IIci shutdown issue (It is now safe to switch off... message)

[AppleTree]

Hi all,

I got my 30 year old IIci from my parents' place a couple of weeks ago.  It had not been switched on for over 20 years.

It did power on but there was no sound.  But eventually it stopped powering on as well after a day or so.

I looked at the board and not surprisingly the surface mount electrolytic caps had leaked so I recapped the entire logic board (including horizontal radial caps) and after that it powered on fine and the sound started working as well.  The power on worked with both the keyboard power switch and the push switch at the back.

I thought all was well but no...  When I went to shut down the machine, instead of a complete shut down, it gave me the message "It is now safe to switch off your Macintosh" as if it was a compact Mac with a physical on/off switch.  The only way to turn it off was to unplug the power cable.

So I recapped the PSU as well (Astec model).  There was just one capacitor I could not replace, as it was in a place that was impossible to reach unless I unsoldered a daughter board, which was too much work.  But apart from that one, I recapped all the others, including RIFA caps.  Unfortunately it didn't fix the issue.  I am still getting the 'It is now safe..." message and it does not switch off.

I am not sure where to go from here.  I looked at the board and all traces look good.  Could the cap leakage have damaged the chips in start up circuit?

(The logic board picture is taken after recapping.  I replaced SMD caps with same ones, to be historically correct ) 

 

[Bolle]

I am not sure where to go from here.  I looked at the board and all traces look good.  Could the cap leakage have damaged the chips in start up circuit?

Traces around there do look good, but a lot of those vias where all the solder has been eaten away do not.

Grab the schematics and check all traces in the startup circuit. Even if they look good they might have failed in places you can’t see.

Those 74s logic chips could be bad as well but I would start with checking traces first.

 

[AppleTree]

I finally got around to looking at this again (only after 9 months  ).  I've fixed the issue.  I am posting my result for anyone who is having the same issue.

I checked all traces and vias using the IIci startup/shutdown circuit schematic.  I found no issues.

In the end, the only thing left to do was to replace the three chips (UB13 - 74HC74, UD13 & UE13 - 74HC132) involved in startup/shutdown.  I desoldered them, checked the traces under them (no issues) and put in new chips and soft shutdown is now working without a problem.

Since I replaced all three with new ones, I do not know which one was actually the issue, but it was probably either UB13 or UD13 since they seemed to be in the shutdown circuit.

And in fact, maybe it had not died.  Maybe it was the cap goo under the chip that I could not clean without desoldering the chips that was causing the issue.  But I didn't want to put the original chips back in after cleaning to find that it still doesn't work so I just replaced them with new ones.

Now my IIci is working as it did 30 years ago when I first got it.  I am so happy!  Next task, upgrade my IIci to 128MB and add more disk space, maybe SCSI2SD   

PS. The sticky residue you see in the bottom of the photo is flux I used during chip removal.

 

[David Cook]

I also received the "safe to power off" message on a Mac IIci rescued from eBay. Attempting to power off by pushing the power button also did not work. Here's how I diagnosed and fixed it.

Aside: The computer also would not reset from the programmer's reset button. That turned out to be unrelated to the power problem. The reset button had corroded for some unknown reason -- whereas everything else around it was fine. Replacing that switch with new-old-stock restored reset functionality.



Now let's focus on the power circuitry. Powering on via switch and keyboard worked fine. Staying powered on worked fine. Swapping power supplies with a functioning IIci did not fix the problem. Therefore the portion of the circuitry that is related only to powering off (highlighted below) is likely at fault. (For impatient readers, I've marked in red the problem that I discovered on my Mac.)


The soft power section of the board looks uncorroded. Recapping was necessary, but didn't fix the problem.



There was one questionable via between pin 1 of UE13 and C12. If scraping or drilling the via results in brittle gray dust, rather than metallic strings, then the metal has been compromised. Threading a wire and resoldering repaired the via. However, this is part of the power-on circuitry, not power off.



With power off, checking continuity for all the other traces and pins did not reveal any faults. Because the power supply physically covers this section of the board during operation, it is necessary to create an extension cable to perform live voltage measurements. I bought a Cerrxian ATX 20 pin male-to-female extension cable (CER-739) from Amazon (26121629). I lightly fit it against each Mac IIci power connector to mate the proper plastic keying and marked the cutoff locations.



Because the cable is wired straight through, it is not necessary to swap any wires. Before you use a hacksaw or snips to cut the connector to size, please make sure your cut marks are the same for both ends of the cables such that the same groups of wires will remain after trimming.

This results in an cheap ($10) power supply extension cable that permits the power supply to be placed outside the case (not resting on the nubus slots as illustrated below).



At this point, I compared the voltages of a working IIci and this one. First, make sure the power supply is providing good 5V, 5V continuous, 12V, and -12V. Although I swapped in a known-good power supply in an earlier step, there could be a short or fault on the bad motherboard that is straining the power supply.

Next, I found that all the pins between the good and bad board had similar voltages except for pin 3 of UD13. That's the power off output to the transistor. What should happen is that when UD13 pin 2 changes state, pin 3 should always be the opposite value (because pin 1 of the NAND is always pulled high and thus the gate acts as an inverter.) In the non-working Macintosh, pin 2 properly changes state during shutdown or pressing the power button, yet pin 3 remained low (actually -- probably broken open). When I used a resistor to drive pin 3 high (5V), the Mac properly powered off. Therefore, I knew UD13 (or nearby corrosion) was the fault.

This section of the board was prepped with Kapton tape and aluminum foil. A desoldering air station made for quick removal of the chip.



As you can see below, the pads and traces are fine underneath the removed chip. There is a little bit of staining, but I concluded that the chip was at fault. I don't know how one chip gate could fail like that, particularly given the series of diodes before the transistor. But, there you are!



I hope this helps other people with steps to track down your soft power issues on similar machines. The replacement power cable is probably the most useful.

- David

 

[AlexADB]

My IIci showed the same odd 'Safe to turn off' message and I could not power it off with the power button - all that started after my recapping which I am a rookie in. Reset Button worked all the time. After scrubbing the UD13 chip's legs with a toothbrush and IPA it immediately started working again. Hopefully it stays like this

 

[Phipli]

Could the cap leakage have damaged the chips in start up circuit?

It will be a trace on the board most likely. You can probably work it out from the description of the power circuit in Guide to Macintosh Family Hardware

https://vintageapple.org/inside_o/pdf/Guide_to_Macintosh_Family_Hardware_2nd_Edition_1990.pdf

 

[rikerjoe]

I had the same behavior with my IIci and found several broken traces around the pad for UD13 pin 6, which is supposed to connect to pin 2 and pin 10. I recommend checking continuity in the power on circuitry.

 

[imactheknife]

The caps all around the startup circuit leak and contaminate the pads and sometimes eat traces / vias. You need a microscope or a good magnifying glass to see up close. I had a mint iici board and it was a short right beside a pad of 47uf 16v. But most of the time the hc 74, hc 132 chips and the 6 little black diodes need to be removed and cleaned especially the one’s closest to leaky caps