Microcircuit AD8422 is very noisy.

In addition, I put the circuit in a shielded iron case, and soldered the middle power point to the iron, the situation has not changed, the noise is still 800 millivolts.

Some recommended - put the positive and negative power supply tracks on opposite sides of the board, but on top of each other for magnetic coupling.

Should I run the power supply tracks on shielded wire?

Please help me to make a circuit with low noise, point out my mistakes.

Hi slava1 ,

Good day. Can you share the schematic so we can visualize your circuit further? Several factors could contribute to the measured noise, such as the supply noise and the resistor noise. I think what you are measuring here is the 0.1 to 10 Hz noise, which is similar to Figure 42.

Can you also provide the measurement of the output of the first AD8422? Thank you.

Regards,
Gilbeys

Привет,  ГЛАДИОН.

Так схема в первом посте или она у вас не работает?

вот диаграмма

I tried measuring the noise on the first chip, where the gain is 1000 times, and the noise was 5 - 6 millivolts.

Привет,  ГЛАДИОН.  I'm writing because all the files with videos and pictures that I uploaded appeared.

It is possible to connect as I have connected according to the diagram, or I have a mistake somewhere?

Hi GLADION

I drew the diagram more nicely

Could you please simulate the circuit for noise?

Hi GLADION

The board is not made neatly, it is a test sample, I did not take everything into account, on this re-soldering more than once, supplemented, the circuit itself later I will do a better quality, and shielding.

Hi GLADION

I made a mistake when I drew the circuit, resistor R3 is connected to ground. I corrected the diagram and removed the wrong drawings.

Could the noise be going from output to input, kind of looping, on the yellow track? If so, what should I do about it?

Hi slava1 ,

Good day. I tried simulating your setup, and the noise you measured somehow correlates with the simulation.

The noise of the system should be around ~757 mV at the output. Since noise is further amplified by the gain of the amplifier, you should observe that amount of noise. Also, the 50k resistor can add up with the overall noise present.

Regards,
Gilbeys

Hi GLADION

Thank you very much. And please simulate four pieces of microchip connected in parallel AD8428 + AD8429 with amplification of 500 times, and a total of 100000 times. If you can please videotape how you model the noise, it will be an instruction for me. Because I watched the program LTSpice, and for me there is not all clear.

And maybe you already have a developed circuit for two AD8422's connected in series, or suggest what errors to eliminate to significantly change the noise. With a total gain of 100000 times.

Hi slava1 ,

Good day. Since your total gain is very high, the noise at the input could considerably be amplified. I can not disclose the modeling of the noise. But I can share the simulation I had to check your setup. May I know what is your input signal range? Thank you.

Regards,
Gilbeys

Hi GLADION

The input signal range is from 5 microvolts to 200 microvolts. But you can simulate ±25 microvolts at a gain of 100000. If anything, the gain will be adjustable from 2000x to 100000x.

And also simulate this circuit for the amount of noise when the inputs are shorted and nothing is fed to them.

four pieces of microchip connected in parallel AD8428 + AD8429 with amplification of 500 times, and a total of 100000 times.

And also simulate this circuit for the amount of noise when the inputs are shorted and nothing is fed to them.

four pieces of microchip connected in parallel AD8428 + AD8429 with amplification of 500 times, and a total of 100000 times.