Need a way to sense water pressure inexpensively!

[viper]

Hoping to brain storm with some EEs.  We need to sense upto about 100psi of water pressure.  Extreme accuracy is not required but reliability and cost is.  The catch here is this really needs to be an electronic sensor and be completely submerged.  I realize there are cheap threaded transducers but as soon as we examine "submersible" sensors, the price point is pushed well beyond what is acceptable. 

We are working on a potential production device and imagine as an example of lowering a sensor on a wire down to the bottom of a lake.  That is basically what we are doing.  Using a tube or pipe to read air pressure just is not practical. 

We are also in manufacturing so we could examine self mfg of these sensors for our need but I'd like to explore other, more practical solutions. 

The main issue to resolve is that darn wired connection as all that water pressure will seek to penetrate a typical sensor so we know this will require a little thought.  I am certainly open to other possibilities in otherwise reading water pressure.  This is all in 'fresh' water application, and only in water at normal 40-60F temp ranges. 

[T3sl4co1l]

What's wrong with a traditional two port (or one port vs. ambient) transducer?

And how cheap are we really talking?  Those things are ca. $10 I think, if you need less, good luck with that.

Note that you need a reference pressure in any case -- probably an airspace within a rigid, sealed enclosure.

As for isolation, that's easy enough: use a diaphragm, with water or other hydraulic fluid inside, to transmit pressure to the sensor.  (The less air is trapped in the line, the less displacement of the diaphragm is needed.  Check if your sensor needs a headspace or if it can have whatever kind of liquid directly on it.)

I suppose you could shortcut both and use some kind of displacement sensor on the enclosure, but now you need a calibration for the enclosure's spring constant, and it will likely vary widely with manufacture and temperature.  Still not impossible to calibrate out, but how much part cost are you willing to trade for dev and cal cost?

Note that you need very good connectors or sealing, regardless of what type of sensor is used.  Submersible hardware isn't cheap any way you cut it.  Is your application even reasonable to begin with?  Can the cost be offset by an alternative business plan? Etc.

Tim

[geggi1]

If it is going to sitt in a container of some sort you can use a manometer that drives some kind of light moving potensiometer. In your case the manometer must sit in a pocket that is open to the atmosphere above.
The arm on the manometer can have a wiper on the arm that slides on some kind of reststive path. This is how a presure transmitter where made in the old days.
Another option is to use a strain gauge in a chamber and a diaphragm that moves the strain gauge.
But you can get very cheep pressure transmitters from china. By getting the transmitter from china you will probably get it for less cost than making inhouse.
https://www.ebay.com/itm/265124922381?hash=item3dbaad440d:g:MUAAAOSwoZxgeDmE
https://www.banggood.com/Pressure-Transducer-Sensor-Oil-Fuel-Diesel-Gas-Water-Air-Sensor-p-1007341.html?utm_source=googleshopping&utm_medium=cpc_organic&gmcCountry=NO&utm_content=minha&utm_campaign=minha-no-en-pc&currency=NOK&cur_warehouse=CN&createTmp=1&utm_source=googleshopping&utm_medium=cpc_bgs&utm_content=xibei&utm_campaign=xibei-SSC-no-en-tools-0316-2021&ad_id=506701494165&gclid=Cj0KCQjwp86EBhD7ARIsAFkgakjaQpkc6B31DAUZHIREEFBewUZdaZpQfJEfOIefQC_-73raWT8j7tEaAq0CEALw_wcB

From experience it it is often cheepest to get something off the shelf stuff and modify it. If i where to make a sensor according to your description i would use some kind of pipefitting to make a housing. Installing the sensor in a air thigh compartment so that the sensor got a reference pressure. Have the cable going through a potted penetration the keep the reference chamber water thigh. The sensor would be like a insideout of the normal installation.

[ajb]

Use a through-mount threaded transducer in reverse?  IE, mount it in some sort of rigid housing with the sensing end pointed out, then a suitable water-tight connector/gland for the cable.  With some care in part selection the housing can maybe be a relatively simple tube with threads for the sensor on one end and threads for an off-the-shelf cable gland on the other, and relatively simple to churn out on a CNC lathe if you need quantity. 

[floobydust]

You'll need to design a gland that seals the cable's jacket and each conductor.
Most sensors have no external pressure rating, only a range for the monitoring port. I've had lots of failures with hydrostatic pressure transducers. For long-term drift and reliability, piezometers are the best and I've used them where they end up mostly covered by cement and need a service life of at least 10 years. You need to say lifetime and resolution and cost, to see if what you want is even feasible.

The main problem is the cable, being properly sealed at the sensor end, and able to withstand the pressure. Water likes to get in the cable and no spec is given with low cost stuff or considered by most.
At 200ft length (100psi hydrostatic) you get limited to what sensor technology you can use. Corrosion of the membrane is how they usually fail.

[JohnnyMalaria]

We are working on a potential production device and imagine as an example of lowering a sensor on a wire down to the bottom of a lake.  That is basically what we are doing.

Are you trying to measure the ambient water pressure at depth? If so, reliable and cheap technology exists for that (scuba). There may be options available with tethered signal wires for monitoring at the surface.

[langwadt]

Use a through-mount threaded transducer in reverse?  IE, mount it in some sort of rigid housing with the sensing end pointed out, then a suitable water-tight connector/gland for the cable.  With some care in part selection the housing can maybe be a relatively simple tube with threads for the sensor on one end and threads for an off-the-shelf cable gland on the other, and relatively simple to churn out on a CNC lathe if you need quantity.

instead of trying to seal with a cablegland, put a hose barb on the "house", run a hose to the surface with wires inside. Always provides an
air path if measuring relative pressure is needed

[Robert Smith Eco Warrior]

Try remote tyre pressure sensors.
They have them fitted on race cars. They have no connections just a radio link from sensor in the wheel to a box on the car. I am not sure how water affects radio signals but I should think making the sensor a waterproof item shouldn't be too difficult.

[Mr.B]

instead of trying to seal with a cablegland, put a hose barb on the "house", run a hose to the surface with wires inside. Always provides an
air path if measuring relative pressure is needed

I have done exactly this before.
I used a stainless steel hose barb, good quality garden hose and a stainless steel hose clip/clamp.
I was measuring to 100m salt water. (328ft)

[Ground_Loop]

It sounds like you're looking for water level/depth rather than pressure. Nonetheless, knowing one will give you the other.

[floobydust]

I have done exactly this before.
I used a stainless steel hose barb, good quality garden hose and a stainless steel hose clip/clamp.
I was measuring to 100m salt water. (328ft)

A hose barb and with spring clamp is not suitable for 100psi, nor is a garden hose in this application.
You have atmospheric pressure inside the hose, and 100psi fluid pressure outside which is a negative pressure differential and it collapses the hose which is designed for the opposite. It's unnoticed unless you have an underwater camera or the pressure readings mysteriously go wrong.

Polyurethane vacuum hose has a coil inside and weighs around 40lb for a 200ft run plus the electrical cable. That alone is likely $500.
Diving hose and fittings might be something to look at.

[Mr.B]

You have atmospheric pressure inside the hose, and 100psi fluid pressure outside which is a negative pressure differential and it collapses the hose which is designed for the opposite. It's unnoticed unless you have an underwater camera or the pressure readings mysteriously go wrong.

Very good point.
I should have added that I was using a ceramic transducer for absolute pressure.
It is sealed with a known internal pressure - vacuum I think, so any pressure changes behind the sensor do not interfere with the readings.

[CatalinaWOW]

As others have said, you appear to be measuring depth, in which case the technology of choice is a variant of a tape measure.  Those made with plastic tape and a reel are corrosion resistant, cheap and easily good for 100 meters.  If you want an electronic output the conversion is far simpler done in air at the surface.  The most baroque way I can think of is to use a video camera and AI software to read the tape.  There are far simpler methods readily available.

Even if pressure is what you want, given your not very accurate comment a simple constant times the depth will do the job.  Adding corrections for salinity and temperature can improve the accuracy, and even deal with the minor compressibility effects.  All a simple job for that cheap microchip that no one can buy these days.

[floobydust]

It's just very difficult MecE design.
One company I worked for came up with a low cost solution, it was all unicorns and rainbows. They won many bids and deployed sensors- only to see them fail after a few months. The time-constant to failure is usually much longer with wet mechanical stuff compared to electrical. So they tried to fix and redeploy under warranty and repeat et. al that loop until all customers hate you and you're almost out of business.

Example: Sensor readings are bad, I opened up a surface junction box and found it was full of water. Turned out to be from capillary action of the wire that went down to the sensor. Wire makes great tubing I discovered, if you give it a few weeks.

The DP sensor versions a few got condensation drops in the vent hose, others a kink in the hose, plugging that vent. Some sensor ports got silt and sand plugging it, used sintered filter. All unexpected problems. The company tried filling the backsides with grease, silicone oil etc. in desperation it was kind of funny to watch people learn about engineering.
An absolute pressure sensor has barometric pressure to correct for if high accuracy is needed.
Best we got was a sub to enclose the sensor and the cable gland and we made our own. Even though a sensor is cheap, the finished solution sure adds up.

A commercial tape measure Geokon/Solinst 101 sorta cheap cable USD $520. Note Geokon has no pressure spec for their cables.
There are also float and GWR methods if you just want to track smaller changes in level vs the 200ft absolute. It depends on what OP is monitoring water level for.

[Conrad Hoffman]

I'd think a simple metal tube or chamber, with a strain gage glued to the inside would do the trick. Electronics in the chamber most likely, so only remaining problem is getting wires to the surface, as discussed above.

[fcb]

Try using an FSR and just pot it in silicone, you could put signal conditioning electronics etc.. if you need to drive a long distance.

Digikey: 1027-1000-ND
Interlink FSR400, 5.08mm dia area, 0.2N to 20N range.

My thinking is:
100psi = 444.8N over 645.16mm²
5.08mm dia = 20.2683mm²
Force on sensor =14N, so within range

[twospoons]

FSRs are horribly non-linear and wildly variable.  Maybe you could calibrate it, but the money spent on calibrating every sensor would be better spent on a better sensor.

The truth here is that there really is no 'cheap' solution - you want a waterproof electronic sensor to read up to 100psi, and this is simply not a cheap thing to do.

[IDEngineer]

Look at water or oil pressure sensors from the automotive industry. Thanks to the huge volumes that industry tends to drive cost down and reliability up, with wide availability. Just another idea.

[voltsandjolts]

Maybe the 'dive watch' type sensors would suit your needs/budget.
e.g. MS5837 from TE or such like
https://www.digikey.co.uk/products/en/sensors-transducers/pressure-sensors-transducers/512?k=ms5837
https://www.te.com/usa-en/products/sensors/pressure-sensors/board-level-pressure-sensors.html?tab=pgp-story

Maybe this is a better option:
https://www.te.com/usa-en/product-MS580307BA06-00.html