Useful MyDAQ Info (Part 1): MATLAB Interfacing and Control

MATLAB and Acquisition Devices

Data acquisition devices have the potential to be directly controlled by code on your computer, code other than the software that one installs on your device. If one had the required interface drivers, one could use MATLAB to directly control a particular data acquisition device. These techniques are possible for both the Analog Discovery and myDAQ devices. With the right drivers, a range of programming languages can achieve these results.

Over the last 10 years, MATLAB has developed a data aquisition toolbox for interfacing in such cases. For the GT site licence, this toolbox should be included; in general, MATLAB does not mind selling yet another toolbox. There are other driver code approaches to interface to MATLAB, although they take alot more work. As one who has written a few drivers in the past, I can state it takes work, although it can be worth it.

The concepts are primarily focused for the MyDAQ device as it seems to be the device that most requires these concepts. The concepts should be extendable to other devices (e.g. Analog Discovery)

Some have mentioned that the MyDAQ element only seems to work with the Data Acquisition Toolbox for Windows based machines.

Link for MATLAB DAQ aquisition documentation from MATLAB.

The following material started from contributed material by Nickolas Lee (Fall 2020) as a way to use the myDAQ directly with MATLAB and be controlled by MATLAB. Having the data taken by MATLAB allows for your plots rather easy (plots are required to be a MATLAB format; Excel type plots are not accepted).

Install the Data Acquisition Toolbox. Applications installed include

  • Analog Input Recorder
  • Analog Output Generator.
Each of these applications has a library of functions.
For example for the voltage divide experiment, one could place the ai0+ and - around the second resistor. One can easily measure voltages up to 5 volts.
The Matlab code used is below. Analog Input Recorder was used to record the results. The first function will send 0 - 5 volts over 10 seconds.

d = daq("ni");
d.Rate = 8000;
n = d.Rate * 10;
output = linspace(0,5,n);
output = output';
addoutput(d,"myDAQ1","ao0","Voltage");
write(d, output);
write(d, 0);
clear d;


One can use both outputs and run them simultaneously. saveas(stackedplot(data), "name.png") saves the graph.

d = daq("ni");
d.Rate = 8000;
n = d.Rate * 10;
output1 = linspace(0,5,n);
output2 = linspace(0, 5, n);
output = [output1; output2];
output = output';
addoutput(d,"myDAQ1","ao0","Voltage");
addoutput(d, "myDAQ1", "ao1", "Voltage");
write(d, output);
write(d,[0, 0]);
clear d;

In response, we had some additional useful comments from Xinpei Ni on using he myDQA directly with MATLAB. The following material started from that material:

To read the data from MyDAQ to our computer, the following code collects and saves the data in MATLAB's timetable format.

d = daq("ni");
d.Rate = 8000;
n = d.Rate * 10;
input = linspace(0,5,n);
input = input';
addoutput(d,"myDAQ1","ao0","Voltage");
addinput(d,"myDAQ1",'ai0','Voltage');
[out, st] = readwrite(d, input);
%the data read by myDAQ from the circuit is stored in out.
write(d, 0);
clear d;

The data records the voltages to out with respect to time, not input voltage. You have the input voltage with time (input).

In response after lab, we had some additional useful comments from Christopher Saetia. The following material started from that material:

When running the code above to control the myDAQ, one might get the error: "Unrecognized function or variable 'daq'." Here is a workaround:

d = daq.createSession("ni");
d.Rate = 10;
n = d.Rate * 10;
%Leads to 100 samples being taken
output = linspace(0,5,n);
output = output';
%Outputs voltage from the ao0 port
addAnalogOutputChannel(d,'myDAQ1','ao0','Voltage');
%Collects the voltage drop data by using ai0+ and ai0-
addAnalogInputChannel(d,'myDAQ1','ai0','Voltage');
d.queueOutputData(output);
d.startForeground();
clear d;

After running these lines of code (takes about 5-10 seconds), the collected data for Vout and the supplied voltages will appear in the workspace pane under the matrices ans and output respectively.
ans: the collected Vout data. Can copy to a new variable.
Do not forget to save the data, and make a script using the lines of code used.

You might get this warning:
Warning: Session contains channels from devices that are not synchronized because digital triggering is not supported) when running the code. It seems one can ignore this error.

The error of daq function not being recognized: likely because not running the most up to date 2020 version of Matlab (the code is very recent MATLAB additions).

To collect meaningful data, use the myDAQ ports ai0+ and ai0- to measure the voltage across the resistor, and ao0 and AGND (on the ao side) to provide the voltage to the power rails.

  • For those who choose to use the MyDAC, a few more pointers (even though I don't support this device):
    • I have not seen an explicit MyDAC function for getting a DC sweep of points. It might be there, but I have not seen it and I have not had anyone else point me to it.
    • One solution is manually setting recording multiple points manually (like 40 points is around the range expected). I know this is frustrating, but is the limitation of the software controlling the device.
    • It seems that NI (who makes MyDAC) does not enable these features, but if you use LabView, their larger software package, there might be ways apparently to make something work.
    • One other way to get a sweep is to put in a triangle waveform at a fairly slow speed (like say 1Hz frequency) and measure the output waveform (and record the input waveform). If you have both parts, you should be able to get the plot of output voltage verses input voltage. I think the triangle waveform + scope is likely the better approach, but either approach is acceptable (as long as it is explained).