1 system structure and principle 1.1 sampling principle The system adopts the counting method instead of the phase-based principle. The advantages are: (1) no frequency multiplication processing is performed, and the corresponding hardware circuit is omitted; (2) the original sensor resolution is utilized to the utmost. If each pulse signal P2 at the low speed end is used as the sampling reference, the error value ΔK of the Kth sampling can be expressed as: 121) (NiPk(1) where: ΣkP1 is the number of high-speed end pulses P1 for this sampling; 3 For the transmission chain transmission ratio; λ1, λ2 are the number of high and low speed sensor grid lines; N1 is the high speed sensor pulse equivalent.
Continuous sampling and accumulation, that is, a set of discrete transmission error data: ikiTE0) ((2) This counting sampling is done by using the "programmable counting" inside the microcomputer.
1.2 Software subdivision The number of conventional subdivision methods is 10, mostly supported by a large number of hardware circuits, and there are various restrictions. There are not many software subdivision methods, some still need some hardware, some require too much original signal, and some have too narrow adaptation range and poor effect.
The basic principle of the subdivision method we use is to treat the two signals P1 and P2 emitted by the sensor as spatial scales. Their comparison can not involve the amount of time t. Now add a high-frequency time for the purpose of subdivision. Pulse Pt. Each pulse does not have any spatial significance until it is related to the spatial scale. Now, before and after each sampling, the program control is used to establish a one-to-one correspondence between the counted P1i and P2i as shown.
From the engineering point of view, it can be considered as to use the space ruler to instantaneously calibrate the time rule before sampling, and then use the time ruler to perform real-time subdivision of the space ruler when the P2 sampling reference signal arrives. The point of view of the calculation method is more precisely, it is some sort of interpolation. The dotted line Φ1(t) is the actual displacement space-time curve of the high-speed axis. Now, based on the measured known point Φ1(tj), a curve 1(t) approximating the variation law of Φ1(t) is fitted. Calculate the corresponding ΣkP1 value when P2k arrives. If only the space-time relationship between the two points before the sampling is extrapolated, this is the linear interpolation. If the three-point space-time relationship is used before and after the sampling time, this is the quadratic parabolic interpolation and so on. Using the n-point information before and after the sampling point, you can use n - 1 Lagrangian interpolation.
Unlike the original integer type ΣkP1, the calculated kP1 is decimal. Substituting it into equation (2) for ΔK achieves the purpose of subdivision. From this, we can see several characteristics of this subdivision method: 1 It is clear that the properties of the two pulse signals are spatial scale and time scale respectively; 2 the problem of subdivision is reduced to the problem of reading decimals; 3 the combination of subdivision and sampling It is completed by CTC together; 4 pairs of single-channel signal work can also be subdivided. At this time, the signal reference is time Pt; 5 can be flexibly selected according to the motion properties of the measured object.
A large number of experiments and actual tests have proved the correctness of this "segmentation method for fractional space-time conversion".
1.3 Data Processing After obtaining the error information, the computer turns to the data processing work. Digital filtering, error separation, various feature values, and CFTT analysis fault diagnosis are all completed by the program. Plus Chinese and English menu-style input and output, constitute a complete test system.
2 System Features and Technology Key The CFTT test system integrates testing and data processing with a computer as the core, which has a series of advantages.
2.1 Accuracy index The signal resolution derived by time pulse subdivision is:
Ifnifnti12160060129600 (angular seconds) (5) where: n1 is the high speed end speed; ft is the clock pulse frequency, its size reflects the quantization error caused by the time pulse.
The gear ratio i is also converted to the resolution at the same time as the error value measured at the high speed end is converted to the low speed end.
If n1 = 40, i = 40, ft = 1 MHz, then the resolution derived by equation (5) is 0.0216".
The test accuracy is 0.14; the change rate of the indicator value (repetition) of the repeated test is 0.89. These two better indicators prove that the system does introduce less link errors because of fewer links.
In addition, using the sensor 6-time indexing, the system integrated error was ±1.78′′.
2.2 Speed ​​Range The dynamic measurement of large gear processing machines at very low transmission speeds is considered to be a major problem in the field. The specific performance is P1P2P10P11P12P2kt0t1t2tsPt1(t)1(t)
In two ways: first is the limitation of the sensor; second is the frequency limit of the phase circuit and the recorder. When the speed of the table is slow to a certain extent, the error frequency of the transmission chain is close to the frequency of the phase signal, and the filter circuit cannot separate it. For the traditional magnetic grid tester, it is more difficult to perform when the table speed is lower than 8 minutes.
In order to solve this problem, the predecessors have done a lot of work, such as the one mentioned in [3], which uses a special magnetic grid sensor, using a stepping motor and a synchronous appliance attached to it. An additional relative motion is produced, which can be measured normally even when the speed is close to zero.
The low speed measurement of the CFTT system was carried out on a Y38 gear hobbing machine with a minimum speed of 47.5 rpm and a gear ratio of 1200 (there was no way to find a larger gear ratio hanging wheel at the site). The table speed was as low as about 25 minutes. A turn. Both the high and low speed ends use ordinary magnetic grid sensors to measure accurate and clear curves. Usually, the measurement of the large hobbing machine requires the rotation speed of the table to be about ten minutes, so the CFTT system is fully capable, which opens up a new way for the dynamic measurement of the large transmission ratio under very low speed conditions in a relatively simple way.
High speed restrictions are another difficulty. The counting frequencies of the CTC, 8253, and 8254 counter timers are 2MHz, 4MHz, and 10MHz, respectively, and there is ample margin. Therefore, the real limitation lies in the mechanism and frequency characteristics of the sensor, the frequency characteristics of the interface circuit, and the system and software can achieve. Real-time indicators. For the gear processing machine of gear hobbing machine, the high speed index is not high, and the design speed range of CFTT system is 1/25300r/min.
2.3 Real-time CFTT system uses software and computer internal hardware resources to complete sampling and subdivision. The advantage is to reduce hardware, reduce link error, cost and failure rate; and the problem is how to ensure real-time. The response time of the hardware circuit is in the nanosecond level, and the software is fast in the microsecond level. This lag in response time primarily affects the subdivisions in the sample (as is evident from the obvious), especially if real-world measurements are required. The CTC external pulse count has the real-time nature of the hardware circuit, but the CPU extracts the data from the CTC and in turn subdivides the external signal by software. If only the software method is used, the design interrupt lag time can be up to 4 T states at most, the Z804M is moderate, the maximum lag time is 1 μs; the 20 MHz clock computer lags up to 0.2 μs. For engineering requirements, the impact is negligible. This has been demonstrated by the accuracy and repeatability of the aforementioned assays.
For high-speed sampling and other situations, you must consider how to take advantage of the external triggering capabilities of the CTC.
2.4 Data Processing Functions Due to the organic addition of computers, the data processing and input functions of the test system are greatly enhanced. The whole process of testing and processing is completed by the computer, which is beneficial to the popularization and use of advanced technologies such as spectrum analysis at the production site, and solves the problem that the test and processing are disconnected for a long time, and the test data is not easily converted into computer data. From the perspective of the use effect, the spectral line is clear, and each harmonic component is accurately corresponding to each rotating link. It has a good error diagnosis function, and a variety of graphics, data and feature value output, which brings convenience to the operator.
2.5 Adaptation surface width Currently, the CFTT system uses the TP805 industrial computer, and the FMT system uses the AST586 computer. Therefore, the test system can theoretically select any computer model or match with the existing computer. The sensor can be selected from a grating, a magnetic grid or an inductive synchronizer depending on the actual situation. In addition to measuring the transmission error of the relative motion of the two ends, it can also measure the rotational non-uniformity, speed and acceleration of the single-ended output. Special gear ratios such as non-integer ratio, variable ratio and clearance ratio can also be measured. Due to the simplicity and compactness of the system, it is easier to achieve productization.
3 Conclusions The computer-aided gear transmission error intelligent detection and analysis system is a software-based test and data processing system that can realize real-time display and monitor the entire measurement process on a computer screen.
The test has high precision, convenient data processing and wide application. The test method of this system has certain reference value for the development of other computer-aided detection systems.
We also have a variety of processing plastic parts, dedicated parts processing, one-stop parts processing service manufacturers. Single pick up, to figure custom, real quotation, high-end quality. Welcome to plan custom, professional design, quotation is true, fine workmanship. Plastic parts have a wide range of applications, which can be used in automotive parts, intelligent home appliances, security home appliances, digital office, medical instruments, electronic instruments, graduation models, etc. Wholesale reasonable price of wrought iron front door quality is the best, plastic injection molding, custom plastic molding, injection molding services, precision plastic injection molding, welcome to consult.
plastic injection, custom plastic molding, moulding service, precision plastic injection molding
Sung Precision Mould & Plastic Co., Ltd. , https://www.dginjectionmolding.com