high accuracy non-contact level measuring


Thread Starter


i need a non-contact non-laser level measuring device that can read with an accuracy of 0.0005". i've got a 30" x 30" x 30" tank, and I can mount the device inches from the liquid.

whew! hoo boy....

lasers can do it, but my liquid optical properties won't cooporate.

pressure differential devices might have the resolution, but the liquid is really thick, and i shouldn't contact the liquid.

other than those, i can't find a level technology that will have this high resolution. any clues or leads?


Robert Scott

For .0005" accuracy to make sense, you would have to have a perfectly calm and clean surface. How about using a linear one-axis servo system that positions a flat plate to within .005" of the surface, servoing for a fixed capacitance with the liquid (assuming it is somewhat conductive) to maintain the fixed offset? Infer the liquid level from the postion of the servo system, which can be measured from a linear encoder tape. To keep the plate clear and dry you need to ensure that the servo is always active, and that the level does not change too fast for the servo to follow it.

Robert Scott
Real-Time Specialties
Embedded Systems Consulting

Curt Wuollet

First you need a block of granite set on bedrock to put your tank on. I don't believe I've ever seen a liquid that still.



Chris Jennings

If legislation is similar to in Australia you can only use a nuclear device if there is no other appropriate technique for measuring the value (i.e. last resort). Make sure you keep that in mind.

Chris Jennings
Those are great ideas. I notices some sensors ultrasonic sensors from Consense that boast 0.0002" accuracy. I have emailed them, but they haven't contacted me yet.

I haven't looked into nuclear level control yet. I'm kind of scared by the name, but I'll investigate.

I like the capacitance-on-servo idea. I'll investigate that also. Any leads on where I can get started?

As for the still liquid, it's very viscious, and the tank is supported and positioned with dampened servos that have an accuracy of 0.0001".

thanks for all the input
Place load cells under your tank and weigh the tank and its contents. Load cells are the most practical, and often the cheapest, non-contact, non-invasive method of measuring fluid in a tank. Load ceall do not interact with the fluid, don't clog, dont care about dirt, dont care about fluid turbulence, and are maintenance free and simple to use.

Gerald Beaudoin

With that kind of accuracy... should make a pretty good siesmic tremor sensor too!

Gerald Beaudoin
Ultra sonic... never. If you haven't already, check with SaabTankControl for their radar systems.

Manning Environmental
Your Solution for Fluid Samplers
I like the idea of load cells, but I have different fluid containers, and with load cells, I will have to control the weight of the containers and accomodate for their weight tolerance in my fluid heigh uncertainty.

I keep reading that radar is more accurate (although more expensive) than ultrasonic. But i have found ultrasonic level detectors with an accuracy of 0.0002", and have yet to find a radar that get under 0.02". The highest Saab product that i found (TankRadar Rex) has a precision of 0.5mm.

Thanks again for all the input. Any leads into a radar system with an accuracy of 0.0005"?
A few suggestions:
1. If you haven't ploughed through the proximity sensor category in the Thomas Register then that might help you find something. http://www.thomasregister.com/
2. Directindustry might be worth a try/search.
3. Try searching for the non-contact transducers/sensors used in the reference position measurement in machine tools. (CNC machines)
4. Try some calculations on the position of a reflected low angle light beam. What could you do with that? Very viscous.... might be a problem.

Vince Dooley
You are misunderstanding the specs of an ultrasonic level instrument. What you are confusing is precision with accuracy. Show me an ultrasonic level instrument with an accuracy of four-zeroes-two and I will eat it.

Control magazine
Putman Media
555 West Pierce Road, Suite 301
Itasca, IL 60143-2649
Ph. 630-467-1300 ext. 368
Fx. 253-981-0285
[email protected]
I would judge the .0002 from ultrasonics is under a controlled environment with little to no attention being given to changes in vapor density or temperature which will most certainly effect accuracy.

Daniel Moroncini

My first reaction is to say you: pls be kind and tell us why you need that accuracy!. This figure is difficult also for contact technologies

ultrasonic is not addecuated: litle changes in ambient temperature will causes errors

I think you have to search in the world of numeric control more than in process instrumentation

good luck

Daniel Moroncini