At this time, the displacement or error signal of the N-fold frequency is subdivided by the microcomputer clock to realize the purpose of improving the resolution and the measurement precision, thereby realizing high-precision measurement. The hardware implementation block diagram is as shown. In the middle, let the phase modulation sensor output signal frequency be: fi = fe + fv (3) where: fe - carrier frequency, fv - displacement signal frequency. A reference frequency signal f r1 ( f r1 < fe) which is similar to the carrier frequency fe is generated by the signal generating circuit. The signal fi is mixed with fr 1 and low-pass filtered to obtain a signal: f 1 = ( fe - fr 1) + Fv(4) then performs phase-locked multiplication processing on f1, and obtains the signal: f 2 = N < ( fe - f r1) + fv > ( 5) where: the multiplication factor N can be performed by the microcomputer according to the frequency of the measured displacement signal Variable presets, phase-locked multipliers are designed to increase the stability of low-frequency signals. After f 2 is mixed with the reference signal fr 2 = N ( fe - f r1) and subjected to low-pass filtering, a stable signal with only effective information such as displacement can be directly obtained: f out = Nf v( 6) vs f After the out signal is processed by the microcomputer precise clock subdivision, the mechanical displacement can be obtained. The above-mentioned subdivision processing technology is called "mixing-frequency multiplication-mixing" subdivision technology, and is abbreviated as "FM-FD-FM" (Frequency Mixing-Frequency Doubling-Frequency Mixing) subdivision technology. In order to achieve as high a precision as possible, the excitation frequency fe, the reference frequency f r1, and the reference frequency fr 2 are all generated by the same precision clock division. The literature only uses a first-stage mixing circuit to improve the measurement resolution. However, when obtaining the absolute displacement information, it is necessary to perform a large amount of processing on the subdivided sampling data, and separate the low-frequency carrier signal from the measured phase. - fr), and then the phase information representing the amount of mechanical displacement can be obtained. The method adopted is a software-implemented digital wave-limiting filter. In this process, there are both principle errors and a large amount of computation time, resulting in poor real-time performance and a compromise in measurement accuracy. Because this article adopts the full hardware "FM-FD-FM" processing, the displacement effective information is directly obtained, thereby greatly improving the real-time measurement and measurement accuracy. For the inductive synchronizer, if the clock frequency fc = 10 M Hz and the excitation frequency fe = 10 kHz, the ordinary direct subdivision measurement method is used, and the subdivision factor is: d = 1000, measurement resolution: p 1 = 3. 6 angular seconds (circular inductive synchronizer), p 2 = 2 m (linear inductive synchronizer); and using the subdivision method of mixing, taking the reference frequency fr = 9. 85 kHz, the subdivision multiple d' = 67667, The measurement resolution p 1′= 0. 05 angular seconds (angular displacement), the line displacement can reach p 2′= 0. 03 m. The resolution is improved by nearly 68 times. 2 measurement principle and system composition Take the measurement of the accuracy of the hobbing machine indexing drive chain as an example to illustrate the problem. The low speed shaft (workbench) sensor uses a circular induction synchronizer, and the high speed shaft (hob shaft) uses a 1024 line photoelectric encoder. The sampled signal is obtained by dividing the high-speed axis square wave signal. The measurement of the error is calculated using the absolute method, that is, each sampling takes the measurement starting point as the reference point of the specific phase. Let the drive chain transmission ratio be ix, the theoretical rotation angle of the rotary table in one sampling period should be = 360 1024 × ni x (7) When the Kth sampling is set, the count value of the clock is NK, and the number of cycles of the low speed axis is MK. (including non-complete cycle count), according to the principle of measurement and signal processing requirements, as shown in the developed measurement system, the measured signal timing relationship. The sampled signal is divided by the fast encoder signal n and connected to the 8098 external interrupt source Exint, causing the interrupt to execute the measurement procedure. The 20 M Hz clock signal is divided by 256 by 8-bit synchronous counter 1 and then single-stable to 8098HSI0 lead angle detection. The count of NK is completed by 8-bit synchronous counter 1 and 8155. According to the sampling signal, the 8098 sends an 8-bit synchronous counter 1 to count the control signal through the 8155I/O port, and informs the 8098 MCU to read the current value of the 8-bit synchronous counter 1 from the 8155PA port, and together with the current HSI0 port input pulse number, the resolution is 50. The counter of ns gives the NK value. At the same time, the inverted signal of the sampled signal is ANDed with the clock signal and then connected to the 8-bit synchronous counter 2. Whenever the upper edge of the slow-end signal f out appears, the counter 2 will restart counting; and when the interrupt signal arrives The clock signal on the counter 2 is zero, and the synchronization of the non-complete periodic signal count with the arrival time of the sampling signal is ensured from the hardware, thereby completing the subdivision count of the non-complete period M' of the slow-end sampling signal f out . The 8155 internal counter completes the full cycle count M" of the slow side signal, M' and M" constitute MK. 3 conclusions Through the above analysis, the following conclusions can be obtained: (1) For the phase (frequency) modulation type sensor signal, a new subdivision principle and corresponding hardware processing circuit are proposed, which is called mixing-frequency multiplication-mixing ( FM-FD-FM) Microcomputer subdivision principle. The measurement resolution can be increased by 2 to 3 orders of magnitude compared to the direct use of microcomputer subdivision. At the same time, it also greatly improves the real-time measurement, and realizes high-precision measurement of displacement and machine tool chain error and develops precision closed-loop CNC real-time compensation system, so that high-precision machining of parts will have far-reaching significance; (2 The analysis of this paper is based on the hobbing machine drive chain with two actuators as circular motion. For the measurement and real-time compensation of the precision of machine tool chain such as screw lathe, grinding machine and NC lathe, the subdivision measurement principle is also applicable; (3) A practical high-resolution counter implementation scheme suitable for high-frequency clock-filled subdivision counting has been developed, and a corresponding microcomputer measurement system has been developed, which has been used in the Y38 hobbing machine indexing drive chain error measurement to achieve high-precision measurement of errors. . According to the scheme and the idea of ​​this paper, as long as the counter of the corresponding frequency is selected, the clock source with higher frequency can be used to further improve the measurement resolution and measurement accuracy; (4) The measurement principle and system proposed in this paper can also be applied. It is used to solve the accuracy problem on similar synchronous systems and closed-loop systems. (Finish) Gate Valve,Bronze Valve,Bronze Gate Valve Texoon Brass Works Co., Ltd. , http://www.china-brassvalves.com
Machine tool effect synchronization facility high precision measurement system
1 "FM-FD-FM" technical principle For the inductive synchronizer, the excitation signal frequency is fe, the computer clock pulse frequency is fc, and the sampling direct subdivision technique, the subdivision factor is: d = fc / fe( 1 The measurement resolution is: p = / d (the distance of the inductive synchronizer) (2) It can be seen from the above (1) and (2) that there are two ways to further improve the measurement resolution: Fc or reduce fe. The increase of fc is limited by the operating frequency of the electronic device. For example, the current programmable counter has a maximum counting frequency of 10 M Hz. The carrier frequency fe of the sensor signal cannot generally be selected too low, otherwise the sensor signal is weak, the anti-interference ability is poor, and it is easy to be disordered and distorted. The inductive synchronizer signal excitation frequency fe (ie carrier frequency) is generally selected from 2 to 20 kHz, often taking 10 kHz. So how to improve its resolution is to mix the sensor high frequency modulation signal that has picked up the effective information. Frequency processing, re-transfer the effective information such as displacement and error to another frequency signal whose frequency is lower than the original carrier frequency fe, and then perform N multiplication processing and then perform secondary mixing processing to thoroughly demodulate.