erratic reading of ultrasonic level transmitter


Thread Starter


Can anyone tell what are the causes of an erratic reading of an ultrasonic level transmitter?

what can affect the reading of the level transmitter?


email me: [email protected]

you can get erratic readings at very short distances, around the 1 inch or less from the transmitter/receiver head. Is this where the trouble lies?

Issac Issachar

Reasons for jittery/noisy readings:

1) jitter is real:
surface turbulence due to
- mixing/pumping/inflow
- rotating agitators/scrapers

a stilling well can be used to dampen turbulence

2) jitter is induced noise due to
- RF interference from cell phones, hand held radios, bluetooth, wireless
- RF interference from VFD's
- gound loop, ground problem
- poor cabling (not twisted pair) or cabling installation (cable run in same conduit has AC
high voltage)

3) artifact of ultrasonic technology
- installation too close to tank wall
- deposits on stilling well wall

4) For open channel flow level wind induced waves/turbulence can be a factor

There are lots of things that would cause erratic readings. First, define what you mean by erratic readings. Second, what are you measuring the level of? Liquid, solid, powder, slurry? What temperature is the measured surface? What is the temperature at the sensor? Are they different? What is the ultrasonic level sensor mounting arrangement? Is it in a tank? A closed vessel? An open vessel? Exposed to wind? Mounted close to a side wall? Mounted above an agitator paddle? What's the measured distance at zero level? At maximum level?

If you think about it, the answers to those questions are critical, and they'll tell you what's wrong.

For example, I once talked to the plant engineer of an oil-fired power plant (it doesn't matter where except that it is cold in Winter there). He told me he had bought, in November, an ultrasonic level gauge for his primary bunker oil storage tank. It worked wonderfully all they way up to about the first of May, but then the readings became erratic and it didn't look like there was any easily observable reason for it.

Turns out that the weather had warmed up. Hydrocarbon liquids give off a "blanket" of vapor all the time. When it is cold, the distance between the top of the vapor blanket and the liquid is very small. When it warms up, the blanket rises. In some cases, the density differential between the top of the vapor blanket and the air above it is high enough and sharp enough to produce a valid echo to the ultrasonic level sensor. This "ghost level" is an often erratic phenomenon that is well known.

Tell us more.

Walt Boyes
Editor in Chief
Control and

Mailto:[email protected]
Read my blog SoundOFF!! At
Could it also be because of improper sampling techniques used? (box car, etc). By the way, what *is* a good sampling technique? Especially for turbulent or semi-turbulent surfaces?

> Turns out that the weather had warmed up. Hydrocarbon liquids give off a "blanket" of vapor all the time. <

Indeed this blanket issue is very interesting, I have not hear of it before!
Another spring time seasonal ultrasonic anomaly - the air temperature above the water in a closed, but vented tank will drop below the air's dewpoint temperature.

The resulting fog and condensation on the ultrasonic transducer face will either produce a 'loss of echo' type failure or wildly varying level readings.

Yep, that's another good one.

Here's another...light foam with a thickness of 1 cm (or about half an inch) will generally be transparent to ultrasonics, and the level measurement will work. Note that I said _generally_ because it is sometimes NOT.

But if you get thick foam, like nitrate foam at the outfall of a wastewater plant, for example, or the foam in a juice tank, the reflection from the top of the foam looks like a level.

If the foam layer is thick, but soft, so that there is not a reflection, the foam will absorb the ultrasonic signal and you will either get a fail or a zero level indication, depending on how good your level gauge is.

Walt Boyes
Editor in Chief
Control and

Mailto:[email protected]
Read my blog SoundOFF!! At
Short explanation of condensation:

Humidity in the air behaves in the way that it goes from warmer air to the colder air. (This is due to some physical phenomena.) So, if the antenna is colder than air, the concentration of humidity is highest close to the sensor’s antenna. Point of condensation depends on two factors: humidity concentration in the air, and the temperature of the surface (antenna’s surface in this case).

We have this problem during winter, because our level sensor is mounted on concrete plate - below is the big reservoir with warm water, and above is the area where can be quite cold during winter, so the coldness “goes through” the sensor’s body and cools down the antenna. We will change two things to solve this problem:

1. We will not mount sensor tightly to the concrete plate. Instead of this, we will make a little big bigger hole in the plate, and lift the sensor a little bit above the plate. In this way, warm air from reservoir will go out through the hole and warm up the sensor.
2. We will put the sensor in closed and temperature isolated box.

So, we hope that result will be that sensor and antenna have more-or-less same temperature as the air in the reservoir, and there will be no condensation. :)
I came back to inform about results, since we implemented described solution. Only difference is that we didn't put the measuring device inside the closed box, but we put it in the box which is wide open from one side. It happened to be wise, because later we found that the air inside the basin occasionally reached 100 degrees C, and the device would have been cooked if it was closed inside the box. (My English grammar is not so good.. :( :) ) Anyway, it works fine now :)
Your friendly local moderator here. Thanks for reporting back. Letting other readers of the forum know what happened and what worked is great. Your experience can be used by others with the same problem.

Peg Ferraro