On Wed, Sep 05, 2001 at 07:41:52PM +0200, SkinTech wrote:
> Well, I haven't followed the previous suggestions, but it seems to me that a
> megohm is pretty high impedance w.r.t. a human finger. When you touch the
> pin(s), you couple a lot of 50Hz (60Hz) noise into the chip, which modulates
> the internal comparators inputs all over the place. Of course it goes
> beserk!
> I suspect that in your actual app you will need to use really well screened
> cabling (how long, physically, will the resistance string be?).

Yes there are a lot of issues at that high a resistance. Also you're going to
get huge delay times, or if you use a small cap very very unstable delays.

I think you'll find that at that resistance you're very nearly out of spec.
Go and doublecheck the datasheet.

> How about this alternative: a tube with an ultrasonic 'transmitter' at the
> top, alongside a receiver. The us is pulsed, and you time how long it takes
> until the reflection from the water surface is received. Divide by two & the
> speed of sound, and you get water level.

Ouch! that's a really complicated system. It's several orders of magnitude
above the original.

Let me take stab at a simpler one: If you have clearance above the tank you
can do it with a simple float. Attach a pole that gets floated and place a
magnet on the pole. Then put a tower with magnetic reed switches at regular
intervals next to the pole. As the magnet passes each switch it'll switch.
Another possibility is to have magnets at the same intervals along the pole
so that when the float is at the bottom of the tank all of the switches are
engaged and as the float rises fewer and fewer magnets engage the remaining
switches.

You can then go right back to the 555 but with much lower resistances. If
you couple a different resistance across each switch then when that switch
engages, it'll change the frequency. Or you can simply read the digital
inputs of the switches directly.

Depending on the tank material, you may be able to implement this same kind
of scheme along the edge of the tank. Another possibility is if the tank
is transparent then the float can carry a mirror and you detect the mirror
using a LED photodector pair...

> That also takes care of the problem that your water 'resistance' is very
> much dependent on whatever  is solved into the water, which can probably
> vary unpredictable.

Very much agreed on that.

BAJ

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics