dsp.HDLFIRRateConverter

Upsample, filter, and downsample—optimized for HDL code generation

expand all in page

Description

The dsp.HDLFIRRateConverter System object™ upsamples, filters, and downsamples input signals. It is optimized for HDL code generation and operates on one sample of each channel at a time. The object implements an efficient polyphase architecture to avoid unnecessary arithmetic operations and high intermediate sample rates.

The object upsamples by an integer factor of L, applies an FIR filter, and downsamples by an integer factor of M. The object accepts and returns control signal arguments for pacing the flow of samples. For detail of the flow control interface, see Flow Control.

To resample and filter input data:

  1. Create the dsp.HDLFIRRateConverter object and set its properties.

  2. Call the object with arguments, as if it were a function.

To learn more about how System objects work, see What Are System Objects?.

Creation

Syntax

HDLFIRRC = dsp.HDLFIRRateConverter
HDLFIRRC = dsp.HDLFIRRateConverter(L,M,num)
HDLFIRRC = dsp.HDLFIRRateConverter(___,Name,Value)

Description

HDLFIRRC = dsp.HDLFIRRateConverter returns a System object, HDLFIRRC, that resamples each channel of the input. The object upsamples by an integer factor of L, applies an FIR filter, and downsamples by an integer factor of M. The default L/M is 3/2.

example

HDLFIRRC = dsp.HDLFIRRateConverter(L,M,num) sets the InterpolationFactor property to L, the DecimationFactor property to M, and the Numerator property to num.

HDLFIRRC = dsp.HDLFIRRateConverter(___,Name,Value) sets properties using one or more name-value pairs. Enclose each property name in single quotes. For example:

HDLFIRRC = dsp.HDLFIRRateConverter(L,M,Num,'ReadyPort',true,'RequestPort',true);

Properties

expand all

Unless otherwise indicated, properties are nontunable, which means you cannot change their values after calling the object. Objects lock when you call them, and the release function unlocks them.

If a property is tunable, you can change its value at any time.

For more information on changing property values, see System Design in MATLAB Using System Objects.

Upsampling factor, L, specified as a positive integer.

Downsampling factor, M, specified as a positive integer scalar.

FIR filter coefficients, specified as a vector in descending powers of z-1.

You can generate filter coefficients by using the Signal Processing Toolbox™ filter design functions, such as fir1. Design a lowpass filter with normalized cutoff frequency no greater than min(1/L,1/M). The object initializes internal filter states to zero.

Enable ready output argument of the object. When enabled, the object returns a logical scalar value, ready, when you call the object. When ready is 1 (true), the object is ready for a new input sample the next time you call it.

Enable request input argument of the object. When request is 1 (true), the object returns a new output sample on the current call to the object.

Rounding mode used for fixed-point operations. This property does not apply when the input is single or double type. 'Simplest' mode is not supported.

Overflow mode used for fixed-point operations. This property does not apply when the input is single or double type.

Data type of the FIR filter coefficients, specified as a numerictype(s,wl,fl) object with signedness, word length, and fractional length properties.

Data type of the output data samples, specified as 'Same word length as input', 'Full precision', or as a numerictype(s,wl,fl) object with signedness, word length, and fractional length properties.

Usage

Syntax

[dataOut,validOut] = HDLFIRRC(dataIn,validIn)
[dataOut,ready,validOut] = HDLFIRRC(dataIn,validIn)
[dataOut,ready,validOut] = HDLFIRRC(dataIn,validIn,request)

Description

example

[dataOut,validOut] = HDLFIRRC(dataIn,validIn) resamples dataIn according to the InterpolationFactor (L) and DecimationFactor (M) properties. To avoid dropped samples when using this syntax, apply new valid input samples, with validIn set to true, only every ceil(L/M) calls to the object. The object sets validOut to true when dataOut is a new valid sample.

example

[dataOut,ready,validOut] = HDLFIRRC(dataIn,validIn) resamples the input data and returns ready to indicate whether the object can accept a new sample on the next call.

This syntax applies when you set the ReadyPort property to true. For example: 

HDLFIRRC = dsp.HDLFIRRateConverter(...,'ReadyPort',true);
...
[dataOut,validOut,ready] = rateConverter(dataIn,validIn);

example

[dataOut,ready,validOut] = HDLFIRRC(dataIn,validIn,request) resamples the input data, indicates whether the object can accept a new sample, and, if request is true, returns the next available sample.

This syntax applies when you set the RequestPort property to true. For example: 

HDLFIRRC = dsp.HDLFIRRateConverter(...,'RequestPort',true);
...
[dataOut,validOut] = rateConverter(dataIn,validIn,request);

You can connect the ready output of a downstream object to the request input of an upstream object.

Input Arguments

expand all

Data input, specified as a scalar, or as a row vector where each element represents an independent channel.

The data can be real or complex.

double and single data types are supported for simulation, but not for HDL code generation.

Data Types: fi | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | single | double
Complex Number Support: Yes

Validity of the input data, specified as a logical scalar.

When validIn is 1 (true), the object captures the value on dataIn. You can apply a valid data sample every ceil(L/M) calls to the object. You can use the optional ready output signal to indicate when the object can accept a new sample.

Data Types: logical

Request for a new output sample, specified as a logical scalar.

When request is 1 (true), and a new output data sample is available, the object returns the sample with validOut set to 1 (true). If no sample is available, the object returns validOut set to 0 (false). The object accepts this argument when you set the RequestPort property to true.

Data Types: logical

Output Arguments

expand all

Resampled and filtered data sample, returned as a scalar, or as a vector in which each element represents an independent channel.

double and single data types are supported for simulation, but not for HDL code generation.

Data Types: fi | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | single | double
Complex Number Support: Yes

Validity of the output data, returned as a logical scalar.

When validOut is 1 (true), the data output is valid. When validOut is 0 (false), the data output is not valid.

Data Types: logical

Indicator that the object is ready for a new input sample, returned as a logical scalar.

The object returns ready = 1 (true) to indicate that the object can accept a new input data sample on the next call. The object returns this additional output when you set the ReadyPort property to true.

Data Types: logical

Object Functions

To use an object function, specify the System object as the first input argument. For example, to release system resources of a System object named obj, use this syntax:

release(obj)

expand all

stepRun System object algorithm
releaseRelease resources and allow changes to System object property values and input characteristics
resetReset internal states of System object

Examples

collapse all

Open Live Script

Convert a signal from 48 kHz to 32 kHz by using the dsp.HDLFIRRateConverter System object™.

Define the sample rate and length of the input signal, and a 2 kHz cosine waveform. Set validIn = true for every sample. 

Fs = 48e3;                  
Ns = 100;                  
t = (0:Ns-1).'/Fs;         
dataIn = cos(2*pi*2e3*t);  
validIn = true(Ns,1);

Preallocate dataOut and validOut signals for faster simulation.

dataOut = zeros(Ns,1);
validOut = false(Ns,1);

Create the System object. Configure it to perform rate conversion by a factor of 2/3, using an equiripple filter.

Numerator = firpm(70,[0 0.25 0.32 1],[1 1 0 0]);
firrc = dsp.HDLFIRRateConverter(2,3,Numerator);

Call the System object to perform the rate conversion and obtain each output sample.

Note: This object syntax runs only in R2016b or later. If you are using an earlier release, replace each call of an object with the equivalent step syntax. For example, replace myObject(x) with step(myObject,x).

for k = 1:Ns
    [dataOut(k),validOut(k)] = firrc(dataIn(k),validIn(k));
end

Because the input sample rate is higher than the output sample rate, not every member of dataOut is valid. Use validOut to extract the valid samples from dataOut.

y = dataOut(validOut);

View the input and output signals with the Logic Analyzer.

la = dsp.LogicAnalyzer('NumInputPorts',4,'SampleTime',1/Fs,'TimeSpan',Ns/Fs);
modifyDisplayChannel(la,1,'Name','dataIn','Format','Analog','Height',8)
modifyDisplayChannel(la,2,'Name','validIn')
modifyDisplayChannel(la,3,'Name','dataOut','Format','Analog','Height',8)
modifyDisplayChannel(la,4,'Name','validOut')
la(dataIn,validIn,dataOut,validOut)

Open Live Script

Convert a signal from 40 MHz to 100 MHz by using the dsp.HDLFIRRateConverter System object™. To avoid overrunning the object as the signal is upsampled, control the input rate manually.

Define the sample rate and length of the input signal, and a fixed-point cosine waveform. 

Fs = 40e6;                  
Ns = 50;                  
t = (0:Ns-1).'/Fs;         
x = fi(cos(2*pi*1.2e6*t),1,16,14);

Define the rate conversion parameters. Use an interpolation factor of 5 and a decimation factor of 2. Calculate how often the object can accept a new input sample. 

L = 5;   
M = 2;   
stepsPerInput = ceil(L/M);
numSteps = stepsPerInput*Ns;

Generate dataIn and validIn based on how often the object can accept a new sample. 

dataIn = zeros(numSteps,1,'like',x);
dataIn(1:stepsPerInput:end) = x;
validIn = false(numSteps,1);
validIn(1:stepsPerInput:end) = true;

Create the System object. Configure it to perform rate conversion using the specified factors and an equiripple FIR filter. 

Numerator = firpm(70,[0 0.15 0.25 1],[1 1 0 0]);         
rateConverter = dsp.HDLFIRRateConverter(L,M,Numerator);

Create a Logic Analyzer to capture and view the input and output signals.

la = dsp.LogicAnalyzer('NumInputPorts',4,'SampleTime',1/Fs,'TimeSpan',numSteps/Fs);
modifyDisplayChannel(la,1,'Name','dataIn','Format','Analog','Height',8)
modifyDisplayChannel(la,2,'Name','validIn')
modifyDisplayChannel(la,3,'Name','dataOut','Format','Analog','Height',8)
modifyDisplayChannel(la,4,'Name','validOut')

Call the System object to perform the rate conversion and obtain each output sample. Call the Logic Analyzer to add each sample to the waveform display.

Note: This object syntax runs only in R2016b or later. If you are using an earlier release, replace each call of an object with the equivalent step syntax. For example, replace myObject(x) with step(myObject,x).

for k = 1:numSteps
    [dataOut,validOut] = rateConverter(dataIn(k),validIn(k));    
    la(dataIn(k),validIn(k),dataOut,validOut)    
end

Open Live Script

Convert a signal from 40 MHz to 100 MHz by using the dsp.HDLFIRRateConverter System object™. Use the optional ready output signal to avoid overrunning the object as the data is upsampled. The ready signal indicates the object can accept a new data sample on the next call to the object.

Define the sample rate and length of the input signal, and a fixed-point cosine waveform. Create a SignalSource object to provide data samples on demand. 

Fs = 40e6;                                  
Ns = 50;                                    
t = (0:Ns-1).'/Fs;                          
x = fi(cos(2*pi*1.2e6*t),1,16,14);          
inputSource = dsp.SignalSource(x);

Define the rate conversion parameters. Use an interpolation factor of 5 and a decimation factor of 2. Determine the number of calls to the object needed to convert Ns samples. 

L = 5;                                      
M = 2;                                      
numSteps = floor(Ns*L/M);

Create the HDL FIR rate converter System object. Configure it to perform rate conversion using the specified factors and an equiripple FIR filter. Enable the optional ready output port.

Numerator = firpm(70,[0 0.15 0.25 1],[1 1 0 0]);     
rateConverter = dsp.HDLFIRRateConverter(L,M,Numerator,'ReadyPort',true);

Create a Logic Analyzer to capture and view the input and output signals.

la = dsp.LogicAnalyzer('NumInputPorts',5,'SampleTime',1/Fs,'TimeSpan',numSteps/Fs);
modifyDisplayChannel(la,1,'Name','dataIn','Format','Analog','Height',8)
modifyDisplayChannel(la,2,'Name','validIn')
modifyDisplayChannel(la,3,'Name','dataOut','Format','Analog','Height',8)
modifyDisplayChannel(la,4,'Name','validOut')
modifyDisplayChannel(la,5,'Name','ready')

Initialize the ready signal. The object is always ready for input data on the first call.

ready = true;

Call the System object to perform the rate conversion and obtain each output sample. Apply a new input sample when the object indicates it is ready. Otherwise, set validIn to false.

Note: This object syntax runs only in R2016b or later. If you are using an earlier release, replace each call of an object with the equivalent step syntax. For example, replace myObject(x) with step(myObject,x).

for k = 1:numSteps
    if ready
        dataIn = inputSource();
    end
    validIn = ready;
    [dataOut,validOut,ready] = rateConverter(dataIn,validIn);
    la(dataIn,validIn,dataOut,validOut,ready)
end

Open Live Script

Convert a signal from 40 MHz to 100 MHz by using the dsp.HDLFIRRateConverter System object™. Use the optional request input signal to control the output data rate. When the object receives the request signal, it returns a new output data sample. This example models a clock rate of 200 MHz by requesting an output sample every second call to the object.

Define the sample rate and length of the input signal, and a fixed-point cosine waveform. Create a SignalSource object to provide data samples on demand. 

Fs = 40e6;                                  
Ns = 50;                                    
t = (0:Ns-1).'/Fs;                          
x = fi(cos(2*pi*1.2e6*t),1,16,14);          
inputSource = dsp.SignalSource(x);

Define the rate conversion parameters. Use an interpolation factor of 5 and a decimation factor of 2. Determine the number of calls to the object needed to convert Ns samples. 

L = 5;                                      
M = 2;                 
numSteps = floor(2*Ns*L/M);

Create the HDL FIR rate converter System object. Configure it to perform rate conversion using the specified factors and an equiripple FIR filter. Enable the optional ready and request ports.

Numerator = firpm(70,[0 0.15 0.25 1],[1 1 0 0]);      
rateConverter = dsp.HDLFIRRateConverter(L,M,Numerator,...
                                        'ReadyPort',true,...
                                        'RequestPort',true);

Create a Logic Analyzer to capture and view the input and output signals.

la = dsp.LogicAnalyzer('NumInputPorts',6,'SampleTime',1/Fs,'TimeSpan',numSteps/Fs);
modifyDisplayChannel(la,1,'Name','dataIn','Format','Analog','Height',8)
modifyDisplayChannel(la,2,'Name','validIn')
modifyDisplayChannel(la,3,'Name','request')
modifyDisplayChannel(la,4,'Name','dataOut','Format','Analog','Height',8)
modifyDisplayChannel(la,5,'Name','validOut')
modifyDisplayChannel(la,6,'Name','ready')

Generate a signal that requests a new output sample every second call to the object. 

request = false(numSteps,1);
request(1:2:end) = true;

Initialize the ready signal. The object is always ready for input data on the first call.

ready = true;

Call the System object to perform the rate conversion and obtain each output sample. Apply a new input sample when the object indicates it is ready. Otherwise, set validIn to false. The object returns valid output samples when request is set to true.

Note: This object syntax runs only in R2016b or later. If you are using an earlier release, replace each call of an object with the equivalent step syntax. For example, replace myObject(x) with step(myObject,x).

for k = 1:numSteps  
    if ready
        dataIn = inputSource();
    end
    validIn = ready;
    [dataOut,validOut,ready] = rateConverter(dataIn,validIn,request(k));
    la(dataIn,validIn,request(k),dataOut,validOut,ready)    
end

Open Script

Create a rate conversion function targeted for HDL code generation, and a test bench to exercise it. The function converts a signal from 40 MHz to 100 MHz. To avoid overrunning the object, the test bench manually controls the input rate.

Define the sample rate and length of the input signal, and a fixed-point cosine waveform.

Fs = 40e6;
Ns = 50;
t = (0:Ns-1).'/Fs;
x = fi(cos(2*pi*1.2e6*t), 1, 16, 14);

Define the rate conversion parameters. Use an interpolation factor of 5 and a decimation factor of 2. Calculate how often the object can accept a new data sample.

L = 5;
M = 2;
stepsPerInput = ceil(L/M);
numSteps = stepsPerInput*Ns;

Generate dataIn and validIn based on how often the object can accept a new sample.

dataIn = zeros(numSteps,1,'like',x);
dataIn(1:stepsPerInput:end) = x;
validIn = false(numSteps,1);
validIn(1:stepsPerInput:end) = true;

Create a Logic Analyzer to capture and view the input and output signals.

la = dsp.LogicAnalyzer('NumInputPorts',4,'SampleTime',1/Fs,'TimeSpan',numSteps/Fs);
modifyDisplayChannel(la,1,'Name','dataIn','Format','Analog','Height',8)
modifyDisplayChannel(la,2,'Name','validIn')
modifyDisplayChannel(la,3,'Name','dataOut','Format','Analog','Height',8)
modifyDisplayChannel(la,4,'Name','validOut')

Write a function that creates and calls the System object.

Note: This object syntax runs only in R2016b or later. If you are using an earlier release, replace each call of an object with the equivalent step syntax. For example, replace myObject(x) with step(myObject,x).

function [dataOut,validOut] = HDLFIRRC5_2(dataIn,validIn)
%HDLFIRRC5_2
% Processes one sample of data using the dsp.HDLFIRRateConverter System
% object. dataIn is a fixed-point scalar value. validIn is a logical scalar value.
% You can generate HDL code from this function.

  persistent firrc5_2;
  if isempty(firrc5_2)  
    Numerator = firpm(70,[0,.15,.25,1],[1,1,0,0]);   
    firrc5_2 = dsp.HDLFIRRateConverter(5,2,Numerator);
  end    
  [dataOut,validOut] = firrc5_2(dataIn,validIn);
end

Resample the signal by calling the function for each data sample.

for k = 1:numSteps
    [dataOut,validOut] = HDLFIRRC5_2(dataIn(k),validIn(k));
    la(dataIn(k),validIn(k),dataOut,validOut)
end

Algorithms

expand all

This object implements the algorithms described on the FIR Rate Conversion HDL Optimized block reference page.

See Also

Blocks

Objects

Introduced in R2015b

DSP System Toolbox Documentation

Support