Introduction With the increasing demands of industrial automation and the needs of modern industrial development, weighing systems are being used more and more in industrial process control, especially in sensor modules. For example, in the petrochemical industry, many process control applications involve weighing large silos and tanks for non-trade settlement purposes, particularly in situations where flow meters are often replaced by weighing. However, calibrating these systems presents many practical difficulties, as these large silos and tanks are often difficult to load. Common methods include substitution calibration, which involves using water instead of weights to weigh the water – a very cumbersome process. Furthermore, for some massive weighing systems, such as those weighing thousands of tons, the cost of calibrating from the start with weights is extremely high. Even using loading methods, calibration typically only reaches about 5% of full load, compromising accuracy, and in some cases, it's impossible to load at all (e.g., sealed tanks). Therefore, a market demand has emerged for calibration methods that do not require loading. From the user's perspective, this would greatly simplify the process, making system installation and commissioning much easier, significantly reducing calibration costs and improving efficiency. Secondly, from a business and technological advancement perspective, companies that lack calibration-free technology are at a disadvantage in bidding, while manufacturers that can provide users with calibration-free functionality clearly have a competitive advantage. This load-free calibration method can be called electronic calibration, which uses a high-precision simulator to simulate the actual output of the sensor under load. This application has existed before, but its accuracy was very low. The calibration method introduced in this article studies the signal attenuation of various corresponding links in the system under different conditions by measuring or calculating, and how to more accurately simulate the output of the instrument when the real system input is set. This achieves the goal of a relatively precise calibration system without loading, and can achieve an accuracy of one-thousandth when the mechanical error is small. Due to national metrology laws, this calibration-free method is inconvenient for the calibration of scales involved in trade settlement. At the same time, since this load-free calibration cannot eliminate the mechanical error of the scale, it has high practical value in situations where the mechanical error of the weighing platform is small and the accuracy requirement is not very high, and where loading is difficult or impossible. 1. Calibration Principle and Weighing System Composition 1.1 Calibration Principle: Where Wci is the weight value output by the i-th sensor after calibration, Xi is the output value of the i-th sensor, Woi is the output weight value of the i-th sensor when the weighing platform is empty, Fci is the angle difference coefficient of the i-th sensor, Fs is the range coefficient, and Wct is the system output weight value after calibration. If the system is composed of analog sensors, all sensors in the system are connected together through a junction box and input to the instrument. The system output value X is the initial measurement value. To obtain the accurate weight value Wct after calibration, the range coefficient and the output weight value of the empty weighing platform must be obtained (the angle difference coefficient is considered to be 1 here). Calibration is required to obtain these two coefficients. During calibration, the output of the empty weighing platform is recorded first, and then a known weight is added. The resulting system output value can be used to calculate the system's range coefficient, which is the most critical parameter. The system is then calibrated, and the system output value will be the specific weight value in subsequent weighing conditions. 1.2 System Structure The system composition is shown in the diagram below. This calibration-free process mainly consists of sensor modules, junction boxes, instruments, computers, simulators, cables, digital multimeters, etc. As analyzed in the calibration principle section above, we know that as long as the system's range coefficient and output under an empty scale are known, the system can be calibrated. Even without a load, the entire system can be calibrated as long as the system's range coefficient is known. The principle of load-free calibration for digital sensor systems is to calculate the system's range coefficient based on the sensor's sensitivity; for analog sensor systems, it is to simulate the sensor's output under load using a sensor simulator and calculate the system's range coefficient. 2. Load-Free Calibration Implementation Scheme From the above analysis, it is clear that the key to load-free calibration is determining the system's range coefficient. Knowing the range coefficient allows for load-free calibration. The system will be divided into digital and analog systems for discussion. 2.1 Digital System Implementation Scheme Generally, load-free calibration cannot adjust the angle difference, as adjusting the angle difference is equivalent to adjusting the sensitivity of each sensor. Since the magnitude of sensitivity adjustment cannot be known for analog systems, angle difference adjustment is not performed. If a digital system wants to account for the angle difference effect during calibration without loading, it can adjust the angle difference using a weight of unknown weight. For example: The angle difference coefficient is used to balance the weight of each sensor. If the angle difference coefficient of each sensor is 1, the output of the entire system is the sum of the outputs of all sensors. Angle difference adjustment can be achieved by testing the total output of the system at each angle and then calculating the angle difference coefficient for each angle. The process is illustrated below with an example. Using a 30-ton digital truck scale, six 22.5-ton digital sensors (rated output of 100,000 internal code), and a T800 instrument, the angle difference adjustment process is as follows: 1) Load weights at each angle and read the total output value of the instrument at each angle. If the angle difference effect is not considered, the following are the influencing factors for calibration without loading in a digital system: 2.2 Influencing factors of analog systems: 3. Step-by-step loading calibration method: In some cases, high calibration accuracy is required, which necessitates considering the correction of mechanical errors. Calibration without loading cannot meet this requirement. Here is a method that uses small weights and substitutes for accurate system calibration. 3.1 Option 1 ● First, calibrate using a small weight according to the general calibration method to obtain the first range coefficient a1 for application from zero load to a certain weight (the weight of the small weight). ● Add a substitute to a certain weight (the displayed weight is approximately the weight of the small weight; the range coefficient at this point is a1), then add another small weight and recalibrate to obtain the second range coefficient a2 for application from the weight of the small weight to twice the weight of the small weight. ● Add a substitute to a displayed weight of approximately twice the weight of the small weight, then add another small weight and recalibrate to obtain the second range coefficient a3 for application from the weight of the small weight to three times the weight of the small weight. ● Continue in this manner until a certain percentage of the total capacity of the weighing system is reached, resulting in a series of range coefficients. Determine the appropriate range coefficient in the instrument as needed, or use different range coefficients for different weights. Note: Excessive calibration may cause cumulative errors due to weight errors. 3.2 Scheme Two ● First, calibrate the instrument using the standard calibration method with small weights to obtain the first range coefficient a1 from zero load to 1/3 of the system capacity. ● Add a substitute to approximately 1/3 of the system capacity, then add small weights and recalibrate to obtain the second range coefficient a2 from 1/3 of the system capacity to 2/3 of the capacity. ● Add a substitute to approximately 2/3 of the system capacity, then add small weights and recalibrate to obtain the third range coefficient a3 from 2/3 of the system capacity to full scale. ● Calculate the range coefficients using a segmented range coefficient method in the instrument. 4. In conclusion, load-free calibration is an electronic calibration method that replaces weight calibration when it is impossible to perform it on-site. However, it cannot eliminate mechanical errors. Its accuracy depends crucially on accurately simulating the output of the entire system under sensor loading conditions. This requires calculating signal losses under various conditions, clarifying the overall system configuration, the characteristics of each sensor (sensitivity, capacity, 6-wire or 4-wire system, etc.), the functions of each instrument (gradients and graduation value settings, etc.), signal losses at the junction box and on the cables, and the interrelationships between them. The step-load calibration method introduced in this paper is a practical calibration method that can largely eliminate mechanical errors when large weights cannot be loaded on-site, and it has high practical value.