Macintosh Portable won't boot after taxing 12V power

[campbellashe]

Okay. I was getting excited and tried to internally power and LCD replacement off the J15 connector -- using 4 12 volt lines and 4 GND lines to power a barrel connector and a 12 volt LCD -- yep. Shouldn't have done this. The LCD did power on; but now my computer will not boot. NO smoke; no pop, no obvious overheating.

What is working:

• F1 good
• +12V converter working
• Main +5V present
• +5V reaches RAM
• No hard short on +5V
• No hard short on +12V

Where I’m currently stuck​

The machine has valid power rails, but:

• SYS_RST* never releases
• RESET stays low
• HALT stays low

The CPU never starts executing.

Looking at the schematic, R13 (1 kΩ) connects SYS_RST* to the CPU GLU NORM_RST* signal. My next thought was to lift one side of R13 to determine whether the PMGR side or the GLU side is asserting reset.

Questions​

1. Has anyone seen this failure after overloading the J15 +12V rail?
2. Is lifting R13 the best next diagnostic step?
3. Is there a known failure in the PMGR, hybrid, or CPU GLU that will permanently hold SYS_RST* low while +5V and +12V remain normal?
4. Is there another signal I should measure before lifting R13?

Thanks for any guidance.

 

[Froggy814]

@techknight this!

 

[MacPlus2300]

Do a full power manager reset. PMGR dictates whether the SYS_RST is released after rails are stable. Measure both sides of R13. If the PMGR side is low than it’s something before the logic at the Power Manager or its input. Also clock the power manager chip to make sure it is running. If the rails are alive than that is an obvious good sign that power is getting to the board. If you have not already I would replace the caps to remove intermittent failures. If this board has a hybrid chip, (M5120) I would replace as well. Also check any sense lines that feed or derive from the power manager. Before anything remove all batteries, power the board directly from a 6.5v bench power supply in the four pin connector (connect VCC and sense line to 6.5v, connect one or any ground to the ground source) and then press and hold both programmer switches for at least 5-10 seconds, release, and then press a key on the keyboard. If that fails, try it once more.

 

[campbellashe]

Thanks @MacPlus2300 . I've gone through much of what you said.....

Recap: After correcting the J15 power tap wiring, the Portable stopped booting. SYS_RST*/PMGR_RESET* stuck at 0V, F1 good, PMGR clock running.

Traced PMGR_RESET* (J21 pin 46) to U15H pin 18 — the PMGR's hardware RESET input — confirmed via two independent schematic sources. The physical RESET button doesn't feed this pin directly; it goes to a firmware-read GPIO instead.

Pin 18 is not stuck low — reads 5.1V, matching a direct chip measurement. But even released, no chime, no boot, and SYS_RST* still reads 0V.

A/D battery-sense line (pin 52) reads 2.7V, matching the hybrid's output — no active overvoltage right now.

Voltage-scaling test shows the hybrid behaving non-linearly within its own normal operating range (5.4-7.5V):

Input	A/D output
6.5V	2.7V
6.0V	1.7V
5.6V	0.4V

Not proportional — points to genuine hybrid degradation (resistor drift), not a bad test voltage.

Chip pin 12 = SYS RESET (confirmed), held cleanly low — not floating, consistent with the PMGR actively running firmware and choosing not to release the system, not a dead chip??

Logic analyzer testing (Rigol DS1104Z Plus + RPL1116), watching pins 12/18/16 plus the GLU source (61/62) simultaneously during AKD triggers: AKD fires correctly (clean pulse confirmed at C116), but nothing downstream reacts — rules out contact resistance in S1 (suggested earlier — appreciated @daanvdl , but the GLU never even attempts a pulse).

Current theory:
a) PMGR is alive, firmware is gatekeeping??? — most likely on the A/D reading, given confirmed hybrid degradation.
b) PMGR is broken -- I just have to accept it?

Next:
1) Thought about isolating R151, injecting a clean ~2.4V directly into the PMGR's A/D input via a breadboard divider, bypassing the hybrid, to test this directly.
2) Do I just replace the Hybrid... as you suggested (I have a replacement in hand)
3) Replace the PMGR when my donor Powerbook 100 comes??

 

[MacPlus2300]

Thanks for the update. I can always vouch for replacing a hybrid. Even if it doesn't help, you still significantly rule out many parts that could have been bad. First, though, isolate R151, inject a stable current limited 2.4v to the A/D input PMGR side of R151; if it releases and you chime, the A/D path is confirmed as the gate. Stays low; look elsewhere. Second, replace that hybrid and confirm output scales correctly across 5.6-7.5v, confirm main 5v, recheck if SYS_RST releases on normal power up/AKD, replace the PMGR if all else fails; that can be a huge step in the right direction. I have saved a board from simply swapping PMGR before. Just a thing to note: the original overload on the +12 V rail may have accelerated hybrid resistor drift or stressed the sense path, and that non-linear A/D is the smoking gun right now. After the known-good hybrid is in, a PMGR reset is worth it again. If all else fails, once again, if a known-good hybrid is in and the A/D is good, confirm if the PMGR is the component holding the reset, which could be a firmware decision or damaged output; check under and around the PMGR physically; check for classic common board-level failures, and for a diagnostic, you can always do more force-release tests or isolations. Hoping this board can be saved. I have many components for these boards if you ever need extra ones, since many parts can fail inexplicably.

 

[SuperSVGA]

PMGR is alive, firmware is gatekeeping??? — most likely on the A/D reading, given confirmed hybrid degradation.

I don't think the A/D line will stop the system from turning on, though it may shut off the system pretty quick if the charger is not detected as connected (from the power manager's point of view). I'll have to double check the power manager firmware to verify, but typically you can boot with a bad/blown A/D line, at least until the OS handles the interrupt.

 

[campbellashe]

I tried the R151 injection of 2.4 volts without any change. Next is the Hybrid....

 

[MacPlus2300]

Next is the Hybrid....

Definitely replace the module. The old ones can be hard to get off, especially with the ceramic absorbing a lot of heat. Solder suckers or a wick can work. After replacement, measure the A/D line again, check SYS_RST and SYS_PWR, and listen on that speaker. If SYS_RST is still stuck low, it is likely the PMGR itself.

 

[Froggy814]

I just cut my old one of with a pair of pliers. Then cracked it the rest. Then took all of the pins out that were still in the holes becuse they are soldered in. Not the cleanest, but it does work!

 

[campbellashe]

What a mess of a time, with cap goo corrosion things were not fun with the hybrid extraction... but it is out and the new one is in.

I used the Pre-Hybrid check as a post hybrid check and resistances are good. Unfortunately still not chiming.....

I have to do some more testing...to make sure that I didn't break anything else.


Then the next step is transplanting the PMGR...

I got a Powerbook 100 off eBay (non functional, it looks good -- I ordered caps to fix it up and hope to get it working in the future), hope to work on the transplant tomorrow. I don't know 100% that it is working but no cap goo around the one on the 100.

 

[Froggy814]

I wonder how many powerbooks are down for the count because of its older brother...

 

[MacPlus2300]

I've revived these PMGR chips before, enough to save some PowerBook 100s and a few Portables. They do get really weird and finicky, but can be saved under certain measures. I've had a Non-working portable PMGR work in a PowerBook 100 and vice versa. Don't worry about the Hybrid, these ceramic ones are super fragile and breakage is common. I would try to see if the board works now with a PMGR reset. If you replace the PMGR and get nowhere, I have lots of spare board level IC replacements I can send over if you need.

 

[campbellashe]

Okay here is a new problem. No 12 Volt line.....

Current measurements with bench supply at 6.5V:

• ALWAYS ON +5V (hybrid pins 11/18): 5.2V ✓
• Main (5/0) rail (J11 VCC): 5.2V ✓
• +12V rail (J15 pin 9): ~0V ✗
• Q14 gate (IRF9230, P-channel): 6.5V = Vgs 0V = OFF
• Q14 source: 6.5V
• VR1 LT1070 pin 5 (VIN): ~0V

The hybrid is regulating ALWAYS ON and enabling the main rail correctly. But Q14 gate is sitting at supply voltage (not being pulled low by the hybrid), so Q14 is off, VR1 has no input power, and the boost converter never runs.

Adding the power brick alongside the bench supply makes no difference — Q14 gate rises with the supply but stays at Vgs = 0.

Questions:

1. Does the hybrid require a signal from the M50753 PRMG before it enables the +12V generation circuit? Is +12V gated on PRMG being alive?
2. Is there a known sequence where +12V only comes up after a successful power-on, rather than at standby?
3. The PRMG (U19H) also shows no response to keypress with both rails present — is this consistent with PRMG damage from the original 12V event?
4. Is this a new problem related to my new Hybrid board?

 

[MacPlus2300]

Check GDS of Q14. Do another full PMGR Reset. Remove the battery and power. Let the board sit for 10 seconds. Replace the battery, press both buttons, and press the space bar. Listen for speaker crackle or chime. The hybrid is fine; now it is just the PMGR and the /SYS_PWR rail. /SYS_PWR should go low when the system is awake. IF you try to turn on the system and it does not get pulled low, then there is some unforeseen error in your PMGR. You can pull down the gate of Q14 manually to see if the LT1070 produces 12v. If so, then your PMGR is dead or faulting. Check the pinout on the Mitsubishi M50753 and ensure that VCC, GND, etc are all good; check the /SYS_PWR net as well.

 

[campbellashe]

Well after all that.... I replaced the PMGR and the sweet sound of a chime. Working great. Thanks for all the help along the way. It was frustrating that I caused the issue; but I got a new Hybrid Module out of the process and a Powerbook 100 that I can fix up. Now I will work on my RAM board again.

 

[MacPlus2300]

Knew the PMGR was a worthy replacement. Glad to see it alive now, a good replaced hybrid too is always important as well.