main content

analyze circuit simulation data -凯发k8网页登录

analyze circuit simulation data

since r2021a

description

the mixed-signal analyzer app enables you to visualize, analyze, and identify trends in mixed-signal simulation data. with the cadence® virtuoso ade-matlab® integration option you can import databases of circuit-level simulation results in matlab. to gain insights into the data, you can plot trends where you can vary different process parameters and see how the system behavior changes. you can compare the simulation results between different simulation runs and save and export your results.

open the mixed signal analyzer app

  • matlab toolstrip: in the apps tab, under signal processing and communications, click the app icon.

  • matlab command prompt: enter mixedsignalanalyzer.

examples

you can export the cadence® simulation run results to a .mat file after an interactive run result is created. the .mat file can then be imported to the mixed-signal analyzer app. to export the data, launch matlab® from cadence ade by clicking the m button in the toolbar.

the matlab session that opens should have adeinfo in the base workspace.

run the function adeinfo2msa. since no name-value arguments are specified, the function creates a .mat file with only the metrics data from the latest cadence simulation run. the .mat file is saved in the present working directory.

once the .mat file is generated, the app automatically launches with the imported .mat file.

you can also specify the simulation run type, run name, and test name. use the name-value pair argument import2msa to determine whether to launch the app with the generated .mat file or just to save the .mat file in the present working directory. use the name-value pair argument metricsonly to import only the simulation metrics data for faster performance.

you can directly import the cadence® database in the mixed signal analyzer app without explicitly creating the .mat file.

to export the data, launch matlab® from cadence ade by clicking the m button in the toolbar.

the matlab session that opens should have adeinfo in the base workspace.

launch the mixed signal analyzer app.

select import > adeinfo database … from the app toolbar.

in the newly opened dialog box, select the interactive runs you want to import to the app. by default, the app imports only the metrics data for better performance. you can deselect the metrics only option to import all simulation data including waveforms.

importdialogbox.png

this example shows how you can use the mixed signal analyzer app to analyze a clock buffer circuit and understand the effect of varying corner points using trend charts. you can also update the analyses with modified simulation data and export the results to a file.

export data from cadence®

the output is setup for the simulation in the cadence® ade assembler maestro view:

maestroview.png

in this setup one node (/in) is probed for waveforms. there is one expression to generate the metrics data (delay_in_o2). to generate a .mat file at the end of the simulation run, a matlab® expression is added that calls the adeinfo2msa function in this format:

adeinfo2msa('metricsonly',false,'import2msa',false,'filename','clockbuffer1.mat')

altogether 16 cases (2 sweeps of 8 corners) of simulation runs are performed. once simulation finishes in cadence, the generated .mat file is saved in the present working directory.

import data to mixed-signal analyzer

open the mixed signal analyzer app from the app gallery or matlab command prompt.

>> mixedsignalanalyzer

to import the .mat file containing the cadence simulation data, click the import button in the app toolbar, select file..., and then select clockbuffer1.

the transient and ac analysis simulation data, analysis waveform, and performance metrics shows up in the data panel.

plot and analyze data

to plot the transient waveform, click on /o2 under the tran section in the data panel, then click the display waveform button in the analysis tab.

tran_o2_waveform.png

you can plot specific cases of the waveforms currently in focus. for example, to filter out the 1.2v of vdd, click the filter button in the plot options panel and deselect 1.2 v the newly opened dialog box. click the ok button to update the waveforms. the plot now shows the 8 out of 16 cases of waveforms representing waveforms for 0.9v cases.

filteredwaveforms.png

to find the overshoot values for the /o2 waveforms, keep /o2 selected and select the ymaximum function from the built-in analysis section in the analysis tab. the calculated metrics are added under analysis metrics in the data panel.

add custom analysis

you can also add your custom analysis function using the add analysis button.

to add a custom analysis function that computes the mean of the slew rates of the transient output waveform, keep /tran/o2 selected click the add analysis button. in the newly opened pop-up window, select enter matlab expression. enter the expression to evaluate as mean(slewrate(y,x)) and click ok.

add_custom_analysis.png

this produces the mean of the slew rate data of the output transient waveform under the analysis metrics section in the data panel.

mean_slew_rate.png

you can create multiple custom matlab expressions or analysis functions and save them for future use.

plot trend charts

to get better insight about certain parameters, you can add a trend chart by clicking trend chart button in the metrics tab. for example, to find the trend in the delay metrics data, select delay_in_o2 under metrics section in the data panel and click the trend chart button.

trendchartdelay.png

the trend chart shows the delay between the output (/o2) and the input (/in) signals as various process corners are varied. you can modify and add new fields to the trend chart.

from the plot options panel, select cormodelspec and vdd in the trend chart fields. the fields are added to the x-axis layers.

trendchartdelay2.png

you can move parameters inside the x-axis box using the arrow buttons to change the sequence in which they are placed. you can also move parameters from x-axis to legends and vice-versa until you see a trend emerging. at that point, you can draw conclusion about the metrics data and its effect on various design parameters and process corners.

trendchartdelay3.png

update data with modified design simulation

these analyses performed on waveform and metrics data help in making design decisions for the circuit you are working on . you can go back to cadence® after you are satisfied with the analysis results to make changes to your design and perform another simulation run. after importing the modified simulation data, use the update button in the mixed-signal analyzer app to refresh all the plots and figures in the current working session. you do not have to re-configure your trend chart, compute analyzed metrics, waveforms and perform filtering on the waveforms the next time around when you generated the next set of simulation results for the same design.

for example, you can have a second set of simulation results extracted from cadence® and saved under a .mat file named clockbuffer2.mat. now, in a mixed-signal analyzer app working session, where you have all the plots and figures from the first simulation run present, select update > file... and select clockbuffer2.mat.

a new dialog box opens asking you to select the data to be refreshed. in this case, interactive.128 data (from clockbuffer1.mat file) is refreshed/updated with interactive.129 (from clockbuffer2.mat file) data.

click the refresh button to update the waveforms and trend chart with the data from the new simulation run.

updateapp.png

export to reports

once you are satisfied with the results, you can export it as a report in either ppt, pdf, doc, or html file format. you can also rename each plot for your convenience while generating the report using the plot > rename plot option from the app toolbar. you can also select the format, name, and location of the report file. by default, the report is saved in the maestro/documents folder of the design.

programmatic use

mixedsignalanalyzer opens a new blank mixed-signal analyzer app.

mixedsignalanalyzer('') starts the mixed-signal analyzer app, and loads it with the simulation data from the file. the app supports .csv, .xlsx, and .mat files.

more about

version history

introduced in r2021a

see also

| | | | | | | |

网站地图