I. Overview Currently, there are two different understandings of "repeatability tests". One is a more conventional understanding, which is the ability of a weighing instrument to provide consistent results when the same load is repeatedly applied under relatively constant test conditions using the same operation method. This relatively constant test condition includes: the same weighing procedure, the same operator, using the same standard instrument under the same conditions, the same location, and repeated weighing within a short period of time. Another way of understanding it is: increase the test load from the minimum to the maximum in sequence, and then decrease the test load to the minimum in sequence. The process of repeating this test sequence multiple times and comparing the results of the same test values is called a repeatable test. Ii. Several Test Methods Related to Repeatability 1. Repeatability and Eccentric load The difference between repeatability tests and eccentric load tests lies in the fact that they are conducted under the same conditions and at the same location. In the explanation of repeatability test terms, it is very clearly stated that the same load is "under the same conditions and at the same location". The purpose of repeatability tests is to understand whether the weighing instruments under test can provide consistent results. The eccentric load test, on the other hand, involves loading the same load "at different positions of the weighing instrument's load-bearing device" to check if the indication error meets the requirements of the maximum allowable error. The purpose of the eccentric load test is to detect whether the weighing instrument under test can provide the same indication when the load is applied to different positions of the weighing instrument's load-bearing device. 2. Repeatability and Weighing The difference between repeatability tests and weighing tests lies in the "multiple loading of the same load". In the definition of repeatability test terms, it is "the same load is repeatedly applied to the same position within a short period of time." The weighing test, on the other hand, involves "gradually applying the test load from zero to the maximum weighing capacity (Max), and then gradually removing the test load in the reverse order to zero." Iii. Problem Analysis Example 1: An example of the second way of understanding can be traced back to over three decades ago. At that time, when applying for the type approval (type evaluation test) of an electronic weighing instrument with a maximum capacity of 5 tons, the inspectors conducted the "repeatability test" three times in the order of the weighing performance test, and required the difference between any two results of the same test value The absolute value of the maximum allowable error not exceeding the test value. The result was that only one of the test values exceeded the tolerance, and one was judged as unqualified. After analyzing this result, I always felt that there was something wrong. I consulted the explanations of the term "repeatability" in the R3 and R28 international recommendations issued by the International Organization of Legal Metrology at that time. In fact, this was not a repeatability performance test, but a three-time weighing performance test. As a result, the test was conducted based on the correct understanding and passed the appraisal at the first attempt. Example 2: The repeatability test of the weighing instrument load tester method is described as follows [2] : Set the truck scale and the weighing instrument load tester to zero. Apply the minimum standard load to the truck scale using one unit of the standard load unit group, then gradually apply the load to the maximum load using all units of the standard load unit group, and then unload slowly. Repeat the test three times. The difference between the maximum and minimum correction errors at any inspection point should not exceed the absolute value of the maximum allowable error under that load. Why, in the experiment of Example 1, when only 5 tons of weights were used and there were only 9 test values in each cycle, that is, 27 test values in three cycles, did the difference in the indicated values of a group of the same test values exceed the tolerance? In the test of Example 2, 6 to 8 standard load units were used to conduct three cycles of tests in sequence on the minimum weighing capacity, the maximum allowable error change weighing capacity (medium accuracy class: 500e, 2000e), and the maximum weighing capacity of the weighing instrument. The test results were basically within the absolute value range of the maximum allowable error. When conducting tests with weights, whether it is a small 20kg weight or a large 1000kg weight, each time they are loaded or unloaded manually or by equipment, the load-bearing of the weighing instrument will shake to varying degrees, thus causing some uncontrollable changes in the test readings each time. That is to say, it may cause the difference between two readings of the same test value to exceed the tolerance. When using the weighing instrument load testing instrument method, no matter how many units are used, since each time a load is applied or unloaded, it is vertically applied to the load-bearing device under the action of the hydraulic device, it will not cause significant shaking of the load-bearing device. Therefore, even if the load is applied multiple times, there will be no change in the test reading (this is relative). Even if there are variations in the test readings, this situation is far from the actual usage of the truck scale. This repetitive test is impossible to simulate the actual usage conditions. Therefore, it is recommended that for truck scales verified by the weighing instrument load test method, after completing the eccentric load performance test, weighing performance test, tare weighing performance test, and discrimination threshold test in accordance with the verification regulations, when conducting the repeatability performance test with approximately 50% of the maximum weighing capacity load [3], the test should be directly carried out on the load-bearing platform using the loading method of a heavy-duty truck. This way, it can directly simulate the actual use of the same form, and there will be no phenomenon of two skins.