• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

PEx ROM Project

trag

Well-known member
And four TriPEx ROM DIMMs are built. Should go out this morning (Saturday) unless I sleep past the post office hours.

Two of the units have a big asterisk drawn on the U4 label. Those two are less likely to work. Soldered a U3 where a U2 goes and had to undo things and then that U3 was needed on the other marked DIMM. Too lazy to set chip programmer back up to make another U3. Well it's not just the programmer. Have to set up the dedicated XP system to use it. 20-year-old software.

Guess I shouldn't start after midnight. Soldering gremlins. Don't give them chips and PCBs after Midnight.

Things went pretty easily though. The board I pulled the errant U3 off of should be fine and the re-used U3 saw a little extra heat time, but is probably fine.
 

trag

Well-known member
And now a lull while we wait for the Post Office to move firmware modules around. Is there a way to play "The Girl from Ipanema" on this thing?

Maybe just insert some photos:

Blank Modules from the new batch from JLCPCB:
IMG_0380[1].JPG


IMG_0376[1].JPG

I packed up and shipped all the built modules before it occurred to me to take a photo.

Here's a couple of old modules from the old (2003) Sierra Circuits fabrication, configured for use in the Beige G3. The only difference is the resistors just above the edge connector shift position and 3.3V chips are used instead of 5V.
IMG_0382[1].JPG

My 20+ year-old chip programmer. If you ever get one of these, don't wipe the surface with isopropyl alcohol. The ink smears right off.IMG_0383[1].JPG

BTW, yes @Trash80toHP_Mini that means I posted the TriPEx modules this (Saturday) morning.
 

Trash80toHP_Mini

NIGHT STALKER
trag, those cards are beautiful! Can't wait until they get here, I'm amazed that you were able to get up to hit the P.O. on a vacation Saturday! We'll see how soon they arrive. I'm headed out of town Wednesday night after work for a family reunion next weekend. I may not be able to test them until the following Wednesday or later. :(
 

haplain

Well-known member
IIRC, one of hap's machines was identical to mine right down to the legends on the PLCCs.

@haplain can you resurrect your thread of amazing finds? I think you've got detailed pics of the PExboard you have/had that's the twin of mine. I can't seem to find it.
My quest but sadly this site's changed a few times so I think a lot of things were lost. Pretty disappointing overall but it is what it is. Ill try to snap some pics of the PEX I've got here.
 

haplain

Well-known member
Here's some pics of my boards.... I think the one that was fully working/booting is missing the HD at the moment and the backup of said HD. Out on loan to a fellow 68ker
 

Attachments

  • IMG_6816.JPG
    IMG_6816.JPG
    4.2 MB · Views: 20
  • IMG_6818.JPG
    IMG_6818.JPG
    4.6 MB · Views: 19
  • IMG_6819.JPG
    IMG_6819.JPG
    4.1 MB · Views: 17
  • IMG_6822.JPG
    IMG_6822.JPG
    6.4 MB · Views: 14
  • IMG_6823.JPG
    IMG_6823.JPG
    6 MB · Views: 14
  • IMG_6824.JPG
    IMG_6824.JPG
    4.4 MB · Views: 18

Trash80toHP_Mini

NIGHT STALKER
Thanks for the pics, hap. That last looks like the EVTA board you'd posted that's the like mine and TriPEx. Interestingly, your PLCC chips are soldered, not socketed. Don't know offhand, but That may be the only instance where I've seen that.

The others are of an EVTE level board, like the one at DigiBarn. Interestingly, yours is stickered DVT in a couple of spots, what's the dif?

Looks like every PEx we have pictures of shows the same coding on the PLCCs? Makes no sense if the sounds don't match unit to unit does it? My guess is that only the startup chime is in the bootstrap ROMs and the "KILL ME!" variations must be in the ROM DIMMs?
 

trag

Well-known member
trag, those cards are beautiful! Can't wait until they get here, I'm amazed that you were able to get up to hit the P.O. on a vacation Saturday! We'll see how soon they arrive. I'm headed out of town Wednesday night after work for a family reunion next weekend. I may not be able to test them until the following Wednesday or later. :(

Re: Getting up before Noon. I've been pretty strict about getting up in the morning, because I'm coming off of 16 months of working nights. Switching back to sleeping at night is a bear. I needed equipment at the office at least a few times a week.

When I went into the office I had to take the DUTs with me. The DUTs got too big to move conveniently. I got tired of hauling them back and forth. Solution was to work from office 100% of time. By working at night, management had no reason to worry that I wasn't staying isolated from other folks coming in to get and drop off stuff. I never discussed it with them. I just started workign at the office 100% of the time, but at night.

Meetings were a pain. I had to be awake in the day time...

No problem if you're delayed testing them. I just hope they work.
 

trag

Well-known member
The others are of an EVTE level board, like the one at DigiBarn. Interestingly, yours is stickered DVT in a couple of spots, what's the dif?
Apple may do things differently, but EVT is usually engineering samples, where the hardware may still be under revision. DVT are samples where the hardware engineering is finished (everybody hopes) and are used for software/firmware development. There may still be some hardware tweaks if bugs are found or they burst into flame, but DVT samples should be pretty close to final electronics hardware, although the fit and finish may still be way off. Things like the new product's motherboard in an older model's case, even though there will be a new case.
 

jimjamyahauk

Well-known member
Did some more ROM digging and have found some "key dates".
These are based on when Open Transport driver was saved/encoded.

Assuming these are correct, then the TriPex and Digibarn ROMs are only a few days apart. I also did a DIFF comparison using HexFiend between these, and also the TriPex and the 9600 - attached if of interest.

FROM7ROM (9600)
Wednesday, May 21, 1997DeclROM for OpenTxptMacintosh CPU Family 6.0Copyright Š 1986-1995 by Apple Computer, Inc. All Rights Reserved

TriPEX
Monday, November 24, 1997DeclROM for OpenTxptMacintosh CPU Family 6.0Copyright Š 1986-1995 by Apple Computer, Inc. All Rights Reserved.

DigiBarn PEX
Saturday, November 29, 1997DeclROM for OpenTxptMacintosh CPU Family 6.0Copyright Š 1986-1995 by Apple Computer, Inc. All Rights Reserved.

9600 From7ROM
Screenshot 2021-07-31 at 21.26.41.png
TriPEX ROM
Screenshot 2021-07-31 at 21.06.33.png

DigiBarn PEX
Screenshot 2021-07-31 at 21.06.09.png

DigiBarn vs TriPEX ROM
Screenshot 2021-07-31 at 20.36.40.png

TriPEX ROM vs DigiBarn
Screenshot 2021-07-31 at 20.24.26.png
 

Attachments

  • ROM compares HexFiend.zip
    3.5 MB · Views: 1

jimjamyahauk

Well-known member
PS> Does anyone have any settings or tips to view the rest of the ROM info - otherwise assume only text strings are human-readable - the rest is code?
 

jimjamyahauk

Well-known member
Here's some pics of my boards.... I think the one that was fully working/booting is missing the HD at the moment and the backup of said HD. Out on loan to a fellow 68ker
Apols if I've missed this - did we have a full dump of your ROM at somepoint @haplain (assume if an-OS dump then will be combined SIMM and PLCC chip dump - but this might not be the case). Interested to know as the PLCCs are soldered what the final version of the code is on them (unless there was some sort of in-house chip flasher for the OS, but assume it's more likley that they were simply finished. At the very least interesting to know what the OF date build date it reports is?

That's something else if we get into OF on these machines (without ROM SIMM) is to find the declared build date.
 

Trash80toHP_Mini

NIGHT STALKER
Cool stuff, that ROM differentiation info, if Greek to me. :confused: Those soldered PLCC ROMs stuck out like a sore thumb to me.

I'll take a stab at it, each Rev's build date will be the same with identical dumps. Else any feedback to Apple would have been all but useless. If any, minor code tweaks like that between A and E Prototype PLCC Revs could prove significant I'd think?

I'm thinking that SIMM+PLCCs would be unified for the soldered pad implementation on shipping units, with the ROM DIMM socket providing the hardware/traces required to put the kaibosh on the soldered ROMs of shipping units having bare PLCC pads?
 
Last edited:

trag

Well-known member
PS> Does anyone have any settings or tips to view the rest of the ROM info - otherwise assume only text strings are human-readable - the rest is code?

That's my understanding. Viewing it in a HEX editing utility is probably the best, because they typically decode ASCII for you so you can see the human readable bits. But the rest is probably code.

The segments above the Open Transport copyright look like some kind of filler though, but not blank. Weird.
 

trag

Well-known member
On shipping models, if the Power Express had ever shipped, the Firmware would have been put here:

IMG_6818.JPG
 

trag

Well-known member
If the firmware modules I've already built don't work, because of some kind of issue with ROM dumping grabbing PLCC code as well as DIMM based code, or some such, I have an idea for a board with socket that can read the DIMM code out directly in four chunks ready to program to chips -- with no need to disassemble the DIMM. Just plug it in a socket and read it four different ways.

But I hope we don't have to go there. James, do you think there's any chance of getting another loan of the Digibarn DIMM, if that becomes necessary? Or, for that matter, I could ship the reading equipment to them....

One thing that might be an interesting experiment...

If the DIMM doesn't work. Try removing the PLCC chips and try the DIMM with the PLCC chips removed. If the dump grabbed PLCC code as well as DIMM code that should remove contention between the DIMM and PLCC.

@Trash80toHP_Mini jt, you may wish to try the above with the Digibarn DIMM(s) that you have. If you can remove your PLCC chips without risk.
 
Top