AD627 Common Mode Range calculator page

Hello,

I'm working on a weight measurement connected device.

I use an AD627 powered at 5v in single supply (final project will be powered with 3AAA).

The strain gauge is a full bridge (1k ohm for each resistor witch give a differential voltage for about 1mV/V at full scale) and I put the middle points of each half bridge in the AD627 inputs. The result is a 2.5v common mode voltage and a differential mode around mV.

My RG resistor is a 220 ohm. Important : I put 0v to VREF.

The receiving box on the strain gauge provide a differential input voltage of 0.9mV.

In summary, we have Vref = 0 ; Vcm = 2.5v ; Vdiff = 0.0009v ; Vcc = 5v ; Rg=0.2k

In this conditions, the calculator page (Instrumentation Amplifiers : AD627 Common-Mode Range / Gain Calculator | Design Center | Analog Devices) return an Out of range most of the time for my Vdiff comprise in [0.9mV ; 5.3mV]

However, my AD627 give me an output voltage ranging from close to 0V to a close to 5v (which is coherent for a rail to rail OpAmp). My weight system is very good in term of linearity and it cover the range I was looking for.

Then, is there somebody to give me an explanation about the 3.5V limitation for each OpAmp taken by this calculator page ?

Thanks for every coming help.

Eric

  • 0
    •  Analog Employees 
    on Jun 13, 2016 10:53 PM

    Hi Eric,

    It seems that you have found a bug in that old tool. But we have built a new and much improved tool for this purpose. See this link with your circuit conditions already entered:

    Instrumentation Amplifier Diamond Plot Tool | Analog Devices

    The voltages listed in the "Internal Circuitry" section of the new tool match the AD627 datasheet equations in the "Input range limitations in single supply applications" section. The voltage out of the first op amp when you have 0.9mV differential is 3.65V, well-within its range. And clearly the inputs and the output of the AD627 are in range too.

    I would like to point out the non-ideality though. If your bridge has an offset of just -0.9mV or +0.2mV, then your AD627 will saturate. I have not seen a bridge sensor with zero offset tolerance quite this good. The typical solution is to reduce the gain and add some level-shift voltage to the REF pin in order to leave room for offsets and make sure your system won't be saturated even for the worst case values. In cases where that's not possible, there are some other ideas: How to Stay Out of Deep Water When Designing with Bridge Sensors

    Best regards,

    Scott

  • Hi Scott,

    thank you for your advices. I discovered and tested the Diamond Plot Tool yesterday and saw what you write but didn't get the time to post a comment. This is a very nice tool wich inspire me a new exercice for my student. I will illustrate it with your diamond tool and Analog instrumentation amplifier.

    The second part a your message is important but not a problem in my design because the strain gauge is already charged with mechanicals stuffs that cause output exceding +0.2mV before the real charge take is place.

    Best Regards,

    Eric