Is it possible to use an INA134 or INA2134 (INA137/INA2137) with a single supply?
Either 5V, 12V, or 24V.
What would be the common mode and differential mode input voltage range if run from a single supply?
Example circuits/schematics are welcome. Thanks!
Hi Jim,
It is possible to use the INA134 family single-supply with some basic considerations for the circuit design. To start, 5V is below the minimum supply voltage for the INA134 family which is given in the datasheet as +/-4V, this is a total supply voltage of 8V, so an 8V single supply would be the minimum for the part. To use the part single-supply you can apply a voltage to the reference pin equal to VCC/2 which will bias the output mid-supply. However, this DC voltage will also appear at the inputs of the part and so AC coupling capacitors will be necessary to prevent feeding DC back to your signal source. Here is an example:
The buffer op amp on the reference pin is an absolute necessity as the voltage at this pin must be from a low-impedance source to prevent gain errors and degradation of the CMRR of the INA134. The 1 megaohm resistors on the inputs are only there to keep my spice simulator happy (Tima-TI) as spice likes all nodes to have a DC path to ground. One final detail that I should note is that the INA134 is only able to swing about 2V from either rail so if you expect large input signals your power supply voltage will need to be high enough to accommodate this.
Jim,
You are correct that the AC coupling capacitors will affect the CMRR of the circuit. With 10uF capacitors I simulated a 20% mismatch in capacitance value and it caused an approximately 30dB decrease in CMRR at 60 Hz:
Increasing the capacitance value to 47uF resulted in a 16dB degradation in CMRR at 60Hz due to a 20% mismatch. The degradation will occur at low frequencies where the capacitors' differing impedances are a factor. At high frequencies both capacitors are essentially short circuits so it does not degrade CMRR. The trend here is that larger capacitor values cause less of a degradation in CMRR due to their mismatch, however there are other aspects of the circuit that should be considered. First, it would be good to include a discharge pathway for these capacitors so that they do not cause pops during plugging and un-plugging of an audio cable. Also, all though it is often overlooked, a major degradation in common-mode rejection is actually the difference in the impedances of the conductors of the cable being connected. Consider that if the input capacitors were perfectly matched to each other, a 1 ohm mismatch in the resistance of the conductors of the input cable would cause a 10dB degradation of CMRR when using a difference amplifier based line receiver circuit such as the INA134.
The sensitivity of a circuit to this effect is directly determined by its input common-mode impedance. Higher common-mode impedance results in better "real world" common mode rejection. This paper describes this effect in detail and has some unique ways of dealing with this effect: http://www.eetimes.com/design/audio-design/4015777/Design-of-High-Performance-Balanced-Audio-Interfaces--Part-1
