Non linearity with XTR115

Hi Collin Wells

           Thanks for replaying, here i will attach the readings which i got practically. And that 2-wire and 4-wire concept is only for our prototype purpose,but i m using only 2-wire here i m attaching the latest circuit diagram.
and our circuit total current consumption is around 3.5mA when all required peripherals of micro-controller is ON,that current we are passing through IRET pin,
 but the thing is we are not using Vreg(+5V) from XTR instead we are using voltage regulators to generate supply voltage for other circuitry.

please let me know for any further clarification,

regards
Zameer(systems Engg)

Data File :-http://e2e.ti.com/cfs-file.ashx/__key/CommunityServe<span style="font-size:small;"><span>r-Discussions-Components-Files/14/2870.CurrentReadings.xlsx

Hello,

I took a look at the data file and plotted the non-linearity over the span of the measurement.  It can be seen below:

Based on the shape of the curve alone I am not sure what is causing the errors, but I do have a few suggestions and questions. 

1.)  What type of output does your pressure sensor provide and what is the output drive strength?  Is it a buffered voltage or a resistive bridge output?
2.)  Do the non-linearity issues go away when your external regulators are not connected and drawing the full 3.5mA?  I have had success in the past using external regulators instead of the built-in regulators but special care was taken to ensure proper routing and also to ensure that the current of the regulators never caused the zero-scale value to rise above 4mA.  It seems like this is working based on the 0% error at your zero-scale reading.
3.)  What is the linearity of your pressure sensor?

 Thanks for the additional information.

Regards,
Collin Wells
Precision Linear Applications

Hi Collin wells,

1) Basically the sensor we are using is Capacitive sensor.The output of sensor is in differential form.

the output will range from 18mV-to-152mV and i m not using any buffer between Microcontroller and sensor,but directly giving sensor output to differential channels of internal ADC.

2) For this i bypassed the whole microcontroller section from XTR circuit and i gave external analog voltage from precession voltage supplier from 0V-to-2.5V but still there is non linearity in the output current.

and as said its working based on 0% error at zero-scale reading ie 4mA and full scale reading ie 20mA.

3) sensor is working with maximum non linearity of less than 0.2%. 

Please let me know for additional information.

regards

zameer(systems engg)

Hi Zameer,

Thanks for the additional information. 

1.)  In your final application what actually drives the input to the XTR?  Is it a DAC in the microcontroller?

2.)  To clarify, for the most recent testing did you still have the external regulators sinking current into the IRET pin or was this part of the circuit removed as well?  Without any other circuitry in the system, if you apply an ideal current to the input of the XTR115 you should receive the same nonlinearity performance that is specified in the datasheet.  Since we specify a max non-linearity spec this informs me that every device is tested to meet this performance before it is shipped, so the devices you've received should perform as indicated in the datasheet. 

That said, unless there are still other parts of the circuit that are active that could be interfering with the XTR circuit operation we should dig a little into your test setup.  Is it possible that your precision source isn't as linear as you think it may be?  Do you read the source input voltage/current with a precision multi-meter when you perform your non-linearity calculations or are you using the number set on the source? 

Also, would you mind attaching the circuit schematic as a .PDF or as an image file?  The images uploaded to the forum directly are difficult to read sometimes and I can not read the component values in your circuit. 

I'm sorry for all of the questions but I'm still not sure what's causing errors in your circuit performance.       

Regards,
Collin Wells
Precision Linear

Hi Collin

1) ,Yes we are directly giving DAC output from microcontroller to  LIN pin of XTR. we are not using any buffer between this.

2),The whole circuits including Microcontroller,external regulators sinking current into the IRET pin, and as i said earlier  i bypassed the whole microcontroller section from XTR circuit and i gave external analog voltage from precession voltage supplier from 0V-to-2.5V but still there is non linearity in the output current. in this i have not removed the other components.

but one thing i found is that schematic which i have sent to you for that we made one PCB board(only XTR part) then i applied voltage externally from 0.1 to 2.5. i measure the output current  nonlinearity will be around less than 0.1%. no other circuits ,no load.

one thing we conclude here is that either XTR115 is not taking external load or our circuit layout is not proper.

and one more thing the XTR115 IC we are using is 3-4years back we ordered that,this got nothing to do with this nonlinearity right.

i m running out of time collins this month end is my project submission.

regards

Zameer(Systems Engg) 

Hi Zameer,

Unfortunately I still can not read your schematics.  Would you please upload them as a file instead of as an image?  The "Upload File" button is right next to the "Upload Image" button on the toolbar. 

Although I doubt the age of the XTR115 is the cause of the issues you are experiencing, if it was not stored properly it is possible that some of the specs may have degraded.  Do you have any "new" XTR115 devices to test with?

I'm not really sure I understand what you've stated in the text below:

2),The whole circuits including Microcontroller,external regulators sinking current into the IRET pin, and as i said earlier  i bypassed the whole microcontroller section from XTR circuit and i gave external analog voltage from precession voltage supplier from 0V-to-2.5V but still there is non linearity in the output current. in this i have not removed the other components.

but one thing i found is that schematic which i have sent to you for that we made one PCB board(only XTR part) then i applied voltage externally from 0.1 to 2.5. i measure the output current  nonlinearity will be around less than 0.1%. no other circuits ,no load.

one thing we conclude here is that either XTR115 is not taking external load or our circuit layout is not proper.

Here is my understanding, please correct me if I misunderstood. 

First you removed power or connections from the other circuits on the board and tested the XTR with a 0-2.5V signal.  In this first test you still see the non-linearity.

For the second test you began with a new PCB and only installed the XTR115.  You tested this circuit with the same input signal and found that everything seems to work correctly and the non-linearity drops significantly to a more appropriate value.

Is this correct?

If so, it sounds like the external circuitry is causing issues with the XTR device.  You must ensure that you keep the current that flows into the IRET pin under around 3.3mA to ensure the XTR can properly regulate the load.  What is the typical load current of the external circuitry, and it is possible that it is increasing and causing issues?

If you provide a .PDF or other file form of your schematics then we may be able to find the issue in them.

Regards,
Collin Wells
Precision Linear Applications 

Hi Zameer,

One of my colleagues with great deal of experience with these applications offered some additional advice. 

He was suspicious of the potentiometers in the circuit because some poor quality potentiometers can have contact resistance issues on the wiper that can manifest as a nonlinearity.  His suggestion was to measure the potentiometer values and replace them with discrete fixed resistors of the closest value.  The result will be that the endpoints are not as exact as they were with the trim pots, but it would allow you to test to see if the the linearity improves.  This would be a good test to verify that the potentiometers are not affecting the linearity performance of the circuit.

Best Regards,
Collin Wells