Hello

my application is weight scale .

as you know about load cell , With 2-mV/V sensitivity and 5-V excitation, the full-scale output voltage is 10 mV.

and in order to use the most linear portion of the load

cell’s span, only about two-thirds of this range would be used.

So the full scale output voltage would thus be about 6 mV.

for example for AD7799 (something that i'm using currently ) the full range is 39 mv(with maximum gain).

So I just use 6mv out of 39mv .

Is there any solution to use full scale ?

Is it good idea to pre amp (by low noise op-amp) ?

Best Regards

Hello,

as I already said: a preamplifier may improve your dynamic range, but only if the benefit of signal gain is not overcompensated by the drawback of additional noise generated by the preampf. And only a detailed noise analysis will tell you, if you improve or deteriorate your effective resolution.

I won't exercise a complete noise analysis for this problem, but I try to give you some rough estimation.

The 1,1nV/sqrt(Hz) of the opa2211 are excellent, but for youre application not the noise floor at high frequencies is relevant, but the 1/f noise at low frequencies (Fig 1 of OPA2211 datasheet). At least I assume, that with a DC-driven weigh scale you are interested in the low frequencies. The low frequency noise of the OPA2211 is well described by Fig. 7 of the datasheet. You read there, that the voltage noise in the range 0,1Hz .. 10Hz has ~70nV peak peak value. If you use two of these OpAmps for amplification you get a factor of sqrt(2). So your preamplifier adds ~100nV pp noise (reffered to input) which corresponds to ~15nV rms.

Let me assume, that you run the AD7799 with a gain of 128 and an update rate of 4,17Hz. Then the ADC itself has a noise of 27nV rms (table 7 of datasheet). Without preamplifier this limited the resolvable signal at the weigh scale output to 27nV. With a factor of 6.5 preamplification, the ADC-noise limits the resolvable signal at the weigh scale output to 27nV/6.5 = 4nV - a clear improvement due to the signal gain.

But don't forget the 15nV rms additional noise due to the preamplifier. Now the preamplifier limits the resolvable signal at the weigh scale output to 15nV. Numerically this is still an improvement compared to the 27nV without preamplifier, but even the numerical improvement is less than a factor of 2 (less then one bit of effective resolution).

Now it's time to remember, that we did not execute a complete noise analysis. The OPA2211 also has an input current noise. Looking at Fig. 2 of its datasheet I would estimate, that with a 500 Ohm DMS resistance of your weighscale the current noise of the OPA2211 gives you roughly the same noise contribution as the voltage noise. In that case the benefit of the preampliflier starts to vanish - in real life you might start to deteriorate the effective resolution instead of improving it.

This noise analysis was neither complete nor accurate. I have e.g. no idea about the input current noise of the AD7799 - if this would be worse than the input current noise of the OPA2211, then the situation would change again.

But to decide if you improve the SNR or if you worsen it, one has to analyse and to understand the noise contribution of the preamplifier. And getting better than the AD7799 alone might be possible, but it's not trivial.

best regards