Or perhaps they were using bottled water because they love their VW's so
much.

/Patrick

On Fri, 27 Jul 2001, Leon Korkin wrote:

> Larry,
> That's scientific approach! Valuable results too.
> My question is, if trigger point is at 42K than using ordnary tap water will
> not cause that blinking light to blink, would it? Why then some people had
> it blinking with tanks full of coolant?
> Perhaps the sensor is capacitive? Being more sensitive than resistance? Just
> guessing.
> Leon
> 85 Subwagen Westy
>
> Larry Alofs wrote:
>
> > Volks,
> > I finally tracked down a coolant leak that had been bothering me for a
> > month.  Had a small crack in the pressure tank between the pressure cap
> > and the coolant level sensor.  Since I had to disassemble things anyway,
> > I decided to take the coolant level sensor and attempt the definitive
> > study of the resistance of various coolant mixtures and then see how
> > these values compare to the resistance needed to trigger that &%#*
> > blinking light.
> >   The coolant level sensor was placed in each liquid sample about 1/2 in
> > deep in a clean plastic cup.  Initially as removed from the van the
> > probes looked mostly clean and shiny with a few particles of dirt
> > covering less than 20% of the surface.  After the first measurement, the
> > probes were rinsed between measurements.
> >   Note that the resistance of liquids generally cannot be measured
> > accurately with a typical ohmmeter using a DC test current because
> > chemical reactions occur which quickly coat the electrodes and change
> > the readings drastically.  These measurements were made by using 14 V AC
> > applied to a series circuit consisting of the coolant probe and an
> > adjustable resistance as a voltage divider.  When the voltage at the
> > junction was half of the total voltage, I assumed that the adjustable
> > resistor had a value matching the resistance of the liquid sample
> > between probes of that size and shape.  For each sample, I then
> > disconnected the circuit and measured the resistance of the adjustable
> > resistor using a DVM.
> >   The coolant from the van only tested to be good to about -10 F with a
> > hydrometer,. so it was not the recommended 50/50 mixture.
> >
> > 1.  coolant from van, probes "dirty"------------ 333 ohms
> > 2.  coolant from van, probes wiped off--------- 356 ohms
> > 3.  same coolant, probes sanded lightly, rinsed, wiped--- 365 ohms
> >
> > 4. 100% Prestone extended life (silicate & phosphate free) --- 1667 ohms
> >
> > 5. 50-50 mixture Prestone & distilled water from grocery store--- 387
> > ohms
> > 6. Tap water (Chicago)------------------------------------------3603
> > ohms
> > 7.  50-50 tap water & Prestone ---------------------------------- 384
> > ohms
> > 8.  distilled water, two different brands, same reading
> > ------------47000 ohms
> >
> > Since I wanted to investigate the statements I had read about the wrong
> > mixture triggering the coolant level warning, I wanted to see how high
> > the resistance had to be to start the light flashing.  Got the digiwife
> > to sit up front to report on the flashing and reset periodically by
> > turning off the ignition.  This test was done with the engine not
> > running.  I adjusted various potentiometers which I had attached to the
> > connector for the level sensor probe until I found that the circuit
> > triggered at a resistance of about  42000 ohms.
> >   This circuitry in my van has been acting correctly for the past month
> > or so, although it has had bouts of random unexplained flashing over the
> > past year (hence the motivation to gather more info).
> >
> > My conclusion is that most any mixture of water and antifreeze should be
> > sufficient to keep the light from flashing, assuming the rest of the
> > bizarre circuitry is working.
> >
> > Whew.  Sorry this is so long.
> >
> > Larry A.
> > 91 GL (no coolant leaks and no flashing...today)
>