>
> On  5 Oct 98 at 16:41, VNITHIAN wrote:
>
> >   I have a 486 - Intel PC system. When I turn the system on, I get
> > the following error " No-timer tick interrupt". This problem is
> > intermittent.
>
>   One of the basic PC components is a "timer chip".  It has several
> "pulse" outputs, and the frequency of each is programmable.  [One of
> the outputs goes to the speaker built into the PC case, and this
> provides the "beep" sound and a very crude musical capability.]
>   The most important output from this chip is connected to IRQ 0, and
> set to pulse every 55 milliseconds -- about 18.2 times/second.  [I
> may be rusty, but I believe that's awfully close to 64K ticks/hour.]
>   The code that handles IRQ 0 updates the OS's idea of what time it
> is, which is used to time-stamp files and so on.  It's kind of
> critical that this code get called in a timely fashion.
>
>   The ROM POST (Power-On Self-Test code must program the timer ship
> to generate this signal.  It then probably goes into a loop that will
> last for 60 ms or so, unless it's interrupted -- which a temporary
> handler for IRQ 0 will do.  If the loop runs to completion, the pulse
> signal back from the timer chip never arrived.
>   That's my best guess at what this message is trying to report, and
> it probably means that the timer chip is dying.  This isn't a common
> component to fail, and I'm not sure that it's replaceable.  Your best
> bet may be to see if you can contact the manufacturer of the board
> about a replacement, but odds are good that they'll just recommend
> you replace the motherboard with a modern one -- which means
> replacing your CPU, and probably your RAM and video card.
>   You may be able to locate a used or surplus 486 board that will
> work with your existing components, if your budget is really tight.
>
> David G

If you have access to an oscilloscope, clip on to one or other of the
leads of the clock crystal (a slim silver cylinder next to the timer chip,
probably near the keyboard plug and power plug).  It should oscillate at
32,768 Hz, but as the crystal ages, it may take longer to start oscillating.

If this is the problem, just the capacitance of the oscilloscope probe
(typically 15 pF) may be enough to make it stop oscillating, or help it
oscillate and prevent the intermittent failures.  (The stronger signal
is the output from the timer chip amplifier, the weaker signal has gone
through the crystal and goes into the amplifier.)

If you don't have an oscilloscope, you could switch off, and clip a small -
10 to 15 pF capacitor to either lead of the crystal (Try both.). Connect
through the capacitor (It probably is rated for 200 V, but should be
rated for at least 10 V) to ground - the power supply case will do nicely.)

When you switch on, the problem should occur every time (or never).  If the
change in capacitance makes a big difference, you need to replace the crystal
(cost about $1).  If you are confident that you can unsolder and replace
it without damaging the board, go ahead.  If not, get a technician to do
it for you.

Boyd Ramsay

[log in to unmask]

                                  -----
                PCBUILD mailing list -  http://nospin.com
         Bob Wright:[log in to unmask] - Drew Dunn:[log in to unmask]