Hi,
I am using the IC to sense the charge current of a battery, but if by mistake the battery is inverted the IC is damaged.
I can add two resistors to the measurement input of the same value to avoid the short circuit.
Or what do you recommend?
Hey Reinaldo,
I am assuming that you want to prevent the common-mode voltage of the INA138-Q1 to not go below -0.3V, which would happen if the battery is placed incorrectly with the reverse polarity.
The very high currents are definitely a problem and could be limited with some current-limiting resistors placed in the input traces into the IN+ and IN- pins of the INA138-Q1 device; however, placing these resistors (even if around 10Ohms) will starts producing error in the INA138 measurements. We as a rule limit these resistors to nothing greater than 10 Ohms, but if you are willing to sacrifice accuracy you could increase this value, see Option 3.
Option 1 – Use a single diode in the power path
During normal operation you will lose power supplied by the battery according to the forward voltage drop of the diode used. If the battery is incorrectly placed, then diode is reverse biased and will open up the circuit. A Schottky diode would work best here because it could provide the lowest forward voltage drop during normal operation, but you will want to make sure that it can withstand the reverse bias voltage (the battery’s voltage).
Option 2 – Use PMOS FET in the power path
http://www.ti.com/jp/lit/an/slva139/slva139.pdf
Option 3 – Use 2 normal rectifying shunt diodes off IN+ and IN- inputs
This is a straightforward approach where you place two 10-Ohm, current-limiting resistors and then shunt diodes from the IN-/IN+ pin to ground (left circuit) or to the voltage supply (right circuit). The common-mode voltage will be clamped to {VDD - Vf} where Vf is diode foward voltage drop. The problem is the diode current will be high (~5A) if battery is misplaced. It would be up to your power voltage supply (INA138 VCC) to supply this current.
If you connect the anodes to ground (left circuit), then it is up to the battery to supply this current. With this diode biasing, the common-mode voltage will become clamped to the forward diode drop. You would need to find a diode with the lowest possible diode drop (<0.3V), but at the same time a reverse-bias max voltage greater than your battery’s voltage.
The secondary problem with this high current is that your current-limiting resistors (R4, R3) will need to be able to dissipate that power. If these resistors are increased to 50 Ohms, the diode current drops to ~1 A and it seems there are resistors that can handle this power dissipation. Although I am not sure how much accuracy you need to retain for your system. The error produced by input resistors (R4 = R10 = Rin) is due to the input offset current (Ios) generating a voltage offset on top of you shunt voltage signal. This offset error is:
Ios*Rin = (IB+ - IB-)*Rin.
As your shunt voltage increases so does Ios because, the IB+ and IB- curves usually diverge from one another. We do not have this characterized for INA138 as we do with other devices unfortunately (see INA240 datahsheet), but it can be measured with an ammeter in series with the inputs, while the input voltage is increased.
Option 4 – Use Zener and Rectifying diodes
This relies on the fact that Vz and Vd are chosen carefully. The common-mode will be clamped to VCC – Vd – Vz. At the 5A running through the diodes, Vz = 4.189V and Vd = 0.785, so with the power supply at 5V, the VCM becomes 26mV (within INA138 CM absolute maximum rating).
Option 5 – Implement a comparator that will turn a MOSFET off if reverse polarity detected
You can read more about this in a former post here:
Hope this helps.
Peter Iliya
Current Sensing Applications Support
