• 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.

PB 165 clock madness

alexGS

Well-known member
It felt a bit like that today in NZ when daylight savings started, but only by one hour.

Ummm… what a bizarre fault. If I had to guess, I’d say the time-of-day clock isn’t dividing the oscillator frequency correctly. A relatively high frequency is divided into small ‘ticks’ for accuracy (i.e. to keep time to within a second a day, it would have many ticks per second). Yours is either running at a much higher frequency than intended, or not dividing correctly. I think if it was the former and shared with the processor, nothing else would be functioning.
 

3lectr1cPPC

Well-known member
I have definitely seen this before on other threads, although I forget which ones and what the cause was. May have been software, may have been hardware. You aren’t the only one though. Pretty sure someone had it happen with an SE/30 or another compact.
 

alexGS

Well-known member
I have definitely seen this before on other threads, although I forget which ones and what the cause was. May have been software, may have been hardware. You aren’t the only one though. Pretty sure someone had it happen with an SE/30 or another compact.
https://68kmla.org/bb/index.php?threads/se-30-time-frozen.40797/#post-441488 seems to be the relevant thread :) I notice in the last photo, there is a small DIP chip and crystal oscillator dedicated to the Real-Time Clock. Now I guess the task is for croissantking to find the same items in the PB165…
 

Aektann

Well-known member
I had the same issue on my PB180c. Fixed it by removing all daughterboards and memory, cleaned them and reseated in place again.
 

3lectr1cPPC

Well-known member
Worth mentioning that I found that thread I was talking about before. It was an SE/30 and it was caused by a bad RTC (real time clock) chip.
 

alexGS

Well-known member
Worth mentioning that I found that thread I was talking about before. It was an SE/30 and it was caused by a bad RTC (real time clock) chip.
Nice work maintaining continuity of a discussion, it’s what makes this forum great when searching for similar problems :)
I notice that in compact Macs that suffered a battery explosion, the tube-shaped quartz oscillator for the RTC is fragile and affected by corrosion. That thread contains good instructions for testing the 1Hz pulse. Hopefully a Powerbook 165c has the same.
 
Last edited:

3lectr1cPPC

Well-known member
My 180c doesn’t do this. Might be related to if the battery is holding marginal charge that’s screwing with something? Not sure how that would affect anything while it’s running but who knows.
 

rollmastr

Well-known member
My 165 also does this since the PRAM battery is flat. I could order a new one and replace it.

One observation I made... I cannot set the correct time in System 7.5, but I can in System 7.1. Once I've set the time in 7.1 the RTC works correctly, even after multiple reboots and booting back to 7.5.

/edit: Screw it, I just ordered a VL2330. Should arrive early next week.
 
Last edited:

croissantking

Well-known member
I have this on all My PowerBook 180s. From what I've been told The PRAM battery is dead which is shorting some part of the RTC.

All of this is located on the Jedi board. (The board interconnecting the screen and the logic board). Il theory, replacing the battery fixes this but I've never tried yet.

Relevant info : https://68kmla.org/bb/index.php?threads/replacing-powerbook-180-pram-battery.6119/
You have the same problem as me with the clock in ‘fast forward’?

I can confirm it’s not a dead PRAM battery, as I already tried soldering in a brand new one in an attempt to solve the issue.
 

croissantking

Well-known member
Just came across @timtiger ’s thread again today and would really like to identify the RTC chip on my PB165. I think it’s likely to be on the daughtercard as I tried a different main motherboard and the clock issues persisted.

Can anyone help me identify which chip it is?

I have a dead parts daughtercard that I could swap a replacement chip in from.

Incidentally I have identified another issue where the PB won’t boot with more than 8MB RAM installed.
 

croissantking

Well-known member
Would tracing the connections from the battery, across the interconnect board, to see to which chip they lead be a good start?
 

croissantking

Well-known member
Would tracing the connections from the battery, across the interconnect board, to see to which chip they lead be a good start?
So I did this and started sketching out a schematic:

IMG_4096.jpeg

It connects up to LT1020CS, a regulator/comparator. Yes, of course this makes sense, as VL2303 is a rechargeable battery so it doesn’t connect straight up to the RTC. Additional logic is needed to charge the battery and maybe for things like overcurrent protection.

Pin 2 of LT1020CS (VOut) connects to logic ICs at U10 and U11, and then there is a single trace back to an Apple ASIC on the CPU card (343-1086-2). I speculated this might be the chip responsible for RTC functions; I desoldered and replaced it with a spare chip, but it didn’t change anything.

I am probably missing something because with the power supply disconnected, ~3V from the backup battery only goes to LT1020CS. No other ICs seem to be powered. But whichever chip holds PRAM values should also have a steady 3V.
 
Last edited:

rollmastr

Well-known member
My 165 also does this since the PRAM battery is flat. I could order a new one and replace it.

One observation I made... I cannot set the correct time in System 7.5, but I can in System 7.1. Once I've set the time in 7.1 the RTC works correctly, even after multiple reboots and booting back to 7.5.

/edit: Screw it, I just ordered a VL2330. Should arrive early next week.

Followup: I replaced the dead PRAM battery and right after the clock still behaved just as it did before. I booted 7.1, set the date and time and now even after sitting for several weeks without a (main) battery the machine behaves as it should and displays the current date and time.
 

croissantking

Well-known member
This has worked!!
Followup: I replaced the dead PRAM battery and right after the clock still behaved just as it did before. I booted 7.1, set the date and time and now even after sitting for several weeks without a (main) battery the machine behaves as it should and displays the current date and time.
 
Top