| Click here to select a new forum. |
| PMGR replacement compatibilities |
Posted by: MacPlus2300 on 2026-07-29 13:15:07
I’m trying to replace the power management chip in a sure m5126 Macintosh portable, however the one I got to replace it was also dead. Does it have to be from a portable or PowerBook 100? Can it be from a later model board, or is it not compatible with the original generation or series? |
Posted by: David Cook on 2026-07-29 14:17:15
Yes, I'm sure. I have in fact moved a power mgr chip from a PowerBook 100 into a Portable and had it work.
Have you tried the trick of holding down the interrupt and reset button for 5 seconds, releasing, and then pressing a key on the keyboard. I have two boards that struggle their first cold start after being stored without power. |
Posted by: MacPlus2300 on 2026-07-29 16:15:42
Yes, I'm sure. I have in fact moved a power mgr chip from a PowerBook 100 into a Portable and had it work.
Have you tried the trick of holding down the interrupt and reset button for 5 seconds, releasing, and then pressing a key on the keyboard. I have two boards that struggle their first cold start after being stored without power. I'll try that here soon, the issue I am coming across is that /SYS_PWR stays high no matter what, I am able to pull it down manually with a resistor to get a chime, but that is only a temporary solution. |
Posted by: MacPlus2300 on 2026-08-01 19:10:07
Yes, I'm sure. I have in fact moved a power mgr chip from a PowerBook 100 into a Portable and had it work.
Have you tried the trick of holding down the interrupt and reset button for 5 seconds, releasing, and then pressing a key on the keyboard. I have two boards that struggle their first cold start after being stored without power. I have tried this trick and was able to get it to fire up instantly. The machine dies and doesn’t charge off the adaptor but it’s no issue to me as long as it works. |
Posted by: campbellashe on 2026-08-10 02:30:53
Do you guys know if there has been any efforts to find the firmware or reverse engineer the PMGR chip? As I just went through the efforts of sourcing and now understanding the rarity of this chip, this would be a great project. |
Posted by: daanvdl on 2026-08-10 03:44:44
@campbellashe
I think it should be possible to use a modern microcontroller to emulate the original PMGR MCU with some clever engineering.
At least on the non-backlit Portable there is J21, which appears to have been intended for an external Power Manager module, but as far as I know was never actually used. It breaks out essentially all of the PMGR signals needed by the rest of the machine, so that could make experimenting with a replacement much easier.
I also found a disassembled ROM for the PMGR MCU in rezafouladian’s Git repository. That should be a very useful starting point for understanding its behaviour and eventually reproducing it on a modern MCU |
Posted by: MacPlus2300 on 2026-08-10 07:33:43
I’ve been working on a few projects, one similar to this actually. The chips are plentiful but a cheap replacement would be by far more beneficial. A side project of mine to say the least |
Posted by: liamur on 2026-08-11 05:58:40
I'm really interested in how that person made that source code. The git history on that file is just endless "update pmgr.asm" commits with no other context, and the source code itself seems pretty well documented. The M50753 is a mask ROM microcontroller and as far as I know, there's no way to get the contents of the ROM without decapping the chip and analyzing the ROM area in the die. That's a doable task for a hobbyist, but near the edge of what you can do without industrial equipment.
Of course, I could easily be wrong and there could a program output mode or something in the chip (there appears to be no archive of the datasheet for the variant used in these machines), or maybe the ROM image was part of a dump or leak at some point. |
Posted by: Arbee on 2026-08-12 08:23:52
MAME emulates the PMGR (a Mitsubishi M50753, 6502-derived) as part of emulating the Portable and PB100 so the firmware binary's been available for a long time now. The 140/170 use the same chip with a newer firmware, and the 160/180 have a third version of the firmware.
Most Apple microcontrollers have PEEK and POKE commands so you can easily ask them to dump themselves. |
Posted by: Froggy814 on 2026-08-12 08:42:07
Most Apple microcontrollers have PEEK and POKE commands so you can easily ask them to dump themselves. That absolutely sounds hilarious! |
Posted by: liamur on 2026-08-12 19:10:23
MAME emulates the PMGR (a Mitsubishi M50753, 6502-derived) as part of emulating the Portable and PB100 so the firmware binary's been available for a long time now. The 140/170 use the same chip with a newer firmware, and the 160/180 have a third version of the firmware.
Most Apple microcontrollers have PEEK and POKE commands so you can easily ask them to dump themselves. That makes sense, and I see exactly what you mean in the MAME code for these machines. Handy that Apple made ROM dumping so inadvertently easy. |
Posted by: campbellashe on 2026-08-13 14:54:53
Anyone know of a socket for this chip? It would be helpful for moving between machines and testing… |
Posted by: finkmac on 2026-08-13 20:48:10
sockets for things that aren't DIP, PGA, or PLCC are usually a massive pain. |
Posted by: campbellashe on 2026-08-18 08:41:57
Thanks. Next question; anyone design a pogo-pin style adapter for the J21 on the Portable. I think this is the best place to troubleshoot and get logic data for PMGR replacement options. |
Posted by: campbellashe on 2026-08-23 22:28:42
@daanvdl The disassembled code looks very helpfu, thank youl. I'm thinking about using :
AVR128DB64 (64-pin, ~55 I/O).
- Runs 1.8–5.5V natively — direct 5V interfacing, zero shifters
- 24 MHz, 12-bit ADC, and notably MVIO (multi-voltage I/O) if you ever want a mixed-rail section
- Modern, cheap, in stock, good toolchain
My goal would be to reimplement the handlers natively on an AVR128DB64 — chosen mainly because the AVR-DB family runs at 1.8–5.5V, so it interfaces the 5V bus directly and I avoid level-shifting ~40 lines. The Mac's VIA is a CMOS G65SC22, so a 3.3V part would be marginal driving it anyway. 24 MHz gives ~6× the original's 3.93 MHz, and since the handshake is interlocked there's no timing to chase.
I'm reimplementing behaviour rather than running Apple's ROM image, so nothing copyrighted gets redistributed. Also will clamp the A/D input so it can't be killed the same way.
Questions
- Has anyone attempted this before?
- J21 — can I used this as the replacement interface for testing on the Portable
- What's the M50753's exact lead pitch? I need it for an interposer footprint. I believe 0.8 mm for a 60-pin QFP but haven't confirmed it against a datasheet — if anyone has the Mitsubishi package spec I'd appreciate it.
- Interposer advice. If J21 isn't sufficient, I'm weighing a castellated-edge PCB (likely below standard fab pitch capability at 0.8 mm) against harvesting the lead frame off my dead M50753 and soldering a small board to it. Has anyone done the latter successfully? This is uncharted territory for me.
Would be great to have this figured out. I have a PowerBook 100 and two Portables but only two PMGR. |
Posted by: SuperSVGA on 2026-08-24 06:52:31
You could put the entire thing on J21, but that will only work for the original M5120 model of the logic board, since the backlit M5126 model removes this connector.
One method I was looking into was making a flex PCB for the bottom of the M5126 logic board, since most of the needed connections are available there. |
Posted by: Froggy814 on 2026-08-24 07:21:50
You could do like what @croissantking just did with the rom for the PowerBook 540 to 550. Don't know exactly what to call what he did, but it's on the second to last page of the thread (so far). |
Posted by: daanvdl on 2026-08-24 08:45:42
Just like J21 on the M5120, the M5126 also has a lot of unpopulated pads near the PMGR where standard pin headers should fit. I wouldn’t be surprised if they expose the same signals. |
Posted by: croissantking on 2026-08-24 14:51:25
You could do like what @croissantking just did with the rom for the PowerBook 540 to 550. Don't know exactly what to call what he did, but it's on the second to last page of the thread (so far). Interposers+Mezzanine board. |
Posted by: campbellashe on 2026-08-24 15:32:38
If I can get a castellated board to fit the original pads and put small risers and then a small pcb with the new chip …this would be great. I’m inquiring at PCBWAY for their options as the pitch is small. |
| 1 > |