Page tree
Skip to end of metadata
Go to start of metadata

Where To Find This Example

AWR Version 14

This example was removed in V14.  

AWR Version 13

Understanding AWR .emz Files

Design Notes

Linear Regression Measurements in AWR

AWR software has equations for linear regression (LGR), linear deviation (LDV), and linear accumulation (LGA). This project shows an example of how to get these values using output equations. Each output equation has a brief explanation of what it does. The help for output equations will go into more details on each specific output equation.

Results

The graphs in the project show the results of the output equations and shows they are doing what the old LGR, LDV, and LGA measurements used to do. In the original LGR, LDV, and LGA measurements, there were settings for the lower and upper limits of the data to do the regression analysis. The same thing is done in the output equations by using the subsweep equation to pull out only the data range the user requires.

In this project ripple is calculated a couple different ways. The first way is by taking the simulated data at a specific value and subtracting the average of that data. The second way is by taking the simulated data at a specific value and subtracting the linear regression of the data over the specified range. These two values will be the same as long as the linear regression is a horizontal line. To see the difference, please open up the tuner and tune on the fstop variable. The fstop variable changes the upper limit of the data used to process the deviation parameters. You will see that when the fstop goes outside of the passband of the filter, the average value and the linear regression of the specified data range are now different.

In this project the tuning is very quick but this is not because it is a very simple circuit. The tuning is just changing how already simulated data is being displayed. Because the tuner is only changing variables in the output equations without any change to the circuit, new simulations are not performed during tuning; the updated output equations are applied to the same set of simulation results. Therefore, if you have a very long simulation, you could still use this technique to do real time tuning to change how your data gets displayed.

Schematic - BPF

Graph - LDA

Graph - LDV

Graph - LGR