Clock pulse oscillator using crystals

[Konrad Muszynski]

Would anyone know of a circuit schematic for a clock pulse generator using
crystals? I'm trying to use it to build a digital clock that will count to
1/100ths of a second, and need a simple but accurate and precise pulse
generator that doesn't fluctuate with temperature change. I've read that
common 555 timers are pretty unstable when temperature changes and can vary
their frequency.

 

[David L. Jones]

Would anyone know of a circuit schematic for a clock pulse generator using
crystals? I'm trying to use it to build a digital clock that will count to
1/100ths of a second, and need a simple but accurate and precise pulse
generator that doesn't fluctuate with temperature change. I've read that
common 555 timers are pretty unstable when temperature changes and can vary
their frequency.

A 1MHz TTL crystal oscillator module with a divide by 10000 counter
using 2 x 74HC390 dual decade couters will do the trick.
There are other more elegant and cheaper ways, but that's a basic one.

You are correct on the 555.

Dave

 

[The Real Andy]

A crystal, a pic and a thermistor or two.

they say a crystal from a old watch is very accurate.

 

[Doug Laidlaw]

A crystal, a pic and a thermistor or two.

they say a crystal from a old watch is very accurate.

Either DSE or Jaycar has a pre-built unit. If you want extreme stability
against temperature changes, you will need an oven for the crystal.

Doug, VK3KDI.

 

[David L. Jones]

they say a crystal from a old watch is very accurate.

No more so than an ordinary crystal, both are around the 20ppm/degC
mark. Watch crystals are 32.768KHz, great for getting 1Hz, but a
bugger for 100Hz.

Dave

 

[Doug Laidlaw]

Either DSE or Jaycar has a pre-built unit. If you want extreme stability
against temperature changes, you will need an oven for the crystal.

Doug, VK3KDI.

Just opened up my "Amateur Radio" mag for this month. It has an article
about using old crystal oscillator modules from computers. These can be
obtained from Farnell Electronics and RS components, or from junked 386
motherboards. The mag. is available from newsagents.

Doug VK3KDI.

 

[Phil Allison]

Would anyone know of a circuit schematic for a clock pulse generator using
crystals? I'm trying to use it to build a digital clock that will count to
1/100ths of a second, and need a simple but accurate and precise pulse
generator that doesn't fluctuate with temperature change. I've read that
common 555 timers are pretty unstable when temperature changes and can vary
their frequency.

** Just *exactly* how long do you want this 1 / 100 of a second
accuracy to hold ??

For 1 hour that means 1 part in 360,000.

For 1 day means 1 part in 8.6 million.

For one month means 1 part in 260 million.


A crystal based oscillator will do it for an hour if the temp is
fairly steady - after that you are talking serious technology.



............. Phil

 

[Konrad Muszynski]

The crystal oscillator that I'm trying to build has to produce some stable
frequency which will be input into a Schmitt Trigger which hopefully will
shape the pulse into a square wave and then will proceed through a series of
Decade counters DIV10 which again hopefully will be input into a number of
counters that will count in 1 second or 1/10th or 1/100th of a second.
Depending on the number of times I divide the frequency. This will then be
displayed on LED designated as minutes, seconds, milliseconds. The longest
time before the clock would have to be reset is 2 min. And it would have to
operate in temperatures ranging from 15 degrees to 50 degrees.
The actual clock would be activated by a magnetic reed switch that would
be passed over a magnet and toggle the clock into On state intially and then
reset about every 2 min. Does this sound like it will work?

Konrad

 

[Phil Allison]

The crystal oscillator that I'm trying to build has to produce some stable
frequency which will be input into a Schmitt Trigger which hopefully will
shape the pulse into a square wave and then will proceed through a series of
Decade counters DIV10 which again hopefully will be input into a number of
counters that will count in 1 second or 1/10th or 1/100th of a second.
Depending on the number of times I divide the frequency. This will then be
displayed on LED designated as minutes, seconds, milliseconds. The longest
time before the clock would have to be reset is 2 min. And it would have to
operate in temperatures ranging from 15 degrees to 50 degrees.
The actual clock would be activated by a magnetic reed switch that would
be passed over a magnet and toggle the clock into On state intially and then
reset about every 2 min. Does this sound like it will work?

Konrad

** 2 minutes is not a **clock** !!!!!!

That is freaking egg timer !!!!

A 555 would probably do, with polystyrene caps - or any crystal
oscillator that exists.



.................. Phil

 

[David L. Jones]

The crystal oscillator that I'm trying to build has to produce some stable
frequency which will be input into a Schmitt Trigger which hopefully will
shape the pulse into a square wave and then will proceed through a series of
Decade counters DIV10 which again hopefully will be input into a number of
counters that will count in 1 second or 1/10th or 1/100th of a second.
Depending on the number of times I divide the frequency. This will then be
displayed on LED designated as minutes, seconds, milliseconds. The longest
time before the clock would have to be reset is 2 min. And it would have to
operate in temperatures ranging from 15 degrees to 50 degrees.
The actual clock would be activated by a magnetic reed switch that would
be passed over a magnet and toggle the clock into On state intially and then
reset about every 2 min. Does this sound like it will work?

Konrad

Yes, that's a trivial circuit. Any crystal based oscillator module or
circuit will do the trick. No need for the schmitt trigger, as the
crystal oscillator circuit will give you a clean TTL square wave
(unless of course you build a crystal sine wave oscillator, but that
would be silly)
Are you using another reed switch to stop the count too?
A reed switch will have a certain amount of latency time to open or
close, you probably won't be able to get 1/100th second accuracy with
a reed switch, more like 1/10th second. But it depends on what you are
doing.

BTW, you wouldn't start and stop the actual oscillator, it would be
free running. You would gate the clock and latch it on/off with a flip
flop.

Regards
Dave

 

[David L. Jones]

Just opened up my "Amateur Radio" mag for this month. It has an article
about using old crystal oscillator modules from computers. These can be
obtained from Farnell Electronics and RS components, or from junked 386
motherboards. The mag. is available from newsagents.

Good idea.
Would be cheaper to get an old motherboard and rip the module off than
to buy a new module!
Make sure you get some nice fast socketed SRAM with that too, they are
handy.

Dave