I'm using a microphone connected to a TIVA's ADC. The output for the microphone is around 100mV and I used some op-amps to make it work from 0V to 3.3V. However, if the user yell at the microphone, it may go up to 4V. Is there any simple circuit to limit this voltage to 3v3?

I was wondering if I can use a 3.3V Zener diode to limit this, but I don't know what configuration to use.

• Resistor in series and schottky to 3.3V rail, might aswell add another to gnd. i.stack.imgur.com/J1681.png Jun 17, 2016 at 20:27
• Zener in the OpAmp feedback Jun 17, 2016 at 20:34
• Use a rail-to-rail opamp with a 3.3 volt supply. It can't even generate out-of-range signals.
– pipe
Jun 17, 2016 at 20:42
• Also, verify that 4V is actually a problem, if you use series resistors. typical microcontroller ADC's have huge input resistances, and are relative resistant to input voltages > Vcc (not all). READ THE DATASHEET! Jun 17, 2016 at 21:05
• In addition to using clamping diodes or "buffer" op amp circuits to limit the voltage swing into the ADC, some op amp parts provide a built-in "output clamp" feature that works well for this application. See section 4 in this application report published by Texas Instruments: ti.com/lit/an/slaa593/slaa593.pdf Jun 17, 2016 at 22:39

So, the datasheet (something everyone asking a question relating to a specific component should link to) mentions this on p. 1358 under "Electrical Characteristics":

V_in_GPIO: Input voltage on GPIOs, regardless of whether the microcontroller is powered: min = -0.3V max = +5.5V

In other words, you should be fine. Notice that you'd want to have a low-pass filter on the input, anyway for obvious reasons, so something like

simulate this circuit – Schematic created using CircuitLab

Notice that there's no voltage clamping diode here anymore – simply because the voltage follower Opamp isn't able to go below 0V or above its input voltage.

Also note that the values of the low pass filter components are something you must chose to match your signal / your sampling rate – as hinted above, it's absolutely necessary to limit the bandwidth of any signal prior to digitization. Otherwise, aliasing happens. Chose one of the many available RC LPF design tools to determine sensible values. I can also recommend analog devices' Filter Design Wizard, which allows you to create active, op-amp based filters with gain in a single step. Really neat!

• Radc input resistance of the TIVA series is only 2.5 kOhm. Might want to be a bit modest with the low pass filter . Aug 27, 2019 at 6:02

simulate this circuit – Schematic created using CircuitLab

Here is a two-component active circuit that will prevent your analog input from going anywhere above your analog reference voltage. It's advantage over diodes and zeners is that there is no diode-drop to worry about.

I use this clamp on the TI Sitara processors, as those analog inputs cannot exceed Vref (1.8v) by any amount.

I found Mark's solution very interesting if the Opamp has these features:

1. Low input Offset voltage (<<1mV).
2. does not drain too much current (<< 1mA).
3. power supply could be done by what is available (3.3 or 5VDC).
4. High input voltage range (-30 till +30V, that is what you can expect with 4~20mA loops)

AD's LT6015/6016/6017 Precision Opamp looks like it was made for clipping: https://www.analog.com/en/technical-articles/op-amp-precision-positive-negative-clipper-using-lt6015-lt6016-lt6017.html

These answer common requirements for embedded low power applications:

• Input offset voltage typically 50microV!
• Uses only several 100 microAmps!
• Minimum supply voltage 3V!
• Input voltage -25 till +80V!

Check the Datasheet for more detail: https://www.analog.com/media/en/technical-documentation/data-sheets/601567ff.pdf

and Mark: Thanks for the hint:)