• 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.

Difference between IDE controller on 1997 beige G3 and 1998 beige G3

MrFahrenheit

Well-known member
I have a few loose logic boards from the beige G3 series.

One is dated 1997, part number 820-0864-a, and has a 233mhz chip. The other is dated 1997/98, part number 820-0991-b, with a 333mhz chip.

Both boards seem to work just fine. However, there is definitely something different with the IDE controller. On the 820-0864-a board, I cannot boot from my IDE/ATAPI/ATA Fujitsu magneto optical drive.

However, I was playing around with stuff today, and discovered I could boot the exact same drive and disk using the 820-0991-b board.

What’s happening here, and how do I look for another board that might also boot this drive?

Could it be something wrong with the IDE controller on my 820-0864 board?

If anyone has a variety of G3 beige machines and is in the US and is willing to try out things let me know and I can ship you a drive.
 

Phipli

Well-known member
Check the ROM versions for the main computer ROM. Rev A ROMs in beige G3s have a few issues. But if you swap in a Rev B or C the IDE works better... even starts supporting master and slave on both busses!

Only way to check I know of is system profiler. If one of the last 3 characters in the ROM veraion is a "0" it is Rev A. I forget the exact numbers. I'm just bitter because I have three G3s, all Rev A.

Grumble.
 

Phipli

Well-known member
Also, awesome, I've never seen a 333MHz in the wild. I was pleased to find a stock 300 :)

300 probably should have had a Rev B in it, but someone swapped it out.
 

demik

Well-known member
This is a relatively know issue.

Rev-A have a "faulty" IDE controller, which doesn't support all of the IDE specs. For example, a REV-A logic board doesn't support slave devices.
Rev-B fixed most of that, that needed a new ROM as well.
 

Phipli

Well-known member
This is a relatively know issue.

Rev-A have a "faulty" IDE controller, which doesn't support all of the IDE specs. For example, a REV-A logic board doesn't support slave devices.
Rev-B fixed most of that, that needed a new ROM as well.
demik I thought the newer ROM fixed the older machines?
 

Phipli

Well-known member
Yay, found a Rev C ROM :)

Edit : Added System Info
77D.40F2 = Rev A / Rev 1
77D.45F1 = Rev B / Rev 2
77D.45F2 = Rev C / Rev 3
 

Attachments

  • IMG_20220619_234131.jpg
    IMG_20220619_234131.jpg
    3.7 MB · Views: 212
  • IMG_20220619_233435.jpg
    IMG_20220619_233435.jpg
    5 MB · Views: 31
Last edited:

Phipli

Well-known member
Maybe. I didn't check it, it was "from memory"
You have me self doubting now. I've only read it, not tested (believing what you read online is often a mistake :) )... but I have a later ROM and a Rev A board. I'll test them with two drives when I have a bit of time.
 

Powerbase

Well-known member
You have me self doubting now. I've only read it, not tested (believing what you read online is often a mistake :) )... but I have a later ROM and a Rev A board. I'll test them with two drives when I have a bit of time.
I think youmay be confusing it with the Rev A B&W G3s which had a faulty ide controller. I think the later Roms allowed the full use of the ide bus on the Beiges
 

MrFahrenheit

Well-known member
I thought at first it had to do with jumpers on the drive for master/slave/cable select. I thought I had it set to master but it was on CS. Changing it did not change the behavior, I still don’t get any boot on the 233mhz model.

I then checked the ROM versions and the 233mhz model (which does not boot IDE) has 77D.40F2 = Rev A / Rev 1.

I fired up the 333mhz to confirm what it had, and it has 77D.45F2 = Rev C / Rev 3.

I tried swapping the ROM SIMM, but the 233mhz will not chime not boot with the RevC ROM for some reason.

I’m still baffled as to how the behavior can be so fickle. Kind of like some people I know.
 

Trash80toHP_Mini

NIGHT STALKER
His Steveness had returned . . . hence, behavioral issues.
Rushed out, finicky, barely working Rev.1 products shipping, kinda like the 128K. ;)


edit: mines a Model A, how can the 1.0, first of series shipped be called a revision? :unsure:
 
Last edited:

Phipli

Well-known member
I thought at first it had to do with jumpers on the drive for master/slave/cable select. I thought I had it set to master but it was on CS. Changing it did not change the behavior, I still don’t get any boot on the 233mhz model.

I then checked the ROM versions and the 233mhz model (which does not boot IDE) has 77D.40F2 = Rev A / Rev 1.

I fired up the 333mhz to confirm what it had, and it has 77D.45F2 = Rev C / Rev 3.

I tried swapping the ROM SIMM, but the 233mhz will not chime not boot with the RevC ROM for some reason.

I’m still baffled as to how the behavior can be so fickle. Kind of like some people I know.
Might be worth cleaning the ROM contacts and trying again. I swear they're interchangeable. I've swapped A and Cs into a Rev 2 last night and a C into a Rev 1 a few months back.
 

CC_333

Well-known member
edit: mines a Model A, how can the 1.0, first of series shipped be called a revision?
There were probably numerous "revisions" made during the prototyping process, and "Revision A" happened to be the designation given to the first production release.

That being said, it would be interesting to see if there are any earlier versions of the Beige G3 ROM in the wild somewhere, and if so, what, if any differences there may be between them and Rev. A.

I tried swapping the ROM SIMM, but the 233mhz will not chime not boot with the RevC ROM for some reason.
I think the revisions affected more than just the ROM, as the onboard graphics controller was mildly updated as well. If cleaning the pins and/or slot doesn't fix it, this could be why?

Here is a seemingly relevant quote form the Power Macintosh G3 Wikipedia page:
Early G3s with Revision A ROMs do not support slave devices on their IDE controllers, limiting them to one device per bus (normally one optical drive and one hard disk). Additionally, they came with onboard ATI Rage II+ video. G3s with Revision B ROMs support slave devices on their IDE controllers, and had the onboard video upgraded to ATI Rage Pro. G3s with Revision C ROMs also support slave devices on their IDE controllers, but the most significant technical differences are the newer Open Firmware version than the previous two models (2.4 vs 2.0f1) and another onboard video upgrade, this time to ATI Rage Pro Turbo.

c
 

Phipli

Well-known member
There were probably numerous "revisions" made during the prototyping process, and "Revision A" happened to be the designation given to the first production release.

That being said, it would be interesting to see if there are any earlier versions of the Beige G3 ROM in the wild somewhere, and if so, what, if any differences there may be between them and Rev. A.


I think the revisions affected more than just the ROM, as the onboard graphics controller was mildly updated as well. If cleaning the pins and/or slot doesn't fix it, this could be why?

Here is a seemingly relevant quote form the Power Macintosh G3 Wikipedia page:


c
The Rage II+ and Rage II Pro are pin compatible. You could technically desolder one and put the other in.
 

Trash80toHP_Mini

NIGHT STALKER
Revision to the final prototypes? 🤣
HEH! Interesting take. LOL. But common practice for such would would be v.1.0 with Alpha, etc. listed as ZED and decimels leading up to unity.

I know this is a silly tangent in one light, but there's a point to it. What do profiler a/o TattleTech report as the ROM designation and are there variants within the the A/B/C "Revisions" which would be shorthand for such that I don't recall seeing in the DevNotes? Did apple ever announce the Rev.A/B/C designations? Usual Apple practice there would be calling significant change/Rev.B of a model the Enhanced Model.

Here is a seemingly relevant quote form the Power Macintosh G3 Wikipedia page:
Good description. Wikipedia's Mac (all) articles are second source at best, a good start with references listed for looking for source materials of a higher level. WikiPedia (as named) makes Macintosh articles summaries of received wisdom, which often isn't. Such can be seen as an order of magnitude more reliable than LEM articles, but not a whole lot more definitive than what's available on EveryMac?

You won't find the designation Macintosh 128K until after 512K model release, no? Curious about or Rev. A of Macintosh II about Rev. B, dunno about Apple I vs. II?

Just morning musings about about model designations here. Beige is a great example, A/B/C which again would appear to me to be shorthand for such coming into use outside the Infinite Loopiness? If I bother to wake up today I may take look at the DevNotes . . . :sleep:

Meanwhile, casually on the lookout for a Rev. B board and quite interested in this thread.
 

Phipli

Well-known member
Meanwhile, casually on the lookout for a Rev. B board and quite interested in this thread.
I have three Rev B logic boards, but two came with Rev A ROMs, at least one of which is likely to be original. When finding the board you want, look at the video chips, you want one with a Rage II Pro Turbo - there is a mix of Rage II types used on beige G3s.

The Pro does OpenGL (I thought it was the only one that did it at all, but I have been contradicted when I said this before, but have never got OpenGL working on a Rage II +, even with 6MB of VRAM) better / more reliably / at all. It is also a indicator of a later board, which should have the later ROM.
 

Trash80toHP_Mini

NIGHT STALKER
Thanks much for reminding me of that marker. Video capabilities and those relevant here would be the reasons I'm on the lookout. Won't be getting one anytime soon as I'm awaiting the arrival of a "Rev.A" Macintosh II which is very exciting. Reversing this discrete IC NuBus controller implementation/pre-NuChip Controller ASIC variant has been an obsession for many years. So toy money well blown. 😁

Thankfully I'm not at all interested in ever trying to figure out bollixed Beige A and Beige B bodgery. ;)


edit: was so sleepy when I couldn't resist that quip equating the confusion surrounding the "second coming" with the horror show of the Macintosh rollout last night I missed your Profiler pics. documenting the variants. Now I'm talkin' about! Nice work.

77D.40F2 = Rev A / Rev 1
77D.45F1 = Rev B / Rev 2
77D.45F2 = Rev C / Rev 3
 
Last edited:

Trash80toHP_Mini

NIGHT STALKER
Casually wondering if ROM 77D.45F2 was a fix for the rushed, bollixed OS that was the not at all adequate X bodge. Has anyone tried booting from that ROM under X minus .05 on the Rage II variant? That first release "New World" rollout and low level System checks of that gotta pay to try it, public Beta release OS downgrade might be the culprit? I waited for the bugfix release and it still wasn't worth the effort or useful for my production software suite on the DA.

Cannot imagine what it must have been like attempting to install, much less run that BETA bodge on Beige A. :eek:

< /amuse myself commentary mode >

On point: maybe try swapping a known good compatible intro X install boot drive from Beige B into Beige A for immediate crash and burn detection and then swapping out the ROM for your boot test?
 
Last edited:
Top