watering system, second attempt.

[John Fields]

On Sun, 08 Aug 2004 16:29:00 +0100, andy
<news4@earthsong.free-online.co.uk> wrote:

i don't think that would work anyhow - the 4538 triggers on a rising edge
of the output of the 'or' gate between inputs A and B. This will let you
trigger on either a rising or falling edge, by choosing which input to
connect the trigger signal to and using the other to gate it with. but i
couldn't see any way of triggering on both edges using just one 4538.

---
You can do it by holding the inactive (at the moment) input at the
level it needs to be at to allow the other input to trigger. Look at
the truth table; it's all there.

Also check "Watering controller 4" on
alt.binaries.schematics.electronic for a four-chip solution that
should give you everything you want, including a two hour holdoff
timer.
_____
BTW, your 4538 RESET inputs are shown floating last time I looked at
your schematic. Not good; they need to be tied high if you're not
using them. Also, there's no such thing as a 340K pot, you don't have
the "B" sections of the comparator and the one-shot pinned out
properly, and I think you're going to run into leakage current
problems with that 1000ľF cap. Plus, how about if you knock off that
annoying little passive-aggressive trick of not using upper-case "I"'s
where you should?

--
John Fields

 

[John Fields]

I can see how to trigger on one or the other edge, but not both with the
same 4538, without using extra circuitry to make more edges or delays like
in the first design with chips I drew. As far as I can see, the 4538
triggers on a rising edge of the OR gate's output, but how can any
combination of gates on the light level signal make two rising edges where
there's only one in the signal?

---
It's doable, but you're right. You'd have to introduce extra circuitry
to remember what happened last and set up the inputs to react to
what's going to happen next. Hardly worthwhile with another entire
one-shot in the package left to do it with.
---

Also check "Watering controller 4" on alt.binaries.schematics.electronic
for a four-chip solution that should give you everything you want,
including a two hour holdoff timer.

I found the text of your post on a free news server, but not the
schematic. My ISP doesn't do binaries, so it's not easy to read this group.

---
I'll email it to you.

--
John Fields

 

[andy]

On Sun, 08 Aug 2004 11:40:00 -0500, John Fields wrote:

On Sun, 08 Aug 2004 16:29:00 +0100, andy
news4@earthsong.free-online.co.uk> wrote:


i don't think that would work anyhow - the 4538 triggers on a rising edge
of the output of the 'or' gate between inputs A and B. This will let you
trigger on either a rising or falling edge, by choosing which input to
connect the trigger signal to and using the other to gate it with. but i
couldn't see any way of triggering on both edges using just one 4538.

---
You can do it by holding the inactive (at the moment) input at the
level it needs to be at to allow the other input to trigger. Look at
the truth table; it's all there.

I can see how to trigger on one or the other edge, but not both with the
same 4538, without using extra circuitry to make more edges or delays like
in the first design with chips I drew. As far as I can see, the 4538
triggers on a rising edge of the OR gate's output, but how can any
combination of gates on the light level signal make two rising edges where
there's only one in the signal?

Also check "Watering controller 4" on alt.binaries.schematics.electronic
for a four-chip solution that should give you everything you want,
including a two hour holdoff timer.

I found the text of your post on a free news server, but not the
schematic. My ISP doesn't do binaries, so it's not easy to read this group.

BTW, your 4538 RESET inputs are shown floating last time I looked at
your schematic. Not good; they need to be tied high if you're not using
them.

Thanks. I did remember this when I actually built it.

Also, there's no such thing as a 340K pot, you don't have the "B"
sections of the comparator and the one-shot pinned out properly, and I
think you're going to run into leakage current problems with that
1000ľF cap.

That I didn't think of - thanks.

Plus, how about if you knock off that annoying little
passive-aggressive trick of not using upper-case "I"'s where you should?

That's partly just laziness/easy typing, but I know
what you mean.

--
http://www.niftybits.ukfsn.org/

remove 'n-u-l-l' to email me. html mail or attachments will go in the spam
bin unless notified with

 or [attachment] in the subject line.