I have sensor datapoints over time, see graphic below. I need to derive a relationship between the measurement of an external variable and the current, taking into account the transient effect of the sensor. How can I calibrate the sensor taking into account the transient time of the circuit? I presume that I need to use the bias and full cell voltage for this. Any help/pointers would be greatly appreciated, thanks!

 

I have sensor datapoints over time, see graphic below. I need to derive a relationship between the measurement of an external variable and the current, taking into account the transient effect of the sensor. How can I calibrate the sensor taking into account the transient time of the circuit? I presume that I need to use the bias and full cell voltage for this. Any help/pointers would be greatly appreciated, thanks!

View attachment 332349

Hello,

This would come under the general category of curve fitting. Check into that for more information.

 

What is shown is the plot of four variables over time. To determine any relationship you need to know the relationship between the independent variable and the dependent variable. That is what is missing.

 

Does anyone know why bias voltage in a sensor (pictured above) might produce the 4 parallel horizontal lines over time, indicating it fluctuates rapidly between 4 discrete values? Could this be due to digitisation/quantisation of the circuit? And is the bias voltage the same as what is being supplied to power the circuit?

 

Does anyone know why bias voltage in a sensor (pictured above) might produce the 4 parallel horizontal lines over time, indicating it fluctuates rapidly between 4 discrete values? Could this be due to digitisation/quantisation of the circuit? And is the bias voltage the same as what is being supplied to power the circuit?

Hi,

Yes, it does look like a digitization due to the limited resolution of some internal analog to digital converter. Either that or the measurement system has limited ADC resolution. Using any ADC will always result in a range of discrete levels. The minimum difference between levels would be the resolution.

 

A more detailed explanation will probably result in much more useful responses.
In addition, a plot of the result versus that "external variable" will allow an understanding of the relationship.
In addition, the plot of "bias voltage versus time", with apparently four different values at the same time, is not possible. Thus it indicates a flaw in the data acquisition scheme. It is not possible for a voltage to have four different values at the same time, is it???

 

A more detailed explanation will probably result in much more useful responses.
In addition, a plot of the result versus that "external variable" will allow an understanding of the relationship.
In addition, the plot of "bias voltage versus time", with apparently four different values at the same time, is not possible. Thus it indicates a flaw in the data acquisition scheme. It is not possible for a voltage to have four different values at the same time, is it???

If you zoom in, you see it is a scatter plot over long timescale where the voltage fluctuates randomly between these 4 different values. Sorry I should have made that more clear!

 

We are given no hint about the time scale, is it in seconds or weeks?? And I repeat: a plot of the result versus that "external variable" will allow an understanding of the relationship.

 

[MODERATION]

Welcome to AAC.

As a courtesy, your thread has been moved to the Homework Help forum. AAC rules limit the sort of help members may give with coursework related questions to guidance in the student’s own solution. When posting it is important to show your attempt at a solution and ask specific quiestions.

In the future, please post directly to the Homework Help forum. Thank you.