Clock not running....

[prime]

Hi all,

I've been upgrading / restoring a Compaq Portable II (8MHz 80286) that I have been given. Because it hadn't been run for about 20 years I had to replace the CMOS battery with a new one. It takes a 3.6V AA size battery which I was able to source from Amazon.

However even with the new battery in, the clock doesn't seem to be running, though it remembers all the CMOS setting like display and floppy type just fine.
As an example yesterday I set the date to 2020-10-10 (10th oct 2020), and the correct time, when I turned it on today, the date was still set to yesterday and the time was just past midnight.

The odd thing is if I switch the machine off for a short period it seems to remeber things correctly.

Any clues as to what might be the problem?

Cheers.

Phill.

 

[Dwight Elvey]

Hi all,

I've been upgrading / restoring a Compaq Portable II (8MHz 80286) that I have been given. Because it hadn't been run for about 20 years I had to replace the CMOS battery with a new one. It takes a 3.6V AA size battery which I was able to source from Amazon.

However even with the new battery in, the clock doesn't seem to be running, though it remembers all the CMOS setting like display and floppy type just fine.
As an example yesterday I set the date to 2020-10-10 (10th oct 2020), and the correct time, when I turned it on today, the date was still set to yesterday and the time was just past midnight.

The odd thing is if I switch the machine off for a short period it seems to remeber things correctly.

Any clues as to what might be the problem?

Cheers.

Phill.

A lot of these clock chips need to be started by a write to some register. Look us the data sheet for your chip.

Dwight

 

[modem7]

However even with the new battery in, the clock doesn't seem to be running, though it remembers all the CMOS setting like display and floppy type just fine.
As an example yesterday I set the date to 2020-10-10 (10th oct 2020), and the correct time, when I turned it on today, the date was still set to yesterday and the time was just past midnight.
Any clues as to what might be the problem?

When you use DOS' date and time commands, the date and time shown is the DOS clock, not the RTC clock. So I think the first thing to do is to confirm that the RTC clock is advancing when the motherboard is, at the least, powered. One way to check that is to run CheckIt, selecting 'Tests' then 'Real-Time Clock'. Checkit will then show both the RTC clock and DOS clock simultaneously.

If the RTC clock is advancing when the motherboard is powered:

One known cause (i.e. not the only cause) of the 'clock does advance, or clock is slow, but only when the computer is powered off' symptom, is a low battery (low for the particular model of motherboard).
For the Compaq Portable II, the Internet indicates a 3.6V non-rechargeable battery. What voltage does your non-rechargeable battery measure at when the battery is removed from the computer?

The odd thing is if I switch the machine off for a short period it seems to remember things correctly.

Some motherboards have a capacitor in their RTC/CMOS circuitry. An example of that is the IBM 5170 motherboard. On that motherboard, the capacitor is 10 uF and designated C84 (shown in the diagram at [here]). Because of C84, if I want to clear the RTC/CMOS via battery removal, I have to power off the motherboard, remove the battery, then wait a minute or more for C84 to discharge.

The Compaq Portable II may have a similar capacitor.

A lot of these clock chips need to be started by a write to some register. Look us the data sheet for your chip.

I expect that Compaq's SETUP/DIAGNOSTICS tool (SP0308.EXE) for the Portable II will do that.

 

[prime]

Hi all,

I've checked the battery and it measures 3.6V, which seems like it should.

After taking it out and re-placing it I had to re-run the setup program to reset time / date / hdd / floppies / video card.

But it still seems to be suffering the clock problem

Cheers.

Phill.

 

[Chuck(G)]

Got a scope or logic probe? Check to see that the 32.768 KHz crystal is operating. These things have fine lead wires and can get damaged by corrosion or rough handling. Fortunately, they're cheap.

 

[prime]

Got a scope or logic probe? Check to see that the 32.768 KHz crystal is operating. These things have fine lead wires and can get damaged by corrosion or rough handling. Fortunately, they're cheap.

Yep, took the motherboard out, and checked the voltage across VCC and GND of the chip, 3.2Vwhich seems fine and also geting normal 32KHz on pin 2. Chip is a MC146818. I've not checked with the motherboard in the machine as it's hard to get at with the machine assembled.

I suspect this might mean I need a new MC146818.....

Cheers.

Phill.

 

[Chuck(G)]

Should not be hard to find. IIRC, you can drop in a Dallas DS12887/DS1287 into the same socket and be relieved of the external battery issue.