We are using an AD8207 I/O amplifier in our DC Monitor device that is already
produced, built and sold for three years now. We are currently looking into a
redesign of that device. One main issue is the protection against Transients on
the input of the AD8207 as we are using the device on a very noisy surrounding
on vehicles. We have placed two Transient Voltage Suppressor Diodes (SMBJ5V0)
on PIN 1 and PIN 8 (so the inputs) and connected them to ground. PIN 1 and PIN
8 also see a capacitor of 1 uF connected from PIN 1 to PIN 8. Both pins also
have a 4k7 ohms resistor in line. We then connect that up to a 50 mV shunt.
The reference voltages are both set to 2.048 V so we can measure positive as
well as negative current.
The output is then fed into an ADC.
Do you see any problem with our circuit ("AD8207_schematics.jpg")? Do the
transient suppressors change the input voltage coming from the shunt?
In a previous design we had ferrites at the input as a filter (see
"AD8207_schematicsB.jpg"). L5 and L6 are the ferrite beads. Do you see any
problem in using this filter? I would then also add the two transient
suppressor diodes to ground.
Also I was wondering about the fact that we are using the AD8207 on a vehicle
and therefore having the shunt in the negative lead of the battery. Do you see
any problems with that? Are there difference amplifier chips that are designed
to go on to the negative side?
We found that it's very common to have the shunt on the negative side of the
battery on vehicles. What advantages does it have to put the shunt on the
negative side other than it's mainly one connection that goes to (-) and that
all the wires to the shunt are on ground so that any contact with the grounded
body of the vehicle doesn't cause a short. Could we use the AD8207 on both, on
the positive and on the negative side of the battery?
The 4.7 kOhm resistors would affect the circuit in a couple of ways.
The first thing you would do is create a resistor divider between the 4.7k and
the input impedance of the AD8207. The AD8207 has a differential input
impedance of 240kOhms +/-20%, so you would introduce a predictable gain error
of -3.76% with an unpredictable gain error of +/- 0.7%.
The second thing you would do is introduce a degradation in CMRR. The extent of
the degradation depends on how well the two 4.7kOhm resistors match. If they
are perfectly matched, then no CMRR error is introduced.
Given these errors, it is best to minimize the resistance of the series
resistors. We have seen EMI and OVP filters designed with as little as 10 Ohm
The attached schematic using inductors does not look like it would cause any
errors when connected across a shunt.
Regarding measuring a shunt on the negative rail :
It is common to monitor the current on the ground side because the common mode
voltage of the shunt will be ground referenced. This ensures that the input
common mode range of the current shunt amplifier does not need much range,
enabling a large selection of inexpensive parts. If one is required to monitor
current on the high side (12V for example, with some transients going up to
40V) then the selection is much more limited. Most of current sense amplifiers
work on the low side, but the real value of parts like the AD8207 is their
ability to monitor the high side, where it is required.