INA180: Input and output filters

Hi Tomer,

in low-side applications a differential amplifier with a simple OPAmp can do because in a low-side application no big common-mode rejection is necessary, usually.

In a high-side application, on the other hand, a differential amplifier with a simple OPAmp will not do very probably when a high common mode rejection is necessary. Then only a current sense amplifier of the "INA" series will do. A differential amplifier with a simple OPAmp cannot provide a big common mode rejection because of the unavoidable imbalance introduced by the manufacturing tolerances, temperature drifts and long term drifts of feedback resistors. Compared to the common mode rejection of the INA180 of 100dB, e.g., a differential amplifier with a simple OPAmp will only provide 40...60dB common mode rejection. This is a huge difference.

That it is best to filter the output and not the input is not quite true. The truth is that filtering at the inputs of INA180 can introduce a highly unwanted imbalance between the inputs caused by manufacturing tolerances, temperature drifts and long term drifts of the input filtering components and can very easily degrade the awesome common mode rejection of INA180. I have seen applications where the common mode rejection was totally ruined by the choose of unsuited input filtering components.

So, it's wise to introduce only very soft common mode input filtering at the inputs of INA180 and to try to do the filtering mainly at the output of INA180. When it comes to the common mode input filtering the rule of thumb is: "Less is more".

I think that differential input filtering according to figure 9-2 will make sense in many applications. But I would be very careful when adding common mode filtering (additional caps from each input to GND). The common mode filtering caps should have a small capacitance and should be made from NP0 or C0G. And, of course, they must be as equal as ever possible.

And a low pass filter at the output of INA180 can make sense when you need a charge bucket filter for the input of an ADC.

But hey, whether filtering is neccessary for you at all depends on your application. Without knowing your application I cannot say what filtering you need.

Kai

Thank you Kai and Javier.

I think the INA180 is suitable for my application, no need for the INA240.

For the motor, the PWM frequency is 10KHz so the INA180 is more than enough.

I will follow the recommendation of filtering only the output.

For the Battery I actually use the INA181 as it is bi-directional, here I use it to measure the charging/discharging of the battery using column counting so the accuracy is much more crucial.

I have a very accurate ref voltage (using the TL431LIAIDBZR for that) so I believe this combination will give me the accuracy I require.

If you have any comments, I will appreciate it.

Best,

Tomer

Hello Tomer,

Javier is now out of the office for the holidays and will revisit this post when he returns.

Hi Tomer,

if the 10kHz PWM signal contains lots of far reaching harmonics, then the use of differential input signal filtering as shown in figure 9-2 of datasheet of INA180 can make sense.

Also keep in mind that the common mode input voltage range of INA180 allows input voltages only going down to -0.2V. So, keep the distance between the low-side shunt and the INA180 as short as possible to minimize any inducances along the shunt.

Kai