# How to achieve a pulse once every 30 seconds?

#### OrangeKat

Joined Feb 13, 2023
5
I have a rather precious (sentimental value) slave clock that was originally driven by a master clock that provided a pulse once every 30 seconds precisely. The slave is powered by 12vdc and I believe the pulse duration to be about one second.
I would like to make a timing circuit to drive the slave as I do not have the original master clock. Being more mechanically minded rather than electronic, any help in providing me with a working circuit diagram would be most appreciated. Do not assume my knowledge of electronics is much beyond that of high school grade of many years ago!

#### Ian0

Joined Aug 7, 2020
10,054
32.768kHz crystal oscillator driving a 4060, will give 2 pulses every second. Connect the output to a 40103 set to divide by 60, and you will get a half-second pulse every 30 seconds.

#### Ya’akov

Joined Jan 27, 2019
9,237
Welcome to AAC.

@Ian0’s approach is simple and straightforward with just a few components but may be a problem if the .5 second pulse is insufficient.

You mention that the clock is powered by 12VDC, is that also the voltage required for the pulse?

In any case, I would favor using a microcontroller for this because they are cheap, easy to use, and the fact that you can adjust the duration of the pulse to whatever you ultimately need purely by revising software is very attractive.

Another interesting possibility is achieving a very high accuracy by using an inexpensive GPS module’s 1 pulse per second output to trigger the pulse at very precise intervals making the clock virtually perfectly accurate.

The MCU (MicroController Unit) approach has a steepish learning curve but depending on your inclinations no more steep and possibly far less so than the bare IC approach (if the first circuit needs revision in particular).

This could be done with an Arduino compatible device for a few dollars and an investment of a few hours of learning that might provoke further interest in MCU projects which are really great fun.

Is there a remote setting facility for the clock, the way a lot of master clock systems worked? If so, that could be easily done with the MCU and and RTC (Real Time Clock) module.

#### crutschow

Joined Mar 14, 2008
34,697
@Ian0’s approach is simple and straightforward with just a few components but may be a problem if the .5 second pulse is insufficient.
I doubt that a 0.5s pulse would not be long enough to energize the solenoid that pulses the clock mechanism.

#### Ian0

Joined Aug 7, 2020
10,054
It‘s just that the 4060 is annoyingly one stage short to produce 1Hz from a 32768Hz crystal.
If it really needs a longer pulse, then adding a 4013 to divide by two, and setting the 40103 to divide by 30 will fix it, or, adding a mono stable to lengthen the pulse.

#### crutschow

Joined Mar 14, 2008
34,697
Last edited:

#### Ya’akov

Joined Jan 27, 2019
9,237
I doubt that a 0.5s pulse would not be long enough to energize the solenoid that pulses the clock mechanism.
I have no idea what the mechanism that regulates the clock looks like and the TS suggested it was about 1s long, so...

#### Jerry-Hat-Trick

Joined Aug 31, 2022
583
In the UK, I'd buy a quartz clock movement from Amazon (or elsewhere) which runs on a single AA battery. Make sure it comes with a 'second' hand, throw away the minute and hour hands and maybe broaden the width of the second hand at both ends, making both ends the same length. Glue an Opto Interrupter to the housing in such a way that the second hand passes through it every 30 seconds (i.e, each end once a minute). Simple circuit to convert the opto interrupter to a 12V pulse - duration determined by the width of the second hand as modified.

Hope this makes sense - if it's a method you'd consider I could sketch a schematic.

#### crutschow

Joined Mar 14, 2008
34,697
I have no idea what the mechanism that regulates the clock looks like and the TS suggested it was about 1s long, so...
They have a solenoid that mechanically moves the clock mechanism 30s per pulse
The TS made an estimate of the pulse time, which I would think has significant tolerance.
I shouldn't take more than a few hundred ms to activate the solenoid.

Last edited:

#### djsfantasi

Joined Apr 11, 2010
9,183
I doubt that a 0.5s pulse would not be long enough to energize the solenoid that pulses the clock mechanism.
@Ian0 ’s solution also suggested using a monostable to increase the pulse width.
or, adding a mono stable to lengthen the pulse.
Or reverse his components, divide by 30 with the 4013 first then divide by two with the 4013. This gives a 15 second pulse width. Ir something similar… my idea is to use the flip-flip as the last operation in the chain.

#### Ian0

Joined Aug 7, 2020
10,054
@Ian0 ’s solution also suggested using a monostable to increase the pulse width.

Or reverse his components, divide by 30 with the 4013 first then divide by two with the 4013. This gives a 15 second pulse width. Ir something similar… my idea is to use the flip-flip as the last operation in the chain.
The crucial question is how long the pulse width needs to be. I suspect a 15 second pulse will be too long.

#### crutschow

Joined Mar 14, 2008
34,697
The crucial question is how long the pulse width needs to be. I suspect a 15 second pulse will be too long.
Yes, that could overheat the solenoid.

#### Papabravo

Joined Feb 24, 2006
21,264

#### Ian0

Joined Aug 7, 2020
10,054
Using an edge-triggered monostable and a preset would allow the pulse width to be adjusted almost down to zero, and up to a maximum.

#### crutschow

Joined Mar 14, 2008
34,697
Here's from an impulse clock driver description:
EC4A impulse driver for slave clocks - mini master clock. ... Variable Pulse width (62ms,120ms,240ms),
so apparently the minimum required width varies with the clock manufacturer/type, and doesn't exceed 240ms.

Last edited:

#### OrangeKat

Joined Feb 13, 2023
5
32.768kHz crystal oscillator driving a 4060, will give 2 pulses every second. Connect the output to a 40103 set to divide by 60, and you will get a half-second pulse every 30 seconds.
Thank you IanO, I had started to read up on 4060 but failed to understand how I could achieve the required pulse frequency. I will now take a look at 40103. Looking forward to learning some new skills.

#### OrangeKat

Joined Feb 13, 2023
5
Welcome to AAC.

@Ian0’s approach is simple and straightforward with just a few components but may be a problem if the .5 second pulse is insufficient.

You mention that the clock is powered by 12VDC, is that also the voltage required for the pulse?

In any case, I would favor using a microcontroller for this because they are cheap, easy to use, and the fact that you can adjust the duration of the pulse to whatever you ultimately need purely by revising software is very attractive.

Another interesting possibility is achieving a very high accuracy by using an inexpensive GPS module’s 1 pulse per second output to trigger the pulse at very precise intervals making the clock virtually perfectly accurate.

The MCU (MicroController Unit) approach has a steepish learning curve but depending on your inclinations no more steep and possibly far less so than the bare IC approach (if the first circuit needs revision in particular).

This could be done with an Arduino compatible device for a few dollars and an investment of a few hours of learning that might provoke further interest in MCU projects which are really great fun.

Is there a remote setting facility for the clock, the way a lot of master clock systems worked? If so, that could be easily done with the MCU and and RTC (Real Time Clock) module.
Thank you Ya’akov, you have provided me with a lot of very helpful information which I look forward to further study. Your approach regarding the use of MCU is very tempting however my knowledge in this area is very limited but something I could well start to understand given enough time. I’ve read the replies following your first comment and feel that I have plenty of helpful suggestions to start making my first electronic project. To answer your question, I was hoping to use a 12v gel cell battery to power the circuit and the slave clock solenoid. I have tested the solenoid by just a very quick touch on the battery and the mechanism indexed perfectly, hence my thought it might require no more than 1 second duration. There is no remote setting as this clock was manufactured probably in the early 1960’s.

#### crutschow

Joined Mar 14, 2008
34,697
I had started to read up on 4060 but failed to understand how I could achieve the required pulse frequency
The 4060 is a 14-stage binary counter so it divides the input (or oscillator) frequency by a factor of 2^14 or 16,384, thus the 32,768Hz common watch crystal will give an output of 2 pulses/sec.
This is then further divided by 60 using the 40103 to give the desired 2 pulses/min.

Last edited:

#### Ya’akov

Joined Jan 27, 2019
9,237
Thank you Ya’akov, you have provided me with a lot of very helpful information which I look forward to further study. Your approach regarding the use of MCU is very tempting however my knowledge in this area is very limited but something I could well start to understand given enough time. I’ve read the replies following your first comment and feel that I have plenty of helpful suggestions to start making my first electronic project. To answer your question, I was hoping to use a 12v gel cell battery to power the circuit and the slave clock solenoid. I have tested the solenoid by just a very quick touch on the battery and the mechanism indexed perfectly, hence my thought it might require no more than 1 second duration. There is no remote setting as this clock was manufactured probably in the early 1960’s.
What is the make and model of the clock? I attended NYC public schools in the 1960s and we had Simplex clocks with master and slaves. They could do top of the hour correction if given a signal by the master. The procedure involved a third wire for some slaves and a long sync signal for others.

It would be a real pain in a large master clock system the slaves had to be manually set for DST or blackouts—not to say that couldn’t be the case, thpugh.

#### Sensacell

Joined Jun 19, 2012
3,475