'need a very low cost timer'
John Waters wrote:
> I need a timer to turn on/off a small aquarium air pump, but cost is
> very sensitive while accuracy is not a problem, could anyone suggest me
> a cheap solution? it could be a circuit, an off-the-shelf product, or
> whatever method, as long as it is cheap! I need only one on/off
> interval - once every hour (just roughly will do), each time turn on
> for about 5 minutes. It would be perfect if the on/off interval can be
> swtichable to once every ten hours (again on for around 5 minutes) as
> well. If it is not possible to have such a long or switchable delay,
> just once on/off interval every hour will be ok.
Sounds like a perfect application for a 10F200.
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
|>> I need a timer to turn on/off a small aquarium air pump, but cost
>> very sensitive while accuracy is not a problem, could anyone
>> suggest me
>> a cheap solution?
>> ... I need only one on/off interval - once every hour (just roughly
>> will do), each time turn on
>> for about 5 minutes. [also] .. swtichable to once every ten hours
>> (again on for around 5 minutes)
> Sounds like a perfect application for a 10F200.
I'd agree with Olin. A microprocessor solution, be it PIC or some
other (eg ATtiny11 should be competitive) is liable to give you most
power and flexibility per $. But it may not give the lowest $ overall.
If you want a hardware only solution a CD4060 may do what you want.
Digikey price falls to around $US0.15 in thousands and others may be
somewhat cheaper. In similar quantities the 10F200 price falls to
about $US0.55 so the CD4060 solution may be able to be made cheaper
depending on how much flexibility is required. Other costs (PCB,
construction etc tend to make the total prices closer than IC prices
may suggest). The ATtiny11 seems to cost about half the $ of the
10F200 so it would be worth looking at.
CD4060 / 4040 / 4020 Datasheet:
The CD4060 is a 14 stage ripple counter with the input inverter
configured so it can self oscillate. This allows time delays many
tomes longer than the period of the oscillator to be configured. Only
outputs 4 to 10 and 12 to 14 of the divider chain are user accessible
but this usually allows significant flexibility. At maximum divide
ratio of 2^14 this allows clock division by 16,384 so to get 10 hour
you need an oscillator of 16384/(10 x 3600) = about 0.5 Hz. This is
probably pushing things somewhat but should be doable. For a one hour
maximum that's about 5 Hz which is better. (This is the "half period"
of the final stage output if runs as a repeating counter - the usual
method is to gate a faster output and a slower one so that the output
becomes active when both are high and then resets the counter when the
faster one again goes low. By choosing fast and slow outputs
appropriately both the overall cycle time and on delay can be semi
The 5 minute output can be achieved by a monostable, or gating or a
second 4060. Using diodes to make a "diode gate" a very low cost
arrangement can be produced, albeit requiring a bit of thinking.
If the oscillator was set to about 0.5 Hz then delays of 10 / 5 / 2.5
hours could be obtained. Slightly faster clock gives 4/2/1 hours.
If a seperate clock can be afforded (and it probably can here) then
the CD4040 counter gives more flexibility as all divider outputs are
accessible BUT it has less stages, OR a CD4020 also gives 14 stages
and access to all counter outputs except stages 2 & 3.
Cheapest most flexible solution probably requires a CD4020, CD4013
(Schmitt NAND)(if I've got the number right after all these years) and
a few R's and C's. Output to drive the pump could be FET, bipolar
transistor or relay as appropriate. Or even a triac if it's AC
With delays of this duration purely analog solutions are not viable in
a low cost production environment (and almost not in any environment).
More... (looser matching)
- Last day of these posts
- In 2005
, 2006 only
- New search...