PC/electronics help needed

w_tom wrote:

AC electric cannot vary wildly because all grid generators
must match up perfectly with all others - or the entire grid
collapses. To find a clock that varies wildly, visit a PC
clock. That clock is way too inaccurate to control the 50 Hz
grid - where timing accuracy is essential.

Power grid frequency varies automatically with the difference between
supply and demand and is controlled by opening and closing the steam
valves of the turbines that drive the generators. In the continental
European grid (UCTE), the spec says that the time-scale defined by the
rotating grid phase vector must remain within +/- 30 seconds of UTC. The
UCTE control center in Laufenbourg, Switzerland sends out instructions
every 24 hours to shift the nominal frequency of 50.0 Hz, if necessary, to
either 49.99 or 50.01 Hz (+/- 10 mHz) if the network clock starts to drift
away from UTC too far. For comparison: the measured frequency deviation
from the nominal frequency is not supposed to be more than 40 mHz for more
than 10% of the time and not more than 60 mHz for more than 1% of time.

As such, at any moment in time the AC grid can 'vary wildly' (subjective)
compared to the tolerances of a PC clock (typical RTC 32.768 kHz crystal
being 20ppm @ 25C, which can accumulate to being off 52s after a month).
The difference is that the power grid is routinely (24 hours) 'corrected'
to bring it back so, over the long term, it remains within tolerance
whereas a PC clock's error just keeps accumulating, unless one does a
similar thing and syncs it with a more accurate time reference.


Use a PC as the timing source. Therefore you only need
write special software to write to displays- ie driven through
parallel ports. But then why bother with fancy hardware
displays? You can put up clock displays in Windows on the
monitor.

Previously posted was using an alarm clock 'kludge' - every
other display is independent of the actual clock - except when
the display number is first latched from a master display.
Appreciate everything posted up to now is perfectly matching
your request - because details are missing. What are the
details? For example, the grid 50 Hz remains a most accurate
source of timing which is why AC electric clocks remain on
time every day - without resetting every week. If 50 Hz AC is
not accurate enough, then anything in a PC or in the embedded
processor is even worse. IOW you did not provide a necessary
detail such as the number for timing accuracy. You did not
say where those other displays get their numbers.

Are you trying to display a time of day, or time of event.
For that matter, what is wrong with a controller from a
defective microwave oven?

How to connect to a display? First select the display.
Then find the driver chip for that type of display. IOW
download datasheets. Read some 3 or eight pages chock full of
information (assume something like 15+ minutes per page).

If that example (rentron.com) is what you want, then why not
buy it? I remain completely confused by what you are trying
to do. If 50 Hz AC power is not accurate enough, then this
example is even worse. Just more little contradictory details
that completely confused me.

It is possible that you have not yet learned an engineer's
mentality. This project is how you learn so much. Same
reason why to teach why the military want people educated in
the temperament of reality. Notice what you may have
perceived as a simple project might take weeks. Something
that would become obvious once you first list those details.
The military teaches their future officers to be engineers.
To learn that reality is not as simplistic as taught in
secondary school. To learn temperament that comes from putting
initial ideas into reality. Making ideas into reality is
rarely taught in general education. We learn it by doing what
you are trying to accomplish here. Currently, you have starve
your request for information with insufficient details.

Daniel J Beardsall wrote:

The problem with an alarm clock is that the chip is
completely contained. The only information coming out would
be seven segment displays - multiplexed. You would have to
convert that back to some usable form OR build a memory device
to store those multiplexed seven segment outputs and then
drive the other displays. Four digits times seven segments
plus AM/PM is 29 D flipflop latches and LED display drivers.

The other displays would be independent of the actual clock, if that's
what you mean. And yes, when I mentioned butchering an alarm clock that
was mostly as a last-resort kind of option.


What type of time do you want to store? For example, if
time of day, then the alarm clock solution is better. AC
powered clocks synchronize with the 60 Hz power line - whose
accuracy is traceable to the National Bureau of Standards. To
avoid the complexity of same, the alarm clock would be a good
foundation to start construction. You would probably have to
beef up or supplement the alarm clock's power supply to power
your latching registers and two additional displays.

The problem there is that I'm not in the US, I'm in the UK, and our
electrical supply isn't quite so simple. It's mostly 50Hz but can
fluctuate wildly, which often interferes with clocks. Besides, I'm not
looking to run it off direct mains, I'm looking to run it off computer
power, so a battery-powered alarm clock was more what I had in mind.


Another alternative is to start from scrap with a single
chip computer. Again, your timing crystal accuracy is just
another consideration. But a single chip computer affords you
great flexibility to enhance your design.

Now you're talking. But I wouldn't know where to start hooking something
like that into something capable of driving LED segment displays. Which
I suppose is ultimately the help I'm after.

On Fri, 03 Dec 2004 21:01:16 -0600, David Maynard wrote:

w_tom wrote:

AC electric cannot vary wildly because all grid generators
must match up perfectly with all others - or the entire grid
collapses. To find a clock that varies wildly, visit a PC
clock. That clock is way too inaccurate to control the 50 Hz
grid - where timing accuracy is essential.

Power grid frequency varies automatically with the difference between
supply and demand and is controlled by opening and closing the steam
valves of the turbines that drive the generators. In the continental
European grid (UCTE), the spec says that the time-scale defined by the
rotating grid phase vector must remain within +/- 30 seconds of UTC. The
UCTE control center in Laufenbourg, Switzerland sends out instructions
every 24 hours to shift the nominal frequency of 50.0 Hz, if necessary, to
either 49.99 or 50.01 Hz (+/- 10 mHz) if the network clock starts to drift
away from UTC too far. For comparison: the measured frequency deviation
from the nominal frequency is not supposed to be more than 40 mHz for more
than 10% of the time and not more than 60 mHz for more than 1% of time.

As such, at any moment in time the AC grid can 'vary wildly' (subjective)
compared to the tolerances of a PC clock (typical RTC 32.768 kHz crystal
being 20ppm @ 25C, which can accumulate to being off 52s after a month).
The difference is that the power grid is routinely (24 hours) 'corrected'
to bring it back so, over the long term, it remains within tolerance
whereas a PC clock's error just keeps accumulating, unless one does a
similar thing and syncs it with a more accurate time reference.

I believe I know pretty much what you're after. However I have no real hard
fast wiring recommendations. There are many software programs out there
to link with atomic clocks to fix your PC's inaccuracy. I run AtomTime myself
and it's nice to be able to set the PC to within a fraction of a second without
relying on your power main's fluctuating frequency(heh).
If you buy one of the programs, several have options to check/reset the PC
time against the atomic clock on a schedule every so many hours, when the
PC boots up, etc. I fix people's computers (which is more spyware removal
and Winders settings screwups anymore than strictly hardware upgrades) and
am horrified by some of the times (and dates!) I see on some of the systems.

Alternatively you might buy a standalone clock that syncs to the radio signals
transmitted by atomic clocks so it would be totally independant from the PC.
I am assuming they do that in England or Europe. The clock constantly
monitors the broadcasts and updates itself no matter how badly the mains
freq screws with it.

~~~~~~
Bait for spammers:
root@localhost
postmaster@localhost
admin@localhost
abuse@localhost
]

~~~~~~
Remove "spamless" to email me.

Overlord wrote:

On Fri, 03 Dec 2004 21:01:16 -0600, David Maynard wrote:


w_tom wrote:


AC electric cannot vary wildly because all grid generators
must match up perfectly with all others - or the entire grid
collapses. To find a clock that varies wildly, visit a PC
clock. That clock is way too inaccurate to control the 50 Hz
grid - where timing accuracy is essential.

Power grid frequency varies automatically with the difference between
supply and demand and is controlled by opening and closing the steam
valves of the turbines that drive the generators. In the continental
European grid (UCTE), the spec says that the time-scale defined by the
rotating grid phase vector must remain within +/- 30 seconds of UTC. The
UCTE control center in Laufenbourg, Switzerland sends out instructions
every 24 hours to shift the nominal frequency of 50.0 Hz, if necessary, to
either 49.99 or 50.01 Hz (+/- 10 mHz) if the network clock starts to drift
away from UTC too far. For comparison: the measured frequency deviation

from the nominal frequency is not supposed to be more than 40 mHz for more

than 10% of the time and not more than 60 mHz for more than 1% of time.

As such, at any moment in time the AC grid can 'vary wildly' (subjective)
compared to the tolerances of a PC clock (typical RTC 32.768 kHz crystal
being 20ppm @ 25C, which can accumulate to being off 52s after a month).
The difference is that the power grid is routinely (24 hours) 'corrected'
to bring it back so, over the long term, it remains within tolerance
whereas a PC clock's error just keeps accumulating, unless one does a
similar thing and syncs it with a more accurate time reference.


I believe I know pretty much what you're after. However I have no real hard
fast wiring recommendations. There are many software programs out there
to link with atomic clocks to fix your PC's inaccuracy. I run AtomTime myself
and it's nice to be able to set the PC to within a fraction of a second without
relying on your power main's fluctuating frequency(heh).
If you buy one of the programs, several have options to check/reset the PC
time against the atomic clock on a schedule every so many hours, when the
PC boots up, etc. I fix people's computers (which is more spyware removal
and Winders settings screwups anymore than strictly hardware upgrades) and
am horrified by some of the times (and dates!) I see on some of the systems.

Alternatively you might buy a standalone clock that syncs to the radio signals
transmitted by atomic clocks so it would be totally independant from the PC.
I am assuming they do that in England or Europe. The clock constantly
monitors the broadcasts and updates itself no matter how badly the mains
freq screws with it.


Thanks. That's good information but I was just commenting on the power line
frequency and am not the one looking to make the external clock, for what
reason I don't know.

I already sync all my local machines to tick.usno.navy.mil. Or, rather, my
domain controller does and the rest sync with the DC.