General advice for noise resistant circuits

Having a noise resistant circuit is important in the test we aim to do, since setting up a clean trigger on a signal that is having frequent voltage pikes is nearly impossible. Since the circuit we tested this morning happened to be quite noise resistant, we figured out this would be interesting information for everyone.

1. Use short wires

Using long wires implies that you equip your circuit with antennas that are waiting to pick up any noise passing by. This happens because any metal forming a loop is an induction loop. What is more, it adds unwanted, uncontrolled and hardly quantifiable resistance that is problematic in the context of low resistance measurements. Thus, we would advise the use of short, flat wires.

2. Capacitive decoupling

Grounding the power supply pins of your chips through a capacitor (100nF works fine) helps reducing noise (especially in the case of the instrumentation amplifier), and is easy to set up.

3. High frequency differentiator design

Differentiators have a tendency to behave poorly at high frequencies; typically, they can become unstable or even integrators. A possible solution for this might be a more advanced differentiator design, as presented below:

                        Source: www.electronics-tutorials.ws/opamp/opamp_7.html

We would recommend a design such that differentiation is guaranteed up to around 1MHz. Typically, we would advise you choose Rin to be very low (wire resistance was sufficient in our case), and Cf to be as low as possible. Side note: remember that putting two same capacitors in series provides half the capacitance of a single of those capacitors.