Ultrasonic and radar level sensing technologies are usually described as competitors, pitted against each other, in a struggle for level-sensing supremacy. “Ultrasonic is better!” “Radar is the best!” The truth is that they are far more complementary than they are competitive, with each excelling in different situations. Anybody pitching either technology as a One-Tech-Fits-All level measurement solution isn’t telling the whole truth about the technologies and their products.
A basic non-contact level measurement technology assumption is “Go with ultrasonic, because it’s cheaper, unless the application is complicated. Then use radar.” While that assumption is partially true, it doesn’t tell the whole story. Yes, measure for measure, ultrasonic sensors are cheaper. And yes, radar does handle most not-so-simple applications better, but it doesn’t handle all of them better. Sometimes, the simplicity of a “simple application” is in the return signal of the sensor, not in the eye of the engineer or systems manager deciding which sensor tech to buy.
Ultrasonic level sensors, in general, are better at dealing with piling solids than radar. The signal from a single-frequency radar will be partially absorbed, partially reflected, and partially deflected by the angled target surface of the piling substance. Ultrasonic waves, however, will reflect enough to get a decent signal, as long as the sensor is capable of reading flat surfaces twice as far away as the angled surface. And while dual-frequency radar sensors are capable of dealing with the complicated reflections off the angle of repose, the signal processing required is much more complicated than what is need for single-frequency radar or ultrasonic.
Radar level sensors provide better results than ultrasonic sensors when there is atmospheric interference. Dust, foam, and vapors are the three most often named culprits, but almost any physical substance or object that could come between a sensor and the target surface is easier to “see through” with a radar sensor than with an ultrasonic sensor. This includes dual-surface interface detection: radar can be set to look past the initial surface to the point of interface between two (usually) liquids. Ultrasonic sensors just can’t do that. And since radar sensors generally have a narrower beam spread, they can often be used in tighter, more cluttered spaces than ultrasonic sensors.
Here’s the huge caveat: if you use bad tech, it isn’t going to work well. A poorly designed sensor, whether it uses sound waves or electromagnetic waves, won’t produce a good measurement, no matter how well it should fit the application. For example, full-power radar sensors promise to be the only non-contact level sensor you’ll ever need. But because the premise is faulty, the sensors end up causing as many problems as they solve.
Full-power radar sensor manufactures promise a narrow beam-spread angle, which sounds great. But what they don’t tell you is that the secondary side-beams generated by their full-power pulse create just as many echoes as a wider primary-beam. When you have a full-power primary wave and noisy secondary waves, everything around the sensor and the target will generate echoes. Everything.
So, in order to compensate for making all that racket, a full-power radar sensor has to deploy a huge amount of front-end signal processing. Every echo that returns to the sensor has to be analyzed: “Is this the echo I’m looking for?” All that processing means a longer delay between sending the initial pulse and generating an output, and more time means more power being used by the sensor.
Some full-power radar manufacturers will tout the ability of their radar to accept custom parameters, specific to each application. Which is all well and good, until you realize: One, those custom parameters are necessary to blackout echoes from permanently installed objects; two, a technician has to sit with the newly-installed radar sensor to figure out where those blackouts should be, which is extra time and money; and three, loading those parameters from memory means extra boot time every time the sensor is turned on. That doesn’t sound like a feature.
Those blackouts that full-power radar sensors use to ignore echoes from stationary objects are also a problem in and of themselves. In order to compensate for creating too many echoes, a full-power radar sensor creates blind spots within the target range. When the target level moves through a blackout area, the sensor is literally guessing what the output should be. Sure, it can base the output on the immediately previous rate-of-change for the target level, but if the rate-of-change suddenly changes, you have no way of knowing until the echo from the target surface emerges from the blackout area.
Full-power radar sensors also need to “guess” when it comes to dual-surface interface detection. Because the sensor fills the measurement environment with echoes, the return signal from a dual-surface interface looks like one big echo, instead of distinct echoes for each surface. So once again, the differentiating of unwanted echoes from the desired primary surface echo and dual-surface interface echo is left to signal processing. At some point, it’s fair to ask if the sensor manufacturer is as good at designing signal processing software as it is poor at designing radar sensors.
Another design feature of some radar level sensors is a convex antenna. The selling point is that a shorter antenna means the radar level sensor can be used in smaller, tighter areas, where a traditional cone antenna might not fit. There are, however, two problems related directly to choosing an antenna of this shape. First, any condensation that forms on the antenna will collect at the apex (nadir?) of the curve. Excess condensation will drip off, but what remains directly interferes with the radar signal. Second, honest radar level sensors with convex antenna will advertise the necessity of a companion air compressor to supply antenna-clearing compressed air. That’s right, you need to buy an air compressor with these radar level sensors. The less honorable manufacturers won’t tell you about the need for cleansing compressed air until you have problems with your already-installed radar level sensor.
From the booming extra echoes of full-power radar to the required air compressor accessories for convex-shaped antenna, poorly designed radar level sensors simply create more problems on their way to measuring levels. Choosing to use these kinds of radar sensors simply because “Radar is better than ultrasonic” is choosing to trade one set of problems for another. Yes, there are applications where a radar level sensor is a better choice than an ultrasonic level sensor, but that radar level sensor has to be well designed. Poorly engineered equipment always leads to level measurement problems.
Have more questions on full-power radar sensors? Wondering if an ultrasonic sensor or radar sensor would be a better fit for your application?
