3BSE022366R1机器人模块
模拟处理器
简介使用Microchip的TC500A模拟处理器和TC520 16位控制器设计了一台15kg称重秤。要求天平分辨率降至1/8克，并校正至61/2克以内。该项目考虑了功能天平的各个方面：•动态范围•应变仪补偿•归零•过采样•单位转换（千克到磅）TC500A是一种模拟处理器设备，执行双斜率模数转换功能。转换的所有计数和计时必须由外部源控制。在几乎所有的应用中，该控制源都是微处理器。微处理器被编程为监控TC500A的状态并控制其时序。还必须对其进行编程，以计算转换结果。作者：Ted Dabney，Microchip Technology，股份有限公司。TC500A没有自己的议程，因此可以用于生成慢速、高分辨率转换或快速、低分辨率转换。准确度的折衷是每毫秒积分时间大约1000个计数，即16位，TINT大约等于60mS。通常，总转换时间约为积分时间的4倍，但对于TC500A，这是相当灵活的。TINT=CINT RINT VINT/VIN（max）eq1 TC520是一种数字接口设备，可用于替换微处理器执行的所有TC500A计时和计数功能。TC520可以使用晶体或外部时钟作为时基来控制TC500或TC500A的操作。电平移位控制逻辑模拟开关控制信号2 16 VIN–模拟公共线+VREF+–BUF CAZ缓冲积分器SWR–SWIZ Comp 1 Comp 2 Comp输出14 GND 15 11 5 10 7 9 8 6 4 3 1 12 13 B控制逻辑SW1 TC500/TC500A CREF CREF SWR CREF CAZ RINT CINT CINT SWRI+SWRI+SWRI–SWRI SWZ SWI VIN+SWZ VS–VS 相位解码逻辑极性检测–+A B转换器状态0 0零积分器输出0 1自动零1 0信号积分1 1去积分VREF+SWI A AN780 DS00780A第2页©2002 Microchip Technology，股份有限公司。
16位转换结果与极性位和超量程位一起累积在TC520中
这些位形成一个18位串行字，可以以任何速率和任何时间读取。从TC520读取串行数据不会影响TC500A/TC520转换周期，但在读取过程中输出移位寄存器不会更新。使用TC500A和TC520输入级开发天平应用低信号电平源的首要考虑是输入放大器所需的增益量。TC500A有一个CMOS输入缓冲器，由于单位增益相位裕度，其必须不低于约68kΩ 用于RINT。最大缓冲电流VIN（max）/RINT）应不超过约20µA。这意味着TC500A的最大输入电压应约为1.5V。用于此应用的15kG应变计的输出约为1mV/g，其增益要求至少为50。MCP606 CMOS运算放大器因其低噪声和最小漂移而最适合于此。应变计的输出阻抗仅为300Ω 因此单端配置是足够的。实际满量程输出约为1V，而不是1.5V。RINT的值设置为130kΩ, 远高于68kΩ 最低限度这使最大缓冲电流为7.6µa，而不是20µa。积分器级TC500A积分器的信噪比是带宽的函数。15kG标度需要以至少8:1的过采样分辨率解析1g。这意味着至少有120000个计数。上述“每毫秒1000次计数”规则要求TC500A的积分时间至少为120ms。选择200mS将降低带宽并获得50/60Hz的最大抑制。应变计是一个平衡电桥，因此输出将有一些共模分量。VINT的值为3.5V而不是4V将允许一些偏移。重新排列等式1给出了CINT的表达式：CINT=VIN（最大值）TINT/VINT RINT=1V 200mS/3.5V 130k=.439mF eq2下一个更高的通用值是为CINT选择的.47µF。对于非常低的介电吸收，该电容器必须是聚丙烯型。参考电压电路差分参考电压通过标准双斜率比技术得出：VREF=VIN（最大）TINT/TDEINT eq3，其中TDEINT是满量程转换所需的解积分时间
Analog Processor
A 15kg weighing scale is designed using TC500A analog processor of Microchip and TC520 16 bit controller. The balance resolution is required to be reduced to 1/8g and corrected to within 61/2g. This project considers all aspects of the functional balance: • Dynamic range • Strain gauge compensation • Zeroing • Over sampling • Unit conversion (kg to lb) TC500A is an analog processor device that performs dual slope analog-to-digital conversion. All counting and timing of transformations must be controlled by an external source. In almost all applications, the control source is a microprocessor. The microprocessor is programmed to monitor the status of TC500A and control its timing. It must also be programmed to calculate the conversion results. Author: Ted Dabney, Microchip Technology, Inc. The TC500A has no agenda of its own, so it can be used to generate slow, high resolution transformations or fast, low resolution transformations. The compromise of accuracy is about 1000 counts per millisecond integration time, that is, 16 bits. TINT is about 60mS. Generally, the total conversion time is about 4 times of the integration time, but for the TC500A, this is quite flexible. TINT=CINT RINT VIN/VIN (max) eq1 TC520 is a digital interface device that can be used to replace all TC500A timing and counting functions performed by the microprocessor. The TC520 can use a crystal or an external clock as the time base to control the operation of the TC500 or TC500A. Level shift control logic Analog switch control signal 2 16 VIN – analog common line+VREF+– BUF CAZ buffer integrator SWR – SWIZ Comp 1 Comp 2 Comp output 14 GND 15 11 5 7 9 8 6 4 3 12 13 B control logic SW1 TC500/TC500A CREF CREF SWR CREF CAZ RINT CINT SWRI+SWRI+SWRI – SWRI SWZ SWI VIN+SWZ VS – VS phase decoding logic polarity detection –+A B converter status 0 0 Zero integrator output 0 1 Automatic zero 1 0 signal integration 1 1 Integral VREF+SWI A AN780 DS00780A Page 2 © 2002 Microchip Technology, Inc.
The 16 bit conversion result is accumulated in TC520 together with the polarity bit and the overrange bit
These bits form an 18 bit serial word that can be read at any rate and at any time. Reading serial data from TC520 will not affect the TC500A/TC520 conversion cycle, but the output shift register will not be updated during reading. Using TC500A and TC520 input stages to develop the balance and apply low signal level source, the first consideration is to input the gain required by the amplifier. The TC500A has a CMOS input buffer, which must not be less than about 68k Ω for RINT due to the unit gain phase margin. The maximum buffer current VIN (max)/RINT) should not exceed about 20 µ A. This means that the maximum input voltage of TC500A should be about 1.5V. The 15kG strain gauge used for this application has an output of about 1mV/g and a gain of at least 50. MCP606 CMOS operational amplifier is most suitable for this because of its low noise and minimum drift. The output impedance of the strain gauge is only 300 Ω, so a single ended configuration is sufficient. The actual full scale output is about 1V instead of 1.5V. The RINT value is set to 130k Ω, much higher than the minimum of 68k Ω, which makes the maximum buffer current 7.6 µ a instead of 20 µ a. The signal-to-noise ratio of the integrator level TC500A integrator is a function of bandwidth. The 15kG scale needs to resolve 1g with an oversampling resolution of at least 8:1. This means at least 120000 counts. The above "1000 counts per millisecond" rule requires that the integration time of TC500A is at least 120ms. Selecting 200mS will reduce the bandwidth and obtain a maximum rejection of 50/60Hz. The strain gauge is a balanced bridge, so the output will have some common mode components. A VIN value of 3.5V instead of 4V will allow some offset. Rearranging equation 1 gives the expression of CINT: CINT=VIN (maximum) TINT/VIN RINT=1V 200mS/3.5V 130k= The next higher general value of 439mF eq2 is selected for CINT 47µF。 For very low dielectric absorption, the capacitor must be polypropylene type. The differential reference voltage of the reference voltage circuit is obtained through the standard double slope ratio technology: VREF=VIN (maximum) TINT/TDEINT eq3, where TDEINT is the integration time required for full scale conversion
