Maker Pro
Maker Pro

Long time delay

M

M.Joshi

Jan 1, 1970
0
Hello,

I'm not sure if this is the correct place to post this query but I a
hoping someone may be able to help.

I am looking for an accurate way to create a 'one-shot' dela
adjustable between 30 - 45 minutes.

People have suggested using a PIC. Is this easy to implement? Ar
there dedicated I.C's that will accurately do the job?

Thanks
 
J

James Sweet

Jan 1, 1970
0
M.Joshi said:
Hello,

I'm not sure if this is the correct place to post this query but I am
hoping someone may be able to help.

I am looking for an accurate way to create a 'one-shot' delay
adjustable between 30 - 45 minutes.

People have suggested using a PIC. Is this easy to implement? Are
there dedicated I.C's that will accurately do the job?

Thanks.


You certainly could do it with a PIC, AVR or any other microcontroller,
which I would suggest if you're already proficient using them. An easier
solution would probably be something like the 74HC4060 IC which contains an
onboard oscillator and a ripple counter, they're often used to get long
delays. If you go to www.ti.com you can look it up and see the datasheet and
you can even get a few free samples of the SN74HC4060NE4 they offer if
you'd like to play around with one.
 
D

Dave Plowman (News)

Jan 1, 1970
0
I'm not sure if this is the correct place to post this query but I am
hoping someone may be able to help.
I am looking for an accurate way to create a 'one-shot' delay
adjustable between 30 - 45 minutes.
People have suggested using a PIC. Is this easy to implement? Are
there dedicated I.C's that will accurately do the job?

A 555 connected for monostable start on power up will do this.
Do a google on NE555 to find a site with the spec sheet which will give
details of how.
 
J

James Sweet

Jan 1, 1970
0
Dave Plowman (News) said:
A 555 connected for monostable start on power up will do this.
Do a google on NE555 to find a site with the spec sheet which will give
details of how.

A 555 is good for delays of a few seconds to a few minutes, but to get a 45
minute delay will require a huge timing capacitor and the time delay won't
be very accurate at all. You really need to use a counter to scale down a
higher frequency for such long delays.
 
J

Jim Adney

Jan 1, 1970
0
You certainly could do it with a PIC, AVR or any other microcontroller,
which I would suggest if you're already proficient using them. An easier
solution would probably be something like the 74HC4060 IC which contains an
onboard oscillator and a ripple counter, they're often used to get long
delays. If you go to www.ti.com you can look it up and see the datasheet and
you can even get a few free samples of the SN74HC4060NE4 they offer if
you'd like to play around with one.

You can also buy digitally set time delay relays which you can program
by little switches on their tops. These cost $50-70, give you a DPDT
contact set and can be set to delay anywhere from seconds to hours.

Google on "time delay relay"

-
 
A

Arfa Daily

Jan 1, 1970
0
James Sweet said:
A 555 is good for delays of a few seconds to a few minutes, but to get a
45
minute delay will require a huge timing capacitor and the time delay won't
be very accurate at all. You really need to use a counter to scale down a
higher frequency for such long delays.

This is not strictly speaking, a problem - if you use the CMOS version of
the 555, from Harris, for instance. This is capable of producing delays from
" microseconds to hours ", quoting from the Maplin data, for instance. The
only reason that the bipolar 555 had trouble with large values of caps, was
the typical leakage of these exceeding the trigger / threshold pins' input
currents, when large value timing resistors were used to go with the large
value caps. This limited you to a meg or so and a few tens of microfarads if
you wanted a stable predictable timer. This is not the case with the CMOS
version, where the input currents are in pico amps, so you can use large
value resistors and caps to get much longer delays. In terms of sheer
simplicity, I'm with Dave on this one, although your clock and divider
solution is an elegant way of achieving the same result, and one that I've
used in the past ;-)

Arfa
 
R

Ralph Mowery

Jan 1, 1970
0
A 555 is good for delays of a few seconds to a few minutes, but to get a
This is not strictly speaking, a problem - if you use the CMOS version of
the 555, from Harris, for instance. This is capable of producing delays from
" microseconds to hours ", quoting from the Maplin data, for instance. The
only reason that the bipolar 555 had trouble with large values of caps, was
the typical leakage of these exceeding the trigger / threshold pins' input
currents, when large value timing resistors were used to go with the large
value caps. This limited you to a meg or so and a few tens of microfarads if
you wanted a stable predictable timer. This is not the case with the CMOS
version, where the input currents are in pico amps, so you can use large
value resistors and caps to get much longer delays. In terms of sheer
simplicity, I'm with Dave on this one, although your clock and divider
solution is an elegant way of achieving the same result, and one that I've
used in the past ;-)

Another problem with long time delays is how to calibrate the delay. If you
do use the 555 or similar version then it could take a week to set the time.
Remember it is 30 minuits between checks. The beter way is to use some kind
of devider chain so you can set the timer up. He may want to use the 555
and a 100:1 devider. I have not checked to see what is out on the market
now but there should be lots of chips that will devide by almost anything.
Probably running an oscillator very fast and deviding it down will be more
accurate and maybe easier to set.
If the device is operated off AC then deviding the 60 hz will be a good
place to start.
 
A

Arfa Daily

Jan 1, 1970
0
Ralph Mowery said:
Another problem with long time delays is how to calibrate the delay. If
you
do use the 555 or similar version then it could take a week to set the
time.
Remember it is 30 minuits between checks. The beter way is to use some
kind
of devider chain so you can set the timer up. He may want to use the 555
and a 100:1 devider. I have not checked to see what is out on the market
now but there should be lots of chips that will devide by almost anything.
Probably running an oscillator very fast and deviding it down will be more
accurate and maybe easier to set.
If the device is operated off AC then deviding the 60 hz will be a good
place to start.
Provided you use a good quality low leakage cap, and high stability high
value resistors, the delay time should be predictable with reasonable
accuracy, using the standard 555 timing formula. Shouldn't take more than 3
or 4 runs to get the time delay down to what you need within a few seconds.
It seemed to me that the OP needed a 'simple' solution, which the 555
circuit is, requiring only one resistor and one capacitor besides the device
itself ( the CMOS version doesn't need either the supply decoupling against
crowbarring, or the control pin decoupling ).

This is in contrast to the clock and divider solution, which I accept is
totally predictable, but requires a divide ratio of 162000 ie 18 bits to
arrive at 45 minutes,even using a low speed clock such as the 60Hz line
supply ( which will also need clamping and shaping to form a suitable
clock ). Also, some ANDing will need doing on the output of the divider to
get an output at the required time delay. Not as simple as the timer chip
solution, properly implemented.

Arfa
 
D

Dave Plowman (News)

Jan 1, 1970
0
A 555 is good for delays of a few seconds to a few minutes, but to get a
45 minute delay will require a huge timing capacitor and the time delay
won't be very accurate at all. You really need to use a counter to scale
down a higher frequency for such long delays.

You're right of course. I wonder if the U6047 is still available. It was
designed for things like car heated rear window timing and can be set from
a few seconds to many hours with standard components - and drive a 20 amp
relay coil directly with a 12 volt rail. It also gives a choice of simple
triggering methods. Luckily, I've got a small stock.

If there's a modern equivalent I'd be pleased to know of it.
 
J

James Sweet

Jan 1, 1970
0
You're right of course. I wonder if the U6047 is still available. It was
designed for things like car heated rear window timing and can be set from
a few seconds to many hours with standard components - and drive a 20 amp
relay coil directly with a 12 volt rail. It also gives a choice of simple
triggering methods. Luckily, I've got a small stock.

If there's a modern equivalent I'd be pleased to know of it.


There's lots of similar chips out there, you reminded me too that yet
another possible solution would be to use a car window defroster timer, many
are built right into the buttons and can be had for a few bucks at a U-pull
yard.
 
A

Asimov

Jan 1, 1970
0
"Arfa Daily" bravely wrote to "All" (05 Sep 05 12:53:28)
--- on the heady topic of "Re: Long time delay"

AD> From: "Arfa Daily" <[email protected]>
AD> Xref: core-easynews sci.electronics.repair:341392

AD> Provided you use a good quality low leakage cap, and high stability
AD> high value resistors, the delay time should be predictable with
AD> reasonable accuracy, using the standard 555 timing formula. Shouldn't
AD> take more than 3 or 4 runs to get the time delay down to what you need
AD> within a few seconds. It seemed to me that the OP needed a 'simple'
AD> solution, which the 555 circuit is, requiring only one resistor and
AD> one capacitor besides the device itself ( the CMOS version doesn't
AD> need either the supply decoupling against crowbarring, or the control
AD> pin decoupling ).
AD> This is in contrast to the clock and divider solution, which I accept
AD> is totally predictable, but requires a divide ratio of 162000 ie 18
AD> bits to arrive at 45 minutes,even using a low speed clock such as the
AD> 60Hz line supply ( which will also need clamping and shaping to form a
AD> suitable clock ). Also, some ANDing will need doing on the output of
AD> the divider to get an output at the required time delay. Not as simple
AD> as the timer chip solution, properly implemented.

AD> Arfa


The cmos version of the "555" timer is the "7555" and an excellent
choice for stable delays. I've used one to get a very stable 42
minutes delay using a poly cap and high value resistor for the timing
elements. Only problem with the "7555" is the output can't drive as
much current.

The 4060 is a nice binary divider chain with 14 stages but it isn't
the only one. The 4045 has 21 stages for example. However, the nicer
device is the 4536 which has 24 stages and is programmable. In one
mode is accepts BCD inputs to control 14 stages. For example, this
goes nicely with decade trimmer switches which output BCD directly.

A*s*i*m*o*v

.... That was a fascinating period of time for electronics
 
Top