XTR115 beter temperature stable

Hello Marien,

Here are some answers to your questions:

1.)  Is there additional data available about the temperature drift behavoir?
Is it a repetive curve or rather random?

For a single device, the drift curve should be repetitive over multiple temperature sweeps.  However, the drift curve of every device will be different.  For example the initial offset of a device will typically be between -100uV and +100uV, the drift will typically be between +/-0.7uV/C.  Based on that information every device will have a slightly different intial offset and a slightly different slope to the offset drift. 

2.)  Typical the device will suite my requirements, But wat is the spreading between
several devices?

When a specification in a datasheet has both a typical and maximum value, a few things can be interepreted.  First, the typical spec represents roughly a +/-1sigma probability.  Therefore, there is roughly a 68% likelyhood that a given device will be within the typical specs.  The max spec represents the maximum worst case for the spec and roughly a  +/-3sigma probability.  Therefore, there is a 99.7% probability that the device will perform better than the max spec.  Devices that do not fall within the max spec are discarded and are not released.

That said, how do you intend to improve the drift of this circuit?  You would have to characterize every unit first to determine the initial offset and drift that it will have.  Then you would have to design an analog method (diode or Nichrome wire) that cancels the drift of the amplifier which may not be easy. 

Just to be clear the XTR115 is a 2-wire transmitter meant for remote sensor applications (an example of this can be seen in Figure1), where the sensor and XTR are some distance away from the loop supply and receiver.  Is this what you're looking for or are you looking for a 3-wire transmitter that would be local to the control system to accurately push a 4-20mA signal to control other systems. 

Please let me know more details about your goals and we will work to find a solution.

Regards,
Collin Wells
Precision Analog

Hi Marien,

Unfortunately the slope of the drift curve for every device will be different.  Most of the parts will fall within the typical spec of +/- 0.7uV/C for the offset drift.  This means that most parts will have an offset drift with a slope somewhere between -0.7uV/C and +0.7uV/C.  As you can see from the range for the offset drift, some parts will have positive slopes, some will have negative slopes, and some also may not experience much drift at all.  Therefore without characterizing every device there will be no way to cancel out the drift.  The same applies for the slope of the gain error drift curve which will also be different for each device. 

Please let me know how I can help further.

Regards,
Collin Wells
Precision Analog