INA826: INA826

Allen,

First, what is the supply voltage you plan on using?  Also please clarify apparent conflict between " +/-7mV input" AND "input is 5V CM  with a gain 1.4v, excitation voltage 10V with an offset off .5V    Gain ~750 output swing ~0- 5V." This is very difficult to understand.  I would suggest you draw the circuit (even by hand) and attached its image using INSERT button below. Please make sure to show all the conditions including Vsupply, Vcm, Vref, gain,etc.

Keep in mind that the linear input voltage range of most instrumentation amplifiers is a function of the supply voltage, input common-mode voltage, reference voltage and gain.  Therefore, the rail-to-rail output (RRO) may not be possible even when using RRO INA if you do not properly set the conditions of the application. Please download TI Engineering Analog Calculator using link below and use the Instrumentation Amplifier Output Voltage vs Common-mode Voltage function to determine linear input/output voltage range - see below.

https://www.ti.com/tool/ANALOG-ENGINEER-CALC?keyMatch=INA%20CALCULATOR

I'm unable to insert a pic, Is an email address available.

To move forward

supply: 10V

INA 826 Ref : 500mV

Vcm :5V 

the sensor and supply share the same ground 

With the pressure sensor connected to the INA, with no applied pressure. the output of the INA is 500mV +-7mV

thanks 

Hello Marek, comparing the noise of the 826 and 823, ( 21nV @  Gain 1000 vs 20 @ Gain 100), the 823 has lower noise.

For some reason, the engineering cal does not run on my computer. Can you rerun the output curves using the 823 also I don't believe the 823 is rail to rail, can you confirm if this is the case.

thanks 

Allen,

The difference of 20nV/rt-Hz vs 21V/rt-Hz is insignificant taking into account that the normal noise process variation may be as high as 20%.

As far as the output swing goes, the only difference between INA823 and INA826 is 150mV to negative rail vs 100mV, respectively, while they both can get within 150mV of positive rail.  Thus, with Vcm of 4.4V INA823 does exactly the same as INA826 - see below.

However, with Vcm of 5V the output swing of INA823 is higher 9V vs 8.6V)  allowing gain up to 1,200 for 7mV input - see below.

hello Marek, I thought the noise would scale with gain, why use different a different gain value to measure noise.

based on the data sheet, which is the better device?

What happens when the input results in a output at or below 150mV, 

thanks

Hi Allen12, 

INA826 and INA823 are very comparable, see the comparison below. 

Please let us know the design requirements, and we can simulate circuits for comparison purposes.  

What happens when the input results in a output at or below 150mV, 

You may consider to use negative supply rail, if the input differential input is below -291uV. In order to simulate the circuit, we need to know the input operating range of pressure sensing impedance, say the pressure transducer is operating at 997Ω -1kΩ range, IA's output range, gains and Vref etc.. 

Best,

Raymond

The output noise scales with gain BUT 20nV/rt-Hz and 21nV/rt-Hz noise spectral density refer to the input and NOT the output noise - see below. Vout_noise_rms=Vnoise*sq-rt(BWn)*Gain where Vnoise is the input noise spectral density and BWn is the noise bandwidth.  In terms of the noise INA823 and INA826 are virtually identical. Allowing the output to get closer than 150mV from the rail pushes the output transistor into saturation resulting in a signal distortion.