I hadn't put that 7100 ROM away. I don't want to dig it out again. At least I know where it is...
These chips (AM29F800/HY29F800) can work in 8 bit or 16 bit mode and this application requires 16 bit mode. The mode is controlled by pin 33 and it needs to be high to tell the chip to operate in 16 bit mode.
@joevt Have you ever looked at what the difference is between the X100 ROMs? My notes show they originally shipped as follows, although that could be wrong. E.g. the 0741 vs. 0757 might be the original vs. speed bumped models:
PM6100/7100 => 341-0741
PM8100 341-0757
PM9150/80 => 341-0752
PM9150/120 => 341-0759
When swapping them amongst X100 machines I've never seen any difference in operation. They all seem to work in all the X100 machines, although I probably haven't hit all the possible combos.
Last edited:
I remember talking about that back in the day. Seems like there were too many issues to iron out in the end, as I recall.
There were a number of signals listed in the ANS CPU pinout that weren't in the Apple CPU card, or the other way around or something like that. I had an engineer from NewerTech giving me a little help on the sly; I think he's safe if I mention that now...
But in the end, IIRC, there were about ten signals for which I put jumpers on the card, not being sure whether to connect them or Gnd them or hook them up to Vcc. Then I calculated that there were about 1024 different configurations for the jumpers taken as a set.
Today, if I took another look at it, I think I could figure more of those out by careful reading of the PPC601/PPC604 datasheets. But, IIRC, some of the problems were just ANS machine signals that don't correspond to any of the CPU pins.
It's not entirely just a matter of routing signals to the right place. The Macintosh puts the clock buffer (splits one clock into several) on the CPU card. The ANS puts the clock buffer on the motherboard and feeds one of the split clocks back to the CPU card. On both of them, the original Clock starts on the CPU card.
These are the ROMs:
Code:
077d.20f2.3 "Boot PDM 601 1.0" 9feb69b3 04200897.04838d09.037744f6.0418b271 02dcbf0e.03396836.0265e5ff.02bda730.02b66423.03964e19.02883f03.02e38f21:dedc602cfc3b9221 b40998471e33d802b350f699127e36df "1994-03 - 9FEB69B3 - Power Mac 6100 & 7100 & 8100.ROM"
077d.22f1.3 "Boot PDM 601 1.0" 9c7c98f7 041e75d0.04827aa1.03756ab5.04179702 02dbb3c9.0338b767.0265fa83.02bdd228.02b7254b.0395c376.0287d7a3.02e25af2:e9220fe5992ffdf2 992b36209d478fa5cc1e7f3a3a462c18 "1994-04 - 9C7C98F7 - Workgroup Server 9150 80MHz.ROM"
077d.23f1.3 "Boot PDM 601 1.1" 9b7a3aad 041ddffd.0481ff08.0374fcef.0417b1c4 02db8372.03381349.0265487c.02bcd539.02b5f8f7.0394b68b.02870499.02e28525:ed510ac937721e25 6355d8c6841591e284853f878b99720a "1995-01 - 9B7A3AAD - Power Mac 7100 (newer).ROM"
077d.25f1.3 "Boot PDM 601 1.1" 9b037f6f 041da0cc.048203cb.0374c6cd.0417b38b 02db5f56.03381411.02652fb3.02bce648.02b5dde2.0394ba10.0286e616.02e275d6:d20052bd25d88bd6 f08ad7ba1d8435ecfa24282aab1e0eb0 "1995-04 - 9B037F6F - Workgroup Server 9150 120MHz.ROM"There's a lot of differences. I don't know if they are significant. It would take a while to go through them all.
The
tbxi dump command can split the ROMs into different parts to make them easier to compare.But each part will have a different format so you need to translate them to make them comparable.
e.g. for PEF resources, use MPW's DumpPEF command. For ROM resources (icons, etc.), use ResEdit, For DeclData use my ParseSlots, etc. There's also my ROM Fiend template for my Hex Fiend app fork.
The 9B7A3AAD ROM has some nice color icons of the 3 form factors.
Oh hey, my rom got mentioned!
Boots Mac os 9.0.4 great, couldn't get it to boot into 9.1 at all
Did get it to boot Debian 12, via bootX, but it crashed at the internet device finding state, didn't try it again cuz it was late, will try again tn
Currently has 200mhz, 196mb ram, and 256ks of cache, ordered 512mb, and a 512k cache sim, that'll probably help. Oh, as well as a ATI rage 128, dunno if that's been tested yet, but I will.
Not sure, did someone say the 100/10 Ethernet card works or nah, it recognizes mine, but won't dhcp or let me manually set it :// (in mac OS)
Last edited:
I'll do this some point soon, anyone got anything on Mac OS 9 that can make the dump?
I have tried and tried and tried to get it to read on ****ing anything, and the drive crashed any system I attach it to EXCEPT the ANS, I now have a Mac OS rom, so this should be much easier now.
That seems surprising - did you use the 2.0 ROMs or the pre-production ROMs? What happened when you tried?
Since you have an ESB there, it's possible there's a hardware difference that 9.1 doesn't like. My own ESB is busted so I can't confirm that.
Crashed like how? You wouldn't need to mount it, just dump the raw device. To be sure, the drive from my own ESB I also read on an ANS in AIX.
I used the 2.0 rom, when I tried to boot the CD (bluescsi) it would give me a bus error, and freeze, attempting to open 9.1 after installing 9.0.4 does work, and installs, but will freeze at the happy Mac and never go past.
For the drive, I mean it HARD crashes my systems. Like, will not boot into any OS I've tried, Linux, windows, fricking DOS I haven't tried Mac OS yet because ide have to set up my MDD, but it hard crashes my systems. Hella weird, but again, will read on AIX fine.
Also will mention I've tried 3 different SCSI cards with it, the other drive read fine, and I can unplug the drive, and my systems will boot so it's not the SCSI cards, it's defo the drive causing issues.
Interesting. Do you have cache installed? Which speed CPU?
A very stupid question: is the drive terminated/has a sane ID? The ANS backplane does this for you but otherwise you'd have to jumper it connecting it to any other system.
256k of cache, 200mhz CPU
I'm not particularly sure what you are trying to say here? The SCSI card and cables have termination yes, if you're referring to jumper settings on the drive, I haven't touched anything I just removed from ANS, attached to PC.
I did not, I do not have bd25 to db15 or whatever the weird SCSI is on these deives
Definitely confirm the drive jumpers. If the SCSI ID and term jumper blocks were connected to the tray, the ANS backplane is doing the work for you. When you disconnect it, the drive will not be terminated and be likely ID 0.
256K cache was not a standard loadout for the ANS (it came with 512K or 1MB). Was this the cache stick that came with it, or one you added? You could try removing it and seeing if it behaves better.