Re: Damned Projects!

Don <g@crcomp.net> wrote:

Liz Tuddenham wrote:

Don wrote:
>

Electrolysis can cause the DC sensor to degrade over time. An analog
adaptation from resistance to capacitance can curb this effect. The
above sensor needs to be altered to detect a dielectric difference
instead of an ohmic resistance.

>
Perhaps spme sort of adaptation of my conductivity meter circuit would
do the job:
<http://www.poppyrecords.co.uk/other/ConductivityMeter14.gif>
>
It generates floating AC with prevents electrolysis at the sensor and
uses synchronous rectification to eliminate spurious responses.  For
simplicity, the range switch could be replaced by a pot and the meter
drivers arranged to switch a relay.

 
Do the five Schmidt triggers in parallel after the oscillator act as a
buffer?

In this pareticular application, the source impedance of the signal is
effectively in series with the resistance of the conductivity cell, so
it needs to be as low as possible or it will affect the linearity of the
highest conductivity range.  This resistance is made 0.5 ohms by the
bottom resistor of the voltage divider which drives the cell.  (Two
1-ohm resistors in parallel)

That means that the top resistor of the divider needs to be in the
region of 750 ohms to get a reasonable signal.  Such a low impedance
across the output of a CD40106 could cause current limiting which would
give unpredicatable output levels, so several sections in parallel
ensure that there is no overloading.  All the sections must be driven
identically and be in the same chip, so that there is no significant
time difference to give rise to a moment when one output switches before
another and a huge current spike is drawn from the supply.

Besides, what ekse could you do with all those unused gates?


--
~ Liz Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk