INA138: INA138

Hi Mirko,

your circuit with the INA138 is only unipolar working and can properly measure only one half wave of the sine. During the other half wave the output of INA138 goes into saturation and shows a non valid, unspecified output signal. R5//C17 takes the average of both the proper output signal and the non valid output signal. That's why you see such a high drift of measuring signal.

Kai

Hello Mirko,

Thanks for choosing to use Texas Instruments in your design.  Kai raises a valid point as the INA138 is a unidirectional device that will only measure the forward current.  If you want to measure a bidirectional signal, you will need to use a setup like in the figure below (also would consider putting a buffer on the output of the INA138 if it is going to a SAR ADC).  Alternatively you could replace the INA138 with a similar part that was designed for bidirectional measurements, which is the INA170.

Hello Patrick,
Thank You for answer. But is using INA138 for sensing bidirectional currents the reason for the temperature drift of the Output Signal ?

Best regards
Mirko

Hello Kai,

Thank You for answer. But in the datasheet of INA138 (8.2.4) bipolar current measurement is described without mentioned a temperature drift.

Best reagards Mirko

Hi Mirko,

your circuit cannot sense a bidirectional current! The circuit Patrick proposed can do that.

Kai

Hi Kai,

i know, that my circuit cannot sense bidirectional current. In my application i only want to know the avarage value of rectified Output current ( therefor R5/C17) and for this the circuit works.
Do You realy think, that using INA138 in this way will cause a big temperature drift ?

Best regards
Mirko

Yes, of course!

Kai

Hi Kai,

i believe , that i have found a solution to prevent the temperatur drift of the output. When i switch a Diode LL4148 between Output of INA138
(Pin 1) and R5||C17 then the drift disappears.

Best regards
Mirko