Thread Content
In the field of industrial automation, radar level gauges have become the preferred choice for measuring liquid and material levels, thanks to their advantages such as non-contact operation and high reliability. However, when purchasing or using them, parameters such as “Analog-to-digital conversion accuracy: 16bit” and “24bit high precision” are often seen. What exactly do these two numbers represent? How big is the gap between them?
I. What is “digital-to-analog conversion accuracy”? The digital-to-analog conversion accuracy of radar level gauges, in simple terms, refers to the precision with which the analog signal (voltage/current) generated by the radar echoes is converted into a digital signal; this accuracy determines how precisely the level can be measured in millimeters and what the error margin is.
The radar level gauge measures the liquid level height (for example, 3.25 meters), which is a numerical value. However, many traditional controllers or remote transmission systems only recognize 4-20mA analog signals; therefore, the instrument must convert the digital value (3.25 meters) into the corresponding current value (such as 12mA).
Precision (bits) determines how “fine-grained” this conversion process is. To give a concrete example: 2-bit precision means that the range of 0–10 meters can only be divided into 4 segments, with each segment being 2.5 meters long. You measured 5.1 meters, but the output can only be chosen as either the “5-meter range” or the “7.5-meter range”; the error is quite large. 16-bit precision → divided into 65,536 segments. 10 meters ÷ 65536 ≈ 0.00015 meters, which is a resolution of 0.15 millimeters. 24-bit precision → divided into 16777216 segments. 10 meters ÷ 16777216 ≈ 0.0000006 meters, which is 0.6 micrometers. In simple terms: the higher the number of bits, the closer the analog output is to the actual measured value.
II. 16bit vs 24bit: What’s the difference? 1. The significant difference in values: 16 bits = 2¹⁶ = 65,536 scales, while 24 bits = 2²⁴ = 16,777,216 scales. The difference between the two is 256 times.
2. Differences in practical applications: Assuming a range of 10 meters with a 4-20mA output (corresponding to a range of 16mA): It may seem that 24 bits offer much greater precision, but in actual level measurement, a resolution at the micron level isn’t necessary!
3. The measurement accuracy of the radar antenna itself is the bottleneck; the measurement accuracy of most industrial radar level gauges ranges from ±1mm to ±5mm. Even if the analog-to-digital conversion achieves 24 bits (0.6 μm resolution) and the output signal is highly detailed, the physical measurement error remains on the order of millimeters. Excess fine scales will be drowned out by noise.
To give an example: you use a ruler that can measure with precision down to millimeters, and then a digital display that shows numbers to 6 decimal places – the digits beyond that are actually meaningless and keep changing randomly.
III. When is 24-bit meaningful? In a few scenarios, 24-bit is superior to 16-bit: very large ranges (>50 meters): high resolution is required to detect minor level changes (such as for evaporation or leakage monitoring). High repeatability is required: 24-bit D/A conversion chips generally have lower temperature drift and nonlinear errors, resulting in a more stable output. Used in conjunction with high-precision radars (±0.5mm level): such as in laboratories or certain special chemical reactors. Long-distance signal transmission: After passing through long cables, the quantization noise of 16-bit signals can be amplified, while 24-bit signals have more headroom.
But in 90% of industrial applications, such as water treatment, oil tanks, powder silos, and conventional reactors, 16 bits is more than sufficient. Many imported brands have even been using 14-bit or 15-bit for many years.
IV. Purchase recommendations: Don’t be misled by artificially high specifications. 1. First, check the measurement accuracy of the radar itself – the “±2mm” value listed in the product manual is far more important than “24bit”. No matter how high the conversion accuracy is, it’s useless if the measurement at the source is inaccurate. 2. Check the overall compatibility of the system: If the analog input modules of the backend PLC/DCS have only 12–14 bits, it is actually unnecessary to use 24-bit modules on the frontend. 3. Be cautious of “pseudo-24bit” systems: Although some instruments have 24-bit ADCs, due to power supply noise and circuit layout constraints, the effective number of bits (ENOB) may be only 18 bits. Chips and designs that can truly achieve 24-bit stability at an industrial level are very expensive. 4. Dynamic range is more critical than resolution. In environments with strong interference (dust, steam, foam), the signal-to-noise ratio (SNR) is a better indicator of reliability than the simple number of bits.