# how can i make a 1 sec pulse

J

#### Jan

Jan 1, 1970
0
Hi
I want to make a clock, from scratch, without an mcu. First i need to make a
oscilator that gives me a 1 sec pulse (1 hz isnt it?)
How can i do it? with a 555? (and how?)

Jan

E

#### Eugene Rosenzweig

Jan 1, 1970
0
555 would not be precise over temperature variations since you have to use
external resistors/caps.
I got 1Hz pulse using a 4060 driven by 32.768kHz crystal into a d-type
flip-flop. 4060 is a 14bit counter so you get 2Hz pulse at most, from pin
Q14, so divide it further with a D-type flip-flop (74hc74) and you have 1Hz.
Look at the datasheet for 4060 for an example circuit.

Eugene.

H

#### happyhobit

Jan 1, 1970
0
1Hz is 500 msec. on and 500 msec. off. 1sec. on and 1 sec. off is 1/2 Hz.

Jay

F

#### Fins

Jan 1, 1970
0
Jan said:
Hi
I want to make a clock, from scratch, without an mcu. First i need to make a
oscilator that gives me a 1 sec pulse (1 hz isnt it?)
How can i do it? with a 555? (and how?)

Jan
Do you want a 1 second pulse, or do you want a signal with a 1 second
period? A one-shot can be configured for a 1 second pulse, an oscillator
could be configured with a 1 second period. Also, a 1-Hz wave is NOT
necessarily 500 ms on/500 ms off. A periodic signal with 1 ms on /999 ms off
is also a 1 Hz signal, changing the duty cycle within a 1 second period does
not change its frequency.

Fins

R

#### Robert Monsen

Jan 1, 1970
0
Just as a side note, a 555 circuit, while simple to build, will be nearly
impossible to tune so it won't drift, particularly under different
temperature conditions.

If you need accuracy, I'd think about using a watch crystal, or even the
60Hz powerline (which is by some accounts kept accurate by an atomic clock.)
Watch crystals are very cheap, and using a 4049 to drive it is a
cheap/simple way to get an accurate timebase. You would have to divide it
down, though, so your part count will go up.

The other possibility would be an RTC chip, but thats a bit expensive for a
clock, I think.

Regards

N

#### Neil

Jan 1, 1970
0
Robert Monsen said:
If you need accuracy, I'd think about using a watch crystal, or even the
60Hz powerline (which is by some accounts kept accurate by an atomic clock.)
Jan
Some years ago - about 1974 - I made a clock using the utility supply (50Hz
here in UK).
I picked off an AC signal of about 6vrms, fed it through a diode, through a
1K resistor to the top of a 4v7 zener. Other side of zener to 0V. Across
the zener was a 100pF (I think) small capacitor to catch spikes.
The signal then went to a 74LS14 schmitt inverter. The output was a nice
clean 50Hz pulse train.
Putting that through two 7490's resulted in a 1Hz clock pulse for the main
clock circuit.

The fun part was putting the LED displays in a small box, with all the rest
of the circuitry on a cable (hidden at the back of the bedside table).
People thought it was a lovely small clock!

An touch sensitive alarm switch comprised a piece of veroboard with
alternate strips connected so that when I put my finger across some, it
switched a cmos gate - only need a microamp or so to pull down a 500K pullup
resistor.
The alarm was a 74132 quad schmitt nand connected as two oscillators, one at
about 2hz, the other at a nice screechy tone. Gated one into the other, and
conctrolled by clock alarm output and touch switch. Output via transistor
to smal speaker. Woke me up a treat for years.

hth
Neil

B

Jan 1, 1970
0
A

#### A E

Jan 1, 1970
0
happyhobit said:
1Hz is 500 msec. on and 500 msec. off. 1sec. on and 1 sec. off is 1/2 Hz.

Jay

That's nice, but 1Hz pulse is undefined as to the duty cycle. 1Hz is 1Hz wether
it's a peak, a square, or whatever. You're gonna get mighty confused if you
think the way you do. Nobody said anything about 1 sec on....

H

#### happyhobit

Jan 1, 1970
0
You're completely correct. I read something into it that wasn't there.

I stand corrected.

Jay

A

#### AC/DCdude17

Jan 1, 1970
0
X-No-Archive: Yes

Robert said:
Just as a side note, a 555 circuit, while simple to build, will be nearly
impossible to tune so it won't drift, particularly under different
temperature conditions.

If you need accuracy, I'd think about using a watch crystal, or even the
60Hz powerline (which is by some accounts kept accurate by an atomic clock.)
Watch crystals are very cheap, and using a 4049 to drive it is a
cheap/simple way to get an accurate timebase. You would have to divide it
down, though, so your part count will go up.

Might as well rip apart a \$4.99 electromechanical clock/watch and use the motor
drive pulse.

R

#### Robert Monsen

Jan 1, 1970
0
Well, the easiest way is to buy an alarm clock for 5.99 at walgreens... but
where is the fun in that?

Another note on this, in a box of junk parts I recently bought on EBAY, I
got a Motorola H11AA1 optoisolator, which turns out to be just the thing to
recover clock from the powerline. Limit the 120VAC line current using a 15k
2W resistor (also from the junk parts), and run the result through this
puppy to get a stable, isolated 120Hz pulse...

Regards

A

#### AC/DCdude17

Jan 1, 1970
0
X-No-Archive: Yes

Robert said:
Well, the easiest way is to buy an alarm clock for 5.99 at walgreens... but
where is the fun in that?

The thrill of putting the rest of project together. Just because someone's
building a project doesn't me he should reinvent the wheels.

Replies
3
Views
726
Replies
4
Views
910
A
Replies
26
Views
2K
A
M
Replies
13
Views
1K
Phil Allison
P
E
Replies
20
Views
2K
ellendort
E