外文翻译基于Buck变换器的综合实验设计.doc

基于Buck变换器的综合实验设计Buck变换器最常用的变换器，工程上常用的拓扑如正激、半桥、全桥、推挽等也属于Buck族，现以Buck变换器为例，来阐述虚拟综合性实验的操作方法设输入直流电压(VIN)：15V，输出电压(VO)：5V，输出电流(IN)：6A，输出电压纹波(Vrr)：50mV，基准电压(Vref)：1.5V，开关频率(fs)：100kHzBuck变换器主电路如图1所示，其中RC为电容的等效电阻图1 Buck变换器主电路图1.1滤波电感和电容参数设计滤波电感电流脉动估算滤波电感值滤波电容的内阻电解电容第一步首先建立交越频率fc0，在此频率总增益为0dB然后选择误差放大器的增益，迫使总开环增益在fc0为0dB下一步设计误差放大器的增益斜率，以使得总开环增益在fc0以斜率－1交越（图6.4）最后，调整幅频特性达到希望的相位裕度 采样理论指出，为了闭环的稳定，fc0必须小于开关频率的一半但必须远远小于开关频率，否则有较大幅值的开关频率纹波因此，一般经验将fc0定为开关频率的1/4~1/5样选取fc0/fz=fp/fc0fz与fp之间分开越大，在fc0有较大的相位裕度希望较大的相位裕度，但如果fz选择得太低，在100Hz低频增益比选择较高频增益低（图6.8），这样对100Hz信号衰减很差。

选取电路参数,fz=1/(2*pi*R1*C1) fp0=1/(2*pi*R0*C0) fp=1/(2*pi*R2*C2)R1=R2=800O欧姆 C1=C2=10nF上面已经指出如果误差放大器具有单极点、单零点和一个原极点，它的幅频特性如图6.11所示现在证明一个误差放大器的传递函数如何推导，以及图6.7b电路确实具有一个单极点、一个单零点和一个原极点图6.7b电路的增益为 引入复变量s=jω，于是 经过代数处理 同时因为一般C2<

PWM增益 图6.1中由误差放大器输出到电感输入电压Vy的平均值Vav的增益是PWM增益，并定义为Gm这样增益定义可能产生迷惑，直流电压Vea随误差放大器的B点的输入成比例变化，而Vy是可调宽的恒定幅值脉冲 这个增益的意义和幅值说明如下图6.1中PWM输出是直流电平Vea与0~3V(实际上是0.5~3V)三角波Vt比较产生的在所有PWM芯片中，产生两个相差180°的可调宽度脉冲在形成PWM以后，经分频并送到两个分离的输出端在正激变换器中，仅用两个输出的一个 在图6.1b中，Vea＝0,Ton=0，在Vy的宽度为零，Vy的平均值就是Vav＝（Vsp-1）(ton/T),其中Vsp是变压器次级电压，1为整流二极管压降，Vav也为零如果Vea移动到3V，在三角波的峰值，ton=0.5T，Vo＝0.5（Vsp-1）/T则调制器的直流增益为Vav与Vea的比值电解电容R1*C1=50~80微法*欧（R1为滤波电容的内阻） 取 50微法*欧△I=2*输出电流/10=2*6/10=1.2安培R1=Vpp /△I=0.05/1.2=41.6毫欧 C1=50微法*欧/41.6毫欧=1200uf瞎做的，一大拖屎堆出来的！！！！！！！！！！！！！后面的不对！！！！！！！！！！！！！！！！！！！！增设单极点、单零点的PI补偿网络     增设单极点、单零点的Pl补偿网络，有的文献称为单极点、单零点补偿网络。

图1（a）所示即为增设单极点、单零点的PI网络电路图，图中Z1（s）＝R1，z2（s）＝（1/sC∥CR2+1/sC2）符号∥表示并联于是，增设单极点、单零点PI补偿网络的传递函数为      　　除了积分器1/s产生的极点s＝0外，还有一个零点-1/TZ，一个极点-1/TP，均位于左半S平面　　图1（b）为增设单极点、单零点PI网络的幅频及相频特性，图例中，零点频率fz为505Hz，极点频率fp为50kHz 图1 增设单极点、单零点的PI补偿网络       图2为应用这种网络（其幅频特性为|Κ|）补偿后的效果调节对象为考虑电容ESR的Buck开关 图2 Buck开关电源应用单极点、单零点PI补偿网络补偿后的Bode图       转换器幅频特|G丨，含有一个ESR零点；采用增设单极点、单零点的Pi网络后，Buck开关电源的开环传递函数T的幅频特性|T0|＝较为理想低频段具有高增益，中频段以-1率穿越0dB线，保证系统稳定，高频段则快速衰减球传递函数 画波特图、仿真图About positioning error compensationController manufacturers provide a facility which allows you to remove any errors in the machine positioning system by specifying compensation values for each machine axis. Both linear and rotary positioning errors may be compensated. By applying compensation you can reduce errors to such an extent that they are almost completely nulled and the accuracy of the machine is improved considerably.Whilst this is an excellent concept, it must be understood that to obtain error compensation values, you must first be able to measure the small differences between the intended position of the moving part and its actual position, at various points along an axis. Fortunately a solution already exists; the combined use of a Renishaw laser interferometer system and the positioning error compensation software extension package.The errors that need to be measured may be considered to be small, in the order of a micrometre or so, however, the cumulative effect of such an error along an axis can be quite considerable. Using the laser to measure these errors and the compensation software to record them, a table of measured errors at points along an axis can easily be obtained. These errors can then be translated into compensation values which the controller can apply as the moving part is moved along the axis.The Renishaw positioning error compensation software is offered as an additional option to the standard Renishaw Laser10 calibration software. The software provides a 'step-by-step' user interface to guide you through the various stages of the error compensation procedure.How a machine is compensatedTo compensate a machine, you use the facilities provided by the positioning error compensation software to create a part program for the controller. This is a simple task; all you need to do is specify the target positions along the axis where you want to measure the linear error and the software creates the part program for you automatically.Serial communication facilities are then used to transfer the part program into the controller. Once it is there, you set the computer into its data capture mode and start the part program.The position of the moving part is monitored by the software via the laser interferometer. As the controller pauses at each target, the position of the moving part is captured by the software, during a suitable dwell period, bef。