I2S comparing to Analog

[matthew-aedapt]

Hi,

I have an Analog Discovery 2 and a Digital Discovery (Great Products) which I have been using to test some designs.

I want to test a pre-amplifer with ADC board by sweeping a signal across the audio band and plotting the I2S output of the ADC chip.

I have tested the analogue circuit response using the network analyzer (so I get the feedback of the delivered analogue signal), what I want to do now is create a script that runs the network analyzer but instead of using the 2nd analogue input to plot the response, capture the I2S peak values for that frequency and plot/ save them.

Is this the right way to do this and can you give me any scripts you may have to start me in the right direction.

[attila]

Hi @matthew-aedapt

Than would be a bit complicated to do.
Alternatively to visualize analog and digital at the same time, use the Scope with Digital and a slow input signal. With tall digital row the data is also represented as analog signal and the digital data can be used in Math channels.

[matthew-aedapt]

That would work, can you send me the workspace file for that as a starting point please?

[attila]

Hi @matthew-aedapt

You will probably need to make some settings to suit your needs, such as the I2S options:
i2s.dwf3work

[matthew-aedapt]

Further,

If I wanted to save the scope.digitalchannels.I2S value as a variable in a script, how do I assign the value to a variable?

Using the Peak2Peak value of the maths channel, how can I reset it to ensure its valid for the current run?

[attila]

Hi @matthew-aedapt

The Scope.DigitalChannels.I2S.events is an array with the interpreted values from the current capture, .eventStamps the associated relative time.

[matthew-aedapt]

Hi Attila,

I found that the Scope I2S doesn't give me consistent results and have changed to the logic I2S.

This gives me nice consistent results and I can save the results and chart them using the built in logging when its set to I2S.

BUT, If I write a script to save the I2S data, I don't get the same data! 

(its a pity you can't "View" the script behind built in functions)

I want to be able to get the data the same as the logging, I can then process it for the Min, Max and Pk2Pk values I am after for each Input frequency bin.

[matthew-aedapt]

I also tried a script you previously answered someones query with.

This also doesn't give me the correct data (I removed the parseInt and division as I want the signed data.

[attila]

Hi @matthew-aedapt

It would be helpful if you could provide more information, attach a screenshot or workspace with data...

The Format can be specified in the I2S configuration:

 

This is a script for logging the value:

var rg = Logic1.Channels.I2S.events; //events
var rgt = Logic1.Channels.I2S.eventStamps; // event time stamps
var c = rg.length-1; // number of events

var i = 0;
if(rg[0].charAt(0)!="R") i++; // make sure to start with right

var file = File("~/Desktop/i2s.csv");
file.writeLine("Right,Left");

for(; i < c; i+=2){
    // trim leading R/L char
    print(rg[i].substring(2), rg[i+1].substring(2));
    file.appendLine(rg[i].substring(2)+","+rg[i+1].substring(2));
}var rg = Logic1.Channels.I2S.events; //events
var rgt = Logic1.Channels.I2S.eventStamps; // event time stamps
var c = rg.length-1; // number of events

var i = 0;
if(rg[0].charAt(0)!="R") i++; // make sure to start with right

var file = File("~/Desktop/i2s.csv");
file.writeLine("Right,Left");

for(; i < c; i+=2){
    // trim leading R/L char
    print(rg[i].substring(2), rg[i+1].substring(2));
    file.appendLine(rg[i].substring(2)+","+rg[i+1].substring(2));
}

 

[matthew-aedapt]

Thanks Attila, that helped using events rather than data.

When using the scope digital channel for I2S, selecting 2 complement didn't give the right output (The logic one does).

The negative values were being shown as there integer equivalents, So a sine wave wasn't centered around zero.

This made any attempt at Pk2Pk measuring invalid seeing as the small negative values were always over 95% value.

Hope this feedback helps.

 

Thanks for your help

 

Matthew

[attila]

On 10/31/2023 at 3:35 PM, attila said:

Hi @matthew-aedapt

It would be helpful if you could provide more information, attach a screenshot or workspace with data...