- Jul 06, 2017 -
What happens to the amplifier output ringing and overshoot?
I saw ringing and overshoot at the output of the amplifier. I follow the guidelines in the data sheet, and the layout is well organized. Could it be wrong?
Such problems are really confusing and frustrating. Engineering design is science, A plus B results should be C. If you have designed the circuit for some time, then you should know that engineering is also art.
The engineer has read the data sheet. This should be a good start, but it is surprising that many times this is not the case. Therefore, we began to study the problem in depth. The first thing we studied was the schematic, first checking the "ordinary suspect": the amplifier noise gain, the bypass capacitor, the load and the supply voltage. Why are these?
The noise gain determines the stability of the amplifier; if the phase margin is low, the output may be ringing and overshoot. The bypass capacitor excludes noise from the amplifier and stores the charge at the power supply pin. This is especially important when the amplifier requires a stable current supply, because its output is changing rapidly. When the output is pressed, if the supply voltage changes, the change will certainly affect the output. If the capacitor or inductance is too large, or the load resistance is too small, then the load may cause problems. When the supply voltage is too large or too small, the performance of some amplifiers will be reduced, so the supply voltage should be checked against the data sheet.
If there is no problem with the above, what would you do? Continue to find faults in other areas. Next look at the layout and routing. Is there a long route with parasitic inductance? Is there a bypass capacitor away from the power supply pin, making the parasitic inductance and these capacitors form an oscillating circuit? Does the input and output pins below the ground layer creep effect form a parasitic capacitance that causes ringing and overshoot? On the issue of the engineers, the layout and wiring seems to be no problem.
So what's going on? How does the circuit test? Is the input clean and the terminal is correct? The engineer saw a little ring at the input, but not too much. We know that the garbage input is equal to the garbage output, so we try to purify the input. Termination is correct, so we change a signal generator to see if it has a problem. The new generator performs better, but the input and output are still ringing. I asked the engineer whether he was using a cable or an oscilloscope probe to check the signal. He used the oscilloscope probe, so I asked if he had a grounding clip. Indeed, the grounding clip is about 3 inches long. I guess the problem may be here, tell him to remove the clip wire, unscrew the plastic tube at the top of the probe, and use the oscilloscope probe's metal liner to pick up the signal next to the ground. When he does, the ring will disappear. Look, my guess is right! So what exactly is going on?
The grounding clamp has a series inductance, the probe has a capacitance, and the trace on the probe has parasitic capacitance. Capacitance and inductance form an oscillating circuit that oscillates under the rapid rise of the circuit, causing the input and output to ring and overshoot. And then tell you a trick: before the measurement must be calibrated oscilloscope probe, this can also help reduce the peaking. So that it can be closed!