>But - i did not mention - the fluid is cooling water, so it is warm if
>runs and the temperature even changes according to dissipation of the
>laser gun to be cooled. It makes the termal approach a bit complicated I
>think.

The heat transport sensor works like this:


==================== <-- main pipe wall

  +----------------+ <-- sensor shield (eg thin PVC or PTFE tube)
  |  S1    H    S2 |
  +----------------+

====================

H is a heater that runs all the time (resistor or NTC at ~80C). S1, S2 are
thermistors and connected to a differential amp followed by a window
comparator whith 3 outputs (above, equal, below). These outputs indicate
flow forward/no flow/flow backward. A circuit verifies that the heater
current is ok (not zero and not shorted). This is because loss of the
heater will cause the output to misread. The heater current can be
monitored by another window comparator to provide a 'signal good' readout.
Output is independent of the liquid temperature but sensitivity can
depend on flow speed.

Peter

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.