当前位置:网站首页>[basic knowledge] ~ hard core / soft core / solid core, pwm/spwm, Fibonacci sequence, large end mode storage, Fourier transform, Nyquist sampling law, chip selection, Kirchhoff law, fir/iir filter

[basic knowledge] ~ hard core / soft core / solid core, pwm/spwm, Fibonacci sequence, large end mode storage, Fourier transform, Nyquist sampling law, chip selection, Kirchhoff law, fir/iir filter

2022-06-09 11:24:00 AI is very good

1. hardcore / Soft nucleus / Solid core

hardcore (Hard IP Core) : The hard core is EDA Design domain refers to the verified design layout ; Specific in FPGA Design refers to fixed layout and process 、 Front end and back end validated design , Designers can't modify it . There are two reasons why it can't be modified : First of all, the system design has strict requirements on the timing of each module , It is not allowed to disrupt the existing physical layout ; The second is the requirement of protecting intellectual property rights , Designers are not allowed to make any changes to it .IP The feature of hard core can not be modified makes it difficult to reuse , So it can only be used for some specific applications , The scope of application is narrow .

Soft nucleus (Soft IP Core) : The soft core is EDA Design domain refers to the register transfer level before synthesis (RTL) Model ; Specific in FPGA Design refers to the hardware language description of the circuit , Including logical description 、 Net list and help document, etc . Soft core only through functional simulation , You need to go through synthesis and layout to use . Its advantage is high flexibility 、 Strong portability , Allow users to self configure ; The disadvantage is the low predictability of the module , There is a possibility of error in subsequent design , There are certain design risks . The soft core is IP The most widely used form of nuclear .

Solid core (Firm IP Core) : The solid core is EDA The design domain refers to a net list with planning information ; Specific in FPGA Design can be seen as a soft core with layout planning , Usually, the RTL The mixed form of code and corresponding specific process net list is provided . take RTL This paper describes the comprehensive optimization design combined with specific standard cell library , Form gate level netlist , Then you can use the layout and routing tool . Compared to the soft core , The design flexibility of the solid core is slightly poor , But there is a big improvement in reliability . at present , So is the solid core IP One of the mainstream forms of nuclear .

2. PWM/SPWM

PWM, English name Pulse Width Modulation, Short for pulse width modulation , It is by modulating the width of a series of pulses , Equivalent to the required waveform ( Including shape and amplitude ), Digitally code the analog signal level , That is to say, adjust the signal by adjusting the change of duty cycle 、 Changes in energy, etc , Duty cycle means in a cycle , The time when the signal is at high level accounts for the percentage of the whole signal period , For example, the duty cycle of square wave is 50%.

SPWM, English name Sinusoidal PWM, Pulse width changes according to sine law and is equivalent to sine wave PWM wave form .

3. Big end mode storage

3.1 summary

Big end model (Big-endian), The high byte of data , Stored in a low address in memory , And the lower bytes of data , Stored in a high address in memory , This kind of storage mode is a bit similar to processing data in string order : The address increases from small to large , And the data goes from high to low ;

The so-called small end mode (Little-endian), The high byte of data is stored in the high address of memory , The low byte of the data is stored in the low address of the , This storage mode combines the height of the address with the data bits Effective combination of power , The weight of high address part is high , The weight of low address part is low , Consistent with our logical approach ;

3.2 There are different sizes ?

Because in the computer system , We are in bytes , Each address corresponds to a byte , A byte is 8bit. But in C In language, except 8bit Of char outside , also 16bit Of short type ,32bit Of long type ( It depends on the compiler ), in addition , For digits greater than 8 Bit processor , for example 16 Bits or 32 Bit processor , Because the register width is larger than one byte , So there must be a problem of how to arrange multiple bytes . So it leads to big end storage mode and small end storage mode . That we use a lot X86 The structure is small end mode , and KEIL C51 It's the big end mode . A great deal of ARM,DSP It's all small end mode . There are some ARM The processor can also choose the big end mode or the small end mode by the hardware .

4. Fibonacci sequence

The Fibonacci sequence refers to a sequence that looks like this :0、1、1、2、3、5、8、13、21、34、……
This sequence starts at number one 3 A start , Each of these terms is equal to the sum of the first two terms .

5. The Fourier transform

Fourier transformation , It means that a function satisfying certain conditions can be expressed as a trigonometric function ( Sine and / Or cosine function ) Or a linear combination of their integrals . In different areas of research , Fourier transform has many different variants , Such as Continuous Fourier transform and discrete Fourier transform .

6. Nyquist's sampling law

Nyquist sampling theorem means that if the bandwidth is limited , To recover the original signal without distortion from the sampled signal , The sampling frequency shall be greater than 2 Times the maximum frequency of the signal .

The sampling frequency is less than 2 Times the highest frequency of the spectrum , The spectrum of the signal is aliased .
The sampling frequency is greater than 2 Times the highest frequency of the spectrum , There is no aliasing in the spectrum of the signal .

7. Kirchhoff's law

Kirchhoff's law includes current law and voltage law :

Current law : In a lumped circuit , At any instant , The sum of the currents flowing to a node is equal to the sum of the currents flowing from the node .

The law of voltage : In a lumped circuit , At any moment , Go round any circuit in the circuit , In this circuit, the sum of electromotive force is equal to the sum of voltage drops on each resistor .

8. Chip selection

First Basic functions, of course . The function must meet your needs .
secondly Performance needs to be considered , For example, signal-to-noise ratio 、 bandwidth 、 transmission speed 、 storage capacity 、 Working voltage 、 Quiescent current loss, etc , It depends on the function of your chip . Of course, it is necessary to consider both performance and cost .
Besides , We also need to consider the packaging form of the chip 、 Operating temperature range 、ESD Protection level 、 Lead free, etc .
Next Ask from the perspective of production process “ Is this chip solderable ?” For example, it is impossible to weld in a processing plant with a low technological level / test BGA Is a package of components .
Next Ask from the perspective of supply chain management “ Is this chip easy to buy ? Is there a stable and reliable source channel ?” If you choose a chip after a lot of hard work 、 But I found that I couldn't buy it all over the world , That would be useless .

9. FIR/IIR filter

9.1 summary

FIR( Finite impulse response ) filter : Non recursive ( There is no feedback channel ), With linear phase .
IIR( Infinite impulse response ) filter : Recursive structure ( Including feedback channel ), Nonlinear phase .

9.2 Difference between the two

Same order FIR and IIR filter ,IIR The filter has good filtering effect , But it will produce phase distortion .FIR The filter has stable performance , But the order ratio required for the same amplitude index IIR higher 5~10 times , Cost is very high . stay DSP In design ,FIR Than IIR Need more parameters , That is to say, more calculations are needed , More resources are consumed and more computing time is required , Yes DSP The real-time performance of .

9.3 FIR filter

Yes N Samples are weighted and averaged ( Convolution ) Handle .

The treatment process is expressed by the following formula :
 Insert picture description here
chart :( The triangle symbol in the figure means delay )
 Insert picture description here

9.4 IIR filter

Contains both recursive and non recursive parts . One IIR The filter can be seen as consisting of two FIR Filter composition . One of the filters is located in the feedback loop , Design IIR The key of the filter is to ensure the stability of the recursive part .

The treatment process is expressed by the following formula :( have N Feedforward coefficients and M-1 A feedback coefficient )
 Insert picture description here
chart :
 Insert picture description here
FIR filter , Maybe I'm based on FPGA When the digital signal processing system needs filtering , The first thing that comes to mind . Because it is very convenient . About the use of filters ,xilinx The official provided FIR Compiler IP Core to call . The traditional design method is , Use matlab Of FDATool Tools to design filters , And export the tap file , It's the one in the picture above W(0)—W(N-1) These weights , And in vivado Call in FIR IP, Configure import tap file . If using system generator It will be more convenient to design , stay Sysgen We can put FDATool and FIR Compiler Make an association , You do not need to manually export or import the taps FIR Design of filter .

原网站

版权声明
本文为[AI is very good]所创,转载请带上原文链接,感谢
https://yzsam.com/2022/160/202206091040370863.html