🚀 go-pugleaf

In article <6229@videovax.tv.tek.com>, bill@videovax.tv.tek.com (William K. McFadden) writes:
> >I need a device to give me an binary electrical indication that water
> >is flowing/not flowing through a standard half-inch domestic water
> >pipe.  
> 
> This is only a partial solution, but you can detect the turn-on of
> hot water by the temperature rise on the pipe.  You can detect the
> flow rate with two sensors.  This only works at startup, though.

	If one penetrates the pipe with a small point-type heating element,
and uses two temperature sensors, one slightly downstream of the other,
one can measure flow rate based upon the rate of heat transfer determined
by the delta-T.  This is the priciple behind one type of mass flowmeter, such
as manufactured by Hastings.

	However, this is probably far too complex and expensive for the
application presented by the author of the original article.

	If this were an industrial application where $100 to $150 is not
a problem to spend, one of the simplest and least expensive flow sensors
utilizes a spring-loaded vane, which when deflected by fluid flow will
move a preset distance and operate a microswitch or reed switch.

	Inexpensive paddle-wheel flow sensors, such as those manufactured
by Signet, connected to an indicator with a setpoint may also be used in
industrial environment.

	For a DIY project involving minimum expenditure, I will make two
suggestions:

1.	Install a pipe fitting such that a slight restriction is created
	in the pipe diameter.  Couple a piezoelectric ultrasonic sensor
	(often used in alarm motion detectors, and readily available on
	the surplus market for a few dollars) to the outside of the pipe
	using a clamp, or possibly epoxy adhesive.  Ultrasonic energy will
	be created during fluid flow by vortices formed at the restriction,
	with this energy being readily detectable by a suitable ultrasonic
	sensor (a miniature ceramic microphone with good high frequency
	response may be a satisfactory substitute for an actual ultrasonic
	sensor).  The ultrasonic energy created should be sufficiently
	broadband that even ultrasonic sensors with narrow bandpass should
	work.  One of the best restrictions can be created by a sharp annulus
	or orifice plate inserted within the pipe; one should not restrict
	the cross-sectional area of the pipe by more than say, 20%.  The
	ultrasonic sensor output is amplified, fed through high-pass filter,
	rectified and fed to an adjustable comparator.  One might also wish
	to throw in some time delay and hysterisis.  Comparator threshhold
	is set to provide an output when a suitable flow rate is detected.
	With proper design, this device should be immune to extraneous
	ultrasonic energy conducted by the pipe from other sources.

	One could easily begin experimentation with the above approach by
	merely connecting a microphone or ultrasonic sensor to an amplifier,
	and observing the output on an oscilloscope.

	Somewhat more complex variations of the above are used in industry
	to provide quantitative flow measurement, with the most common
	example being the vortex-shedding flowmeter.  The above design
	suggestion does not represent a vortex-shedding approach, however.

2.	Create an internal orifice plate as above, and install tees with
	sidearms as small as possible before and after the orifice plate.
	Differential pressure in the low in-H2O range will be created by
	flow fluid across the orifice plate.  Differential pressure switches
	that are adjustable may readily be found on the surplus market, with
	a common original application being aircraft fuel flow sensing.
	One could also use a surplus differential pressure transducer, but
	the differential pressure switch would be a much simpler approach.
	Be careful in using some differential pressure sensors: they may be
	damaged if inadvertently connected in a non-differential manner
	(i.e., some devices are unable to withstand water line pressure
	on one port if the other were open to atmosphere).
 
Larry Lippman @ Recognition Research Corp.  "Have you hugged your cat today?"
VOICE: 716/688-1231   {boulder, rutgers, watmath}!ub!kitty!larry
FAX:   716/741-9635                  {utzoo, uunet}!/      \aerion!larry