Maker Pro
Maker Pro

Water sensor - dynamic thermal resistance sensor

  • Thread starter Klaus Kragelund
  • Start date
K

Klaus Kragelund

Jan 1, 1970
0
Hi

I wrote an earlier post regarding a water sensor for a pump (see below
for original post):

The first idea dealed with using a capacitive sensor, but that has
several problems for example the problem of the sensor being affected
by incoming water because of an enclosure that is not hermetically
sealed.

Well - another idea is to use an totally sealed enclosure by using a
sensor that will work through the barrier of the stainless steel
housing. It works by measuring the dynamic thermal resistance of a NTC
resistor. The NTC is placed just on the other side of the enclosure
wall next to the water surrounding the pump. The NTC has a good thermal
contact to the stainless steel.

The sensor works by dissipating a known power in the NTC resistor.
After some time the power is turned off and the NTC resistor is used to
measure the temperature and more importantly the negative slope of the
temperature when the power is off. A high slope indicates that water is
present to cool off the NTC through the thermal resistance of the
NTC/steel connection, the steel itself and finally the water. In case
of no water the slope is not as steep.

Thats all fine because a sensor like this may not be as sensitive as
the capacitive one because the impedance levels are lower and in case
of water that penetrates the enclosure the effect of the volume
conductivity will be minimal.

But - can anyone tell me how to calculate the thermal resistance of the
enclosing water and the thermal resistance of the water/steel junction
to get an indication of how good this sort of sensor would work through
a steel enclosure?

Thanks

Klaus

(Sorry for the lenghty post)
OLD POST:

"I'm working on a pump that is suspended in water and I need to add
some
electronics to it (not more than 1 meter water level)


My initial thought was to house the electronics in a sealed
plastic/rubber enclosure, but someone pointed out that plastic/rubber
might over time absorbe water and render the electronics un-usable


Anyone got details of the behavour of plastic enclosures exposed to
long
periods of total suspension i water?"
 
J

John Fields

Jan 1, 1970
0
Hi

I wrote an earlier post regarding a water sensor for a pump (see below
for original post):

The first idea dealed with using a capacitive sensor, but that has
several problems for example the problem of the sensor being affected
by incoming water because of an enclosure that is not hermetically
sealed.

Well - another idea is to use an totally sealed enclosure by using a
sensor that will work through the barrier of the stainless steel
housing. It works by measuring the dynamic thermal resistance of a NTC
resistor. The NTC is placed just on the other side of the enclosure
wall next to the water surrounding the pump. The NTC has a good thermal
contact to the stainless steel.

The sensor works by dissipating a known power in the NTC resistor.
After some time the power is turned off and the NTC resistor is used to
measure the temperature and more importantly the negative slope of the
temperature when the power is off. A high slope indicates that water is
present to cool off the NTC through the thermal resistance of the
NTC/steel connection, the steel itself and finally the water. In case
of no water the slope is not as steep.

Thats all fine because a sensor like this may not be as sensitive as
the capacitive one because the impedance levels are lower and in case
of water that penetrates the enclosure the effect of the volume
conductivity will be minimal.

But - can anyone tell me how to calculate the thermal resistance of the
enclosing water and the thermal resistance of the water/steel junction
to get an indication of how good this sort of sensor would work through
a steel enclosure?

Thanks

Klaus

---
No offense intended, but that's all total bullshit.

What is it that you want to do? Start the pump when the water falls
to a certain level and then stop it when the water rises to another?

Or something altogether different?

Tell us what you want to do without telling us how smart you are
(but how you can't quite work out a solution to your problem) and
maybe you'll get some good answers.

Grrr...
 
M

martin griffith

Jan 1, 1970
0
On 5 Oct 2005 13:53:21 -0700, "Klaus Kragelund"
snip
Tell us what you want to do without telling us how smart you are
(but how you can't quite work out a solution to your problem) and
maybe you'll get some good answers.


Grrr...


Calm down Mr Fields. I've just finished my water divining stick with
USB/Wifi interface. I'll put the extra software in that will handle
the this, no more than 100K of code extra. The Leyline detection code
will have to go though, it was a bit buggy around the summer solstice
anyway :(


martin
 
R

Rene Tschaggelar

Jan 1, 1970
0
Klaus said:
Hi

I wrote an earlier post regarding a water sensor for a pump (see below
for original post):
[snip]

But - can anyone tell me how to calculate the thermal resistance of the
enclosing water and the thermal resistance of the water/steel junction
to get an indication of how good this sort of sensor would work through
a steel enclosure?

The system NTC/stainless steel tube is too slow to detect no-water
when the pump runs dry. This condition should be detected in miliseconds.

Rene
 
R

Rich Grise

Jan 1, 1970
0

Hi. :)
I wrote an earlier post regarding a water sensor for a pump (see below
for original post): ....
Well - another idea is to use an totally sealed enclosure by using a
sensor that will work through the barrier of the stainless steel
housing. It works by measuring the dynamic thermal resistance of a NTC
resistor. The NTC is placed just on the other side of the enclosure
wall next to the water surrounding the pump. The NTC has a good thermal
contact to the stainless steel.

....

Well, if it was me, I wouldn't rely on the thermal impedance of water,
unless it's chilled and/or flowing (like in a hot-wire anemometer) -
I'd think the thermal conductivity of the steel would swamp out any
effect from presence or absense of water, but I'm guessing.

What I might do is use ultrasonics. Get a small transducer, like
they'd use in an ultrasonic thickness gauge, and I'd think that
the difference in acoustic impedance between air and water would
be much easier to detect. But, like I said, it's just one of my
crazy ideas. (I have a little passing familiarity because I once
got a job at RS by fixing an ultrasonic "fish finder" that the
other two techs had been scratching their head over for about
a month. There was a significantly higher return echo when the
transducer was in water vs. in air. :) )

Good Luck!
Rich
 
J

Joerg

Jan 1, 1970
0
Hello Rene,
The system NTC/stainless steel tube is too slow to detect no-water when
the pump runs dry. This condition should be detected in miliseconds.
Tell me about it. The booster pump for our pool sweep committed suicide
within a minute when something in front failed and it got no water. I
could repair it but that was a messy one-hour job.

Best would be to alert when cavitation or air bubbles occur. Ultrasound,
inductive, whatever works best. Cavitation can clearly be heard so
acoustic sensing might work as well. Ours sounded like a gear box that
ran out of oil, very loud.

Another feature that would have saved the pump from seizing would be a
thermal sensor. When that happened I ran down to the pool house and the
enclosure of the pump (not motor) had gotten too hot to touch. That may
be something for Klaus to try.

Regards, Joerg
 
W

Winfield Hill

Jan 1, 1970
0
Rich Grise wrote...
I have a little passing familiarity because I once got a
job at RS by fixing an ultrasonic "fish finder" that the
other two techs had been scratching their head over for
about a month. There was a significantly higher return
echo when the transducer was in water vs. in air. :)

LOL, that's funny!
 
R

Rene Tschaggelar

Jan 1, 1970
0
Joerg said:
Hello Rene,

Tell me about it. The booster pump for our pool sweep committed suicide
within a minute when something in front failed and it got no water. I
could repair it but that was a messy one-hour job.

Best would be to alert when cavitation or air bubbles occur. Ultrasound,
inductive, whatever works best. Cavitation can clearly be heard so
acoustic sensing might work as well. Ours sounded like a gear box that
ran out of oil, very loud.

Another feature that would have saved the pump from seizing would be a
thermal sensor. When that happened I ran down to the pool house and the
enclosure of the pump (not motor) had gotten too hot to touch. That may
be something for Klaus to try.

Joerg,
one approach would be a precision current monitor. When
the pump runs dry, the torque is much lower and this
should be detectable.

The problem with these pumps is :
1) they are using slide bearings not ball bearings. And
the slide bearing runs on a water film, not oil.
2) they may depend on the case being watercooled.

If they could cost more than a few dollars, they could
be made to run dry without problem.

Rene
 
J

Joerg

Jan 1, 1970
0
Hello Rene,
one approach would be a precision current monitor. When
the pump runs dry, the torque is much lower and this
should be detectable.


That's dicey. I tried that with ours. The difference isn't that great
unless you are running a pump motor at almost full torque which tends to
overheat it in summer.
The problem with these pumps is :
1) they are using slide bearings not ball bearings. And
the slide bearing runs on a water film, not oil.


The motors are usually ball bearing types.

If they could cost more than a few dollars, they could
be made to run dry without problem.


Except for the gasket which is two ceramic disks spring loaded to hold
them tight.


Regards, Joerg
 
K

Klaus Kragelund

Jan 1, 1970
0
Joerg said:
Hello Rene,

Tell me about it. The booster pump for our pool sweep committed suicide
within a minute when something in front failed and it got no water. I
could repair it but that was a messy one-hour job.

Best would be to alert when cavitation or air bubbles occur. Ultrasound,
inductive, whatever works best. Cavitation can clearly be heard so
acoustic sensing might work as well. Ours sounded like a gear box that
ran out of oil, very loud.

Another feature that would have saved the pump from seizing would be a
thermal sensor. When that happened I ran down to the pool house and the
enclosure of the pump (not motor) had gotten too hot to touch. That may
be something for Klaus to try.

Thanks - and we do actually have a thermal sensor inside the motor (an
UL demand due to the new rules that an a software system needs a
redundant satefy system - total crap)

Regards

Klaus
 
K

Klaus Kragelund

Jan 1, 1970
0
We have this sort of current sensor designed and working inhouse, but
we are attempting other more cheaper methods

Thanks

Klaus
 
K

Klaus Kragelund

Jan 1, 1970
0
I'm not so smart - otherwise I would be posting to this NG

No bullshit Q:

Need a waterlevel detector for a pump - must cost nothing, and no fancy
electronics must be used (costly), must be slim (placed vertically
along side the metallic chassis of the pump)

Thanks

Klaus
 
J

jtaylor

Jan 1, 1970
0
Rich Grise said:
Hi. :)


...

Well, if it was me, I wouldn't rely on the thermal impedance of water,
unless it's chilled and/or flowing (like in a hot-wire anemometer) -
I'd think the thermal conductivity of the steel would swamp out any
effect from presence or absense of water, but I'm guessing.

What I might do is use ultrasonics. Get a small transducer, like
they'd use in an ultrasonic thickness gauge, and I'd think that
the difference in acoustic impedance between air and water would
be much easier to detect. But, like I said, it's just one of my
crazy ideas. (I have a little passing familiarity because I once
got a job at RS by fixing an ultrasonic "fish finder" that the
other two techs had been scratching their head over for about
a month. There was a significantly higher return echo when the
transducer was in water vs. in air. :) )

What about a pressure switch on the output side?
 
J

Jim Thompson

Jan 1, 1970
0
Hi

I wrote an earlier post regarding a water sensor for a pump (see below
for original post):
[snip]

I take it that pump is "submersible", totally under the water in
normal operation?

Couldn't you simply measure output pressure?

Around here we periodically drain our swimming pools (to get rid of
mineral concentration due to evaporation). A submersible pump is
used, which turns off automatically when the pool is empty. Wonder
what they use?

...Jim Thompson
 
J

Joerg

Jan 1, 1970
0
Hello Klaus,
We have this sort of current sensor designed and working inhouse, but
we are attempting other more cheaper methods
Cheaper? A current sensor could be done with a current transformer and
burden resistor. If that's too expensive you could buy the toroids and
wind your own which should drop the materials cost to less than a buck.
Or a few Kroner in Viking currency ;-)

Regards, Joerg
 
J

Joerg

Jan 1, 1970
0
Hello Klaus,
Thanks - and we do actually have a thermal sensor inside the motor (an
UL demand due to the new rules that an a software system needs a
redundant satefy system - total crap)
Ahem, I respectfully disagree. Read about a software system bug that
sent a large military vessel crashing into a rocky island:
http://www.ganssle.com/tem/tem116.pdf

A few million Dollars later a lesson had been learned.

Regards, Joerg
 
J

Jim Thompson

Jan 1, 1970
0
Hello Klaus,

Cheaper? A current sensor could be done with a current transformer and
burden resistor. If that's too expensive you could buy the toroids and
wind your own which should drop the materials cost to less than a buck.
Or a few Kroner in Viking currency ;-)

Regards, Joerg

I guess it's "cheaper" to replace burned out motors ?:)

...Jim Thompson
 
J

Joerg

Jan 1, 1970
0
Hello Jim,
Couldn't you simply measure output pressure?
Yes, but pressure sensing is more expensive than current sensing. Plus
it needs to be plumbed in and the tap-off tube to the sensor gets
clogged and it stops sensing (happened in our pump house).
Around here we periodically drain our swimming pools (to get rid of
mineral concentration due to evaporation). A submersible pump is
used, which turns off automatically when the pool is empty. Wonder
what they use?
Probably just the usual float switch.

Draining a pool? Now I know why we are always short of water in the
west. Out here nobody would dare to do that since the water bill would
be excruciating. We keep the chems balanced despite evaporation. My test
kit measures seven parameters and at least once a year we carry a sample
to the store where they do more tests, to see what needs to be balanced
out. Since our last pool guy just disappeared I began doing it myself
and for some reason now the pool store never finds an imbalance. No more
scaling of the walls either.

Regards, Joerg
 
J

Jim Thompson

Jan 1, 1970
0
Hello Jim,

Yes, but pressure sensing is more expensive than current sensing. Plus
it needs to be plumbed in and the tap-off tube to the sensor gets
clogged and it stops sensing (happened in our pump house).

Probably just the usual float switch.

Draining a pool? Now I know why we are always short of water in the
west. Out here nobody would dare to do that since the water bill would
be excruciating. We keep the chems balanced despite evaporation. My test
kit measures seven parameters and at least once a year we carry a sample
to the store where they do more tests, to see what needs to be balanced
out. Since our last pool guy just disappeared I began doing it myself
and for some reason now the pool store never finds an imbalance. No more
scaling of the walls either.

Regards, Joerg

The water here is VERY HARD, so every few years it's necessary to do a
draining, or the calcium becomes uncontrollable.

Up where you live the evaporation is probably quite low compared to
here.

(My reef tank, INSIDE, loses ~2-3 gallons per day. I'm contemplating
giving it its own RO system ;-)

...Jim Thompson
 
J

John Woodgate

Jan 1, 1970
0
I read in sci.electronics.design that Joerg
Since our last pool guy just disappeared

Did you check the pH? Maybe he dissolved.
 
Top