Date:         Fri, 14 May 1999 12:35:38 -0700
Reply-To:     Mark Drillock <drillock@EARTHLINK.NET>
Sender:       Vanagon Mailing List <vanagon@gerry.vanagon.com>
From:         Mark Drillock <drillock@EARTHLINK.NET>
Subject:      Re: Source for water tank level sensor
Comments: To: David Beierl <synergx@IBM.NET>
Content-Type: text/plain; charset=us-ascii

As far as I know, the circuits for the water tank levels are much the
same, but there are at least 3 different LED panels out there. They are
not functionally interchangeable as to the LED for the fridge pilot
light or LACK there of. Also one variation has the power to the faucet
pump controlled by a transistor so that the pump will not run if the
tank is empty. This one has a second red LED below the 3 water level
LEDs and when this one is lit, the transistor is switched off to disable
the pump. The pilot LED is in the center on this panel and the one I had
was red. It would be nice if someone worked up a schematic for these
panels. I have found bad solder joints and weak amplifier ICs to be
common causes of intermittent LED functioning. My panel is acting up
right now but I have too many other projects in the hopper.

Mark Drillock

David Beierl wrote:
>
> > > move the float all the way down to within one inch of the bottom,
> > > then I get another click and the Ohm meter goes to 350K (nearly
> > > infinity - open circuit).  If that is of any help to anybody
>
> In this context, 350k is far from an open circuit.  The sender is fed 12v
> through a one-megohm resistor, at least on the earlier arrangement -- and
> the evidence to date indicates that they didn't change the LED panel for
> the new sender.  The LED panel senses voltage and has an input impedance
> that could easily be ten megohms.  In stock form it should light green if
> it sees btw 0 and 3 volts, Yellow from 3 to 5, Red from 5 to 7, and no
> lights above 7 volts.  Again, I'm speaking of the '84 version and we have
> not yet established for certain that the later ones are identical.  Occam
> thinks they are, though.
>
> david
> David Beierl - dbeierl@ibm.net