Need a little help building a 12 volt timer

[Ryan Ashline]

I am looking for a 12 volt timer or looking to build one. What I
would like it to do is when powered wait for approx 3 mins (doesnt
have to be close 15 sec or so +- would be ok) Then trigger a light I
have for 2-3 secs shut off then repeat the entire sequence again. 3
mins off 3 secs on. Would some please point me in the right
direction?

Thanks for your time

Ryan Ashline

 

[mike]

I am looking for a 12 volt timer or looking to build one. What I
would like it to do is when powered wait for approx 3 mins (doesnt
have to be close 15 sec or so +- would be ok) Then trigger a light I
have for 2-3 secs shut off then repeat the entire sequence again. 3
mins off 3 secs on. Would some please point me in the right
direction?

Thanks for your time

Ryan Ashline

Dig up a 555 datasheet and build the astable.
mike

 

[Rich Grise]

I am looking for a 12 volt timer or looking to build one. What I
would like it to do is when powered wait for approx 3 mins (doesnt
have to be close 15 sec or so +- would be ok) Then trigger a light I
have for 2-3 secs shut off then repeat the entire sequence again. 3
mins off 3 secs on. Would some please point me in the right
direction?

Thanks for your time

http://www.national.com/pf/LM/LMC555.html

There's a circuit in there, and a chart of RC values, from there it's
just a little arithmetic. I'm not sure what the best caps are these
days - it used to be polyester, but I've been out of the loop for
awhile - look for "low leakage".

Failing that, look for the 4020/4040/4060 - the 4060 has its own
internal oscillator. You can get an arbitrary time delay with one
of these and a little bit of logic.

Have Fun!
Rich

 

[Joerg]

Hi Rich,

Failing that, look for the 4020/4040/4060 - the 4060 has its own
internal oscillator. You can get an arbitrary time delay with one
of these and a little bit of logic.

Excellent advice. When stuff goes past the one minute marker these 4000
logic choices are clearly better than some RC combination on a one-shot.

Regards, Joerg

 

[CFoley1064]

Subject: Re: Need a little help building a 12 volt timer

From: Joerg [email protected]
Date: 10/11/2004 3:23 PM Central Daylight Time
Message-id: <[email protected]>

Hi Rich,

Excellent advice. When stuff goes past the one minute marker these 4000
logic choices are clearly better than some RC combination on a one-shot.

Regards, Joerg

Joerg speaks truth -- the longer the time delay, the more chance of something
coming around to mess timing up like electrical noise from the relay you're
driving, power supply hash, you name it.

A lot of time delay relays built in the 1970s and 1980s were based on the 4060,
mostly because if noise from the relay caused the oscillator to add a count,
well, there are always 16383 other counts to keep timing stable. These relays
were never specified at better than 1% repetitive anyway.

Here's something that should work for you (view in fixed font or M$ Notepad):


VCCVCC
+ +
| |10K VCC VCC VCC VCC
.-..-. + + + +
VCC 270K| || | | | | |
+ | || |.----o-----o----. | |
| '-''-'| 8 4 | | C|
.-------o---------. | | | | 1N4002- C|RY1
| 16 | | | | | ^ C|
| |.001uF | | | | | |
| | | | | | | |
| | || | | | | | |
| 3o---||--|--o-o2 3o------. '---o
| 4060 | || | | | | |
.-o8 | | | 555 | | |
| | | o----o6 | | |
| | | | | | .-. |
| | | | | | 1K| | |
| | 11 10 9 | o----o7 | | | |
| '-o------o------o-' | | | '-' |
| | | | 10uF| | | | |/
| .-. .-. | --- | | o-----|2N3904
| | | | | --- --- | 1 5 | | |>
| | | 50K| | --- | '----o-----o----' .-. |
| '-' '-' |0.1uF | | 1K| | |
| |220K | | === === | | |
| | | | GND GND '-' |
| '------o------' | |
| | |
=== === ===
GND GND GND
created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de


You're using the built-in oscillator on the 4060 to generate a 91 Hz clock with
the .1uF cap, the 50K resistor, and the 220K resistor. That gets rolled around
in the divide-by-2^14 ripple counter in the 4060, coming out as 1/180 Hz (do
the math). That means you get a NGT to trigger the 555 through the .001uF cap
with the 10K pullup once every 3 minutes. The values of the 555 are set for a
3 second pulse, which drives the relay through the transistor. Try to find a
relay with a coil that needs 100mA or less.

A few notes here. For the 50K resistor in the 4060, you might want to use a
39K resistor and a 25K pot. For the 270K resistor on the 555, you might want
to use a 220K resistor and a 100K pot. Use a CMOS 555 if you can -- leakage
currents on a regular 555 might cause you problems. But the main caveat is,
don't use electrolytics for either cap, especially the 555. Use a tantalum if
you can -- it will generally have lower leakage current. Also, keep your
circuit board clean and out of high humidity environments.

I hope this has been of use. Questions of this type generally receive a better
response on sci.electronics.basics.

Good luck with your project
Chris

 

[Tam/WB2TT]

I am looking for a 12 volt timer or looking to build one. What I
would like it to do is when powered wait for approx 3 mins (doesnt
have to be close 15 sec or so +- would be ok) Then trigger a light I
have for 2-3 secs shut off then repeat the entire sequence again. 3
mins off 3 secs on. Would some please point me in the right
direction?

Thanks for your time

Ryan Ashline

Ryan
If you are AC powered, and don't have to be exact, just take the power line
frequency and pass it through a 12 bit binary counter. You will get a cycle
time of 68 seconds at 60 Hz, and 81 seconds at 50 Hz. Divide that by 2 or 3
to get a roughly 3 minute trigger.

Tam

 

[Rich Grise]

Thank you.

Joerg speaks truth -- the longer the time delay, the more chance of
something coming around to mess timing up like electrical noise from the
relay you're driving, power supply hash, you name it.

A lot of time delay relays built in the 1970s and 1980s were based on the
4060, mostly because if noise from the relay caused the oscillator to add
a count,
well, there are always 16383 other counts to keep timing stable. These
relays were never specified at better than 1% repetitive anyway.

Here's something that should work for you (view in fixed font or M$
Notepad):

What???!!!!???? A 555????? We don' need no steeenkeeeng 555!

VCC VCC VCC VCC
+ + + +
VCC | | | |
+ .----o-----o----. | |
| | 12 7 | | C|
.-------o---------. | | 1N4002- C|RY1
| 16 | | | ^ C|
| 4o------------o2 | | |
| 5o------------o3 | | |
| 6o------------o4 | | |
| 7o------------o5 1o------. '---o
| 4060 13o------------o9 | | |
.-o8 14o------------o10 | | |
| | 15o------------o11 | | |
| | | note: | | .-. |
| | | these pins | 4068 | 1K| | |
| | 11 10 9 | are | | | | |
| '-o------o------o-' inter- | | '-' |
| | | | change- | | | |/
| .-. .-. | able. | | o-----|2N3904
| | | | | --- | 14 | | |>
| | | * | | --- '----o----------' .-. |
| '-' '-' | * | 1K| | |
| | * | | === | | |
| | | | GND '-' |
| '------o------' | |
| | |
=== === ===
GND GND GND
created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

Modified by Rich Grise v0.999.69 Alpha 3 ;-)

* component values for proper timing are left as an exercise for the
reader. (I didn't even check yours, and am too lazy to figure it out
for whatever freq you'd need for the above)

Cheers!
Rich

 

[Terry Pinnell]

Excellent advice. When stuff goes past the one minute marker these 4000
logic choices are clearly better than some RC combination on a one-shot.

Regards, Joerg

Not for the application and precision requested by the OP. Quite
unnecessary. The 555 itself is capable of 1% accuracy (2 secs in 3
mins). With a modern low-leakage electrolytic as timing capacitor, and
reasonably careful construction, you could expect close to that even
for monostable or astable periods of the order of an hour.

 

[Terry Pinnell]

I am looking for a 12 volt timer or looking to build one. What I
would like it to do is when powered wait for approx 3 mins (doesnt
have to be close 15 sec or so +- would be ok) Then trigger a light I
have for 2-3 secs shut off then repeat the entire sequence again. 3
mins off 3 secs on. Would some please point me in the right
direction?

Thanks for your time

Ryan Ashline

After switching on power, is it important that the delay before the
first 3 sec ON should be (approximately) 3 minutes?

If NO, a simple astable (oscillator) could be used, with a roughly
1:60 duty cycle (3 mins OFF, 3 secs ON).

If YES, a monostable (one-shot or timer) needs to be added. This is
because a simple astable, such as from a 555 or 1/2 x 4001 or 4011,
starts at an unpredictable position in its cycle. So the delay could
be *anything* up to 3 mins. As an added complication, if you used a
555 the delay could be over 4 mins, because the *first* cycle of a 555
astable is always 30-40% longer than all subsequent cycles.

You also need to be clear we understand your skill level. Given a
suitable astable or mono+astable circuit, can you add appropriate
lamp-driving circuitry yourself? If not, you need to specify more
details of the lamps and your power supply voltage and current.

 

[Tam/WB2TT]

After switching on power, is it important that the delay before the
first 3 sec ON should be (approximately) 3 minutes?

If NO, a simple astable (oscillator) could be used, with a roughly
1:60 duty cycle (3 mins OFF, 3 secs ON).

If YES, a monostable (one-shot or timer) needs to be added. This is
because a simple astable, such as from a 555 or 1/2 x 4001 or 4011,
starts at an unpredictable position in its cycle. So the delay could
be *anything* up to 3 mins. As an added complication, if you used a
555 the delay could be over 4 mins, because the *first* cycle of a 555
astable is always 30-40% longer than all subsequent cycles.

You also need to be clear we understand your skill level. Given a
suitable astable or mono+astable circuit, can you add appropriate
lamp-driving circuitry yourself? If not, you need to specify more
details of the lamps and your power supply voltage and current.

How about a compromise desig,:
1. Build an oscillator with a period of 3 seconds. Easily doable with
standard component. ( 1 ufd mylar cap + few megaohm)
2. Feed the output to a 6 bit counter.
3. Connect the reset lead of the counter to an RC network so it powers up
000000.
4. Detect the 111111 state to turn on the light.
5. This gives 192 seconds OFF, 3 seconds ON.

Tam

 

[Joerg]

Hi Rich,

What???!!!!???? A 555????? We don' need no steeenkeeeng 555!

To be honest, I have never really used one in a design for a customer
and if I ever did it would only be the CMOS version. Whenever I needed
timing I used CD chips, usually Schmitts. If you take a hex chip that
gives you a timing capability for a silicon cost share of under 2 cents.

Long timers are usually better done using dividers. After all, the cost
impact of a polyester cap or larger varieties in other technologies is
nothing to be sneezed at.

Regards, Joerg

 

[Joerg]

Hi Terry,

Not for the application and precision requested by the OP. Quite
unnecessary. The 555 itself is capable of 1% accuracy (2 secs in 3
mins). With a modern low-leakage electrolytic as timing capacitor, and
reasonably careful construction, you could expect close to that even
for monostable or astable periods of the order of an hour.

For a one-up project and in a home environment with fairly constant
temperature and humidity, maybe. But when you look at the cost of low
leakage electrolytics the picture can change for a commercial product.
Also, ideally you do not want trim pots and I believe it is nearly
impossible to obtain an electrolytic under 5% tolerance at reasonable
cost. Even with polyester that is a challenge as they often cost more
than two resonators would. I guess that's why someone came up with the 4060.

Back in my days at the university one of our professors used to joke
that the quality of a digital designer is inversely proportional to the
number of one-shots he or she uses. Now I don't quite agree with that
and I use them a lot myself, even though they will be "home made" rather
than a 555 or HC123. It's just that the cost picture changes with longer
timings and cost is usually the predominant factor in a commercial design.

Regards, Joerg

 

[Spehro Pefhany]

Hi Terry,

For a one-up project and in a home environment with fairly constant
temperature and humidity, maybe. But when you look at the cost of low
leakage electrolytics the picture can change for a commercial product.
Also, ideally you do not want trim pots and I believe it is nearly
impossible to obtain an electrolytic under 5% tolerance at reasonable
cost. Even with polyester that is a challenge as they often cost more
than two resonators would. I guess that's why someone came up with the 4060.

The labor cost of calibrating an analog timer in the minutes can be a
factor too. For our 3 minute analog timer modules I did a circuit that
would track and then freeze the capacitor voltage in time long enough
that the calibration pot could be trimmed (turned out to be trickier
than it sounds). With digital timers you can bypass most of the
divider chain to speed up testing. Reset of the capacitor voltage
becomes a non-issue.

 

[Joerg]

Hi Spehro,

The labor cost of calibrating an analog timer in the minutes can be a
factor too. For our 3 minute analog timer modules I did a circuit that
would track and then freeze the capacitor voltage in time long enough
that the calibration pot could be trimmed (turned out to be trickier
than it sounds). With digital timers you can bypass most of the
divider chain to speed up testing. Reset of the capacitor voltage
becomes a non-issue.

That is probably the reason why most of the start-up delay relays for
motors use the 4060 or something along those lines.

Testing can often be done with a poor man's spectrum analyzer: One of
those radios from Europe that have the long wave band. Just look where
the harmonics come out. If you let an IF stage oscillate a bit it can be
even easier to detect. I guess nowadays a sound card with enough dBs of
preamp would do.

Regards, Joerg

 

[Spehro Pefhany]

Hi Spehro,

That is probably the reason why most of the start-up delay relays for
motors use the 4060 or something along those lines.

Testing can often be done with a poor man's spectrum analyzer: One of
those radios from Europe that have the long wave band. Just look where
the harmonics come out. If you let an IF stage oscillate a bit it can be
even easier to detect. I guess nowadays a sound card with enough dBs of
preamp would do.

National (IIRC) patented a method of calibrating digital watches by
optically picking up the multiplex frequency of the LED (yes, early
watches) digital display and using something like a period counter as
a display while the trimcap was being tweaked.


Best regards,
Spehro Pefhany

 

[Peter]

I am looking for a 12 volt timer or looking to build one. What I
would like it to do is when powered wait for approx 3 mins (doesnt
have to be close 15 sec or so +- would be ok) Then trigger a light I
have for 2-3 secs shut off then repeat the entire sequence again. 3
mins off 3 secs on. Would some please point me in the right
direction?

Thanks for your time

Ryan Ashline

A little program like this into a PIC would do the trick.

main: low 0
wait 180
high 0
wait 3
goto main

"wait" counts in seconds, Just add a transistor or a relay to drive the
load.

If you are in Australia I could supply the complete unit you are after,
including the programmed chip. I have about 100 of these PCB's hanging
around. 12V ac/dc in, same out, max load 5A.

Just a thought.

 

[Joerg]

Hello Spehro,

National (IIRC) patented a method of calibrating digital watches by
optically picking up the multiplex frequency of the LED (yes, early
watches) digital display and using something like a period counter as
a display while the trimcap was being tweaked.

Hey, if they were able to patent that scheme I could have patented my LW
receiver gimmick. Too late, I guess I just spilled the beans ;-) Oh, well.

Regards, Joerg

 

[Rich Grise]

How about a compromise desig,:
1. Build an oscillator with a period of 3 seconds. Easily doable with
standard component. ( 1 ufd mylar cap + few megaohm)
2. Feed the output to a 6 bit counter.
3. Connect the reset lead of the counter to an RC network so it powers up
000000.
4. Detect the 111111 state to turn on the light.
5. This gives 192 seconds OFF, 3 seconds ON.

Hey, I posted this circuit yesterday, ASCII art and all! Where did my
post go?

Thanks,
Rich

 

[Rich Grise]

How about a compromise desig,:
1. Build an oscillator with a period of 3 seconds. Easily doable with
standard component. ( 1 ufd mylar cap + few megaohm)
2. Feed the output to a 6 bit counter.
3. Connect the reset lead of the counter to an RC network so it powers up
000000.
4. Detect the 111111 state to turn on the light.
5. This gives 192 seconds OFF, 3 seconds ON.

Right here: <lxGad.3480$Rp4.2607@trnddc01>

Cheers!
Rich

 

[Joerg]

Hi Spehro,

National (IIRC) patented a method of calibrating digital watches by
optically picking up the multiplex frequency of the LED (yes, early
watches) digital display and using something like a period counter as
a display while the trimcap was being tweaked.

Just wondering: I remember when a friend in high school had one of
these. Holding it next to an AM radio would muffle all the stations and
you could hear distinct harmonics between all the pulsating hash when
turning the AM dial. They could have just picked a harmonic as well and
attach a pickup coil to the radio, plus feed in a correct frequency or
harmonics off a "golden watch" for beat. Then it's just a simple tuning
to zero, "weeeeoooupp...", done. Acoustic feedback usually works fastest
in a production environment. It does tend to go on other peoples nerves
though.

Now don't ask how I tune our piano. Any pro would get goose pimples.

Regards, Joerg