"Accuracy" describes the level of accuracy of a physical quantity, reflecting the error between the measured value and the true value, while "resolution" describes the scale division, reflecting the smallest change that can be read during the numerical reading process.
To illustrate simply: Imagine a common ruler with a 10-centimeter range and 100 graduations, capable of displaying a minimum effective value of 1 millimeter. We can say this ruler has a resolution of 1 millimeter, and it can only read values from 1 to 100. Its actual precision is unknown, because a 2-millimeter reading doesn't reveal the error between this and the true, absolute 2 millimeters. However, if we heat it with fire and stretch it, then examine it again, we'll find that it still has 100 graduations, so its "resolution" remains 1 millimeter—the same as before! However, its precision has clearly changed.
For encoders , "resolution" is not only related to the number of lines, but also changes due to the influence of electrical signals. It is adjustable and controllable. It can change with the subdivision of the signal. The higher the subdivision factor, the smaller the resolution. However, the higher the subdivision factor, the greater the error introduced.
Precision, on the other hand, is more related to mechanical aspects. Once a product is manufactured, its precision is basically fixed (some high-precision products can improve precision by compensating for signals). This value is obtained through testing and is closely related to the overall performance of the product, such as its workmanship and materials. It is difficult for us to obtain a specific value as a basis for precision through calculation. We can mostly judge the quality of precision during use.
For example, for a 13-bit encoder, the absolute position count on the code disk is 8192. Therefore, the calculated resolution is 158 arcseconds. This means that when reading values, the jump between values must be 158 arcseconds. If the first value to be read is 0, the second value must be greater than 158; if it's less than 158, a smaller resolution is needed. When reading the value 158, due to errors, an absolute 158 seconds cannot be obtained. The error between the 158 seconds read by the encoder and the absolute true 158 seconds depends on the precision. Therefore, precision is discussed based on resolution.
It's not true that the finer the subdivision, the better. Subdivision introduces and amplifies errors, and excessive subdivision will compromise accuracy! The required and achievable subdivision level must be determined within the bounds of maintaining accuracy. High subdivision is irresponsible because accuracy is not visible before use. Higher code disk quality, better engraving, and better signal quality result in smaller errors after subdivision. This is influenced by the overall performance of the encoder, which explains why encoders with the same parameters can be from different brands and at different price points.
For example, if we want to read values 1, 2, 4, 7, and 8, we must choose a resolution of at least one unit. Choosing a resolution of two units is obviously not possible, because if we read the value 1, we cannot read 2. The error between the 1 we read and the true absolute value of 1, based on a resolution of one unit, is the precision. CNC systems on machine tools have resolution settings for linear encoders. If the interval between the values to be read is less than the resolution, the machine tool may vibrate or malfunction.
For absolute encoders with incremental signals, it is possible to accurately keep the absolute position value and incremental value synchronized in serial transmission. The absolute value corresponds exactly to an incremental signal, and the position value is always within the sine cycle of an incremental signal.
For example, with a 13-bit absolute encoder, a 512-line incremental signal, and an absolute position interval of 158 seconds, it's not suitable to read the position between two encoder positions. However, we can subdivide the 1Vpp incremental signal, such as by a factor of 100. This is equivalent to introducing several subdivided positions between the two absolute positions. We can then calculate the number of incremental pulses between the subdivided absolute positions to read the position value between the two absolute positions. For instance, with 512 lines subdivided by 100, absolute position 1 has a value of 0, and absolute position 2 has a value of 158. The position between these two positions can be read by adding one pulse to position 1 (value 0), resulting in 25 pulses. Adding two pulses would result in 25 x 2 = 50 pulses…
However, absolute encoders with incremental signals do not have subdivision capabilities, which requires users to perform subdivision processing on the incremental signals themselves.
