I have a MiniSense 100 Vibration Sensor with me, which I am trying to interface with my MSP430 Launchpad to track the vibrations of refrigerator compressor. However, I am not sure if connecting it directly to my Launchpad, can damage the MSP Board.

When I tried to research on my own, I came across there different suggestions, such as this one v/s this one v/s this one.

MSP430 has 3.3v operating voltage. In the third one, the guy has directly fed the pin to ADC pin of Arduino. Will doing this damage the MSP430 Dev Board? What schematic should I follow?

I had to ask this question because, I couldn't risk damaging the board. I'm completely newbie to designing schematics. Please guide.


  • \$\begingroup\$ The second option seems to be the better one from my perspective. As you say you don't want to risk damaging your board, putting a 3.3 TVS zener in parallel will protect your input from any overvoltage \$\endgroup\$ – Radiohead Apr 24 '18 at 6:44
  • \$\begingroup\$ TVS diode and zener are two different things. Zeners are much slower than TVS and with that 1 MOhm resistor the 3.3 V zener diode won't start to conduct at 3.3 volt \$\endgroup\$ – Long Pham Apr 24 '18 at 7:42
  • \$\begingroup\$ And if you want go get the amplitude of the oscillation, adding source resistance( or impedance) will make the reading lower than reality. \$\endgroup\$ – Long Pham Apr 24 '18 at 7:50
  • \$\begingroup\$ @LongPham Umm. do you mean that by adding any kind of circuitry I would be altering the analog input to my microcontroller and I shouldn't be doing it? If yes, then should I just connect the two parts? Won't it damage the MCU? I have to measure/track the acceleration values \$\endgroup\$ – Doherty Apr 24 '18 at 12:59
  • \$\begingroup\$ No, personally, I would prefer to buffer it with an opamp + DC bias circuitry because it has very high input impedance. but for simplicity, the circuit which Dorian provided would sufficient. \$\endgroup\$ – Long Pham Apr 24 '18 at 13:45

The one without any protection is really bad.

The other two have a major flaw, both are asymmetrical. Remember, the sensor has a capacitive output.

This will make the signal analysis a nightmare, let's take the first example with a short 2V peak to peak sinusoidal burst. Before the burst the voltage is zero, then a sinusoidal signal with the lower peak aligned to -0.6V (the forward voltage of the zenner diode) then after the burst a 1.4V (the average voltage of the output signal) slowly decreasing to zero signal.

You can see in the picture a 30Hz 1.5V 20ms burst (VI1) and the output after the clipper (VO1) The zenner diode forward conducting is raises the output signal adding a DC bias.

The other two are a 50Hz 300mV 20ms burst and the output after the clipper, identical but also with negative values that ADC cannot read.

enter image description here

Even with a lower than 0.6V signal, you can read only the positive voltages which is a drawback for asymmetrical shakings like a bump.

I would use a setup like this with two clipping diodes and two resistors that add a constant DC bias to make the entire range readable by ADC:


simulate this circuit – Schematic created using CircuitLab

  • \$\begingroup\$ Thanks for explanation. If I am not mistaken, then R3 drops the voltage, while R1 and R2 are used a voltage divider. D2 seems to be eliminating negative part. Not sure sure how d1 and C2 affect the circuitry. Could you please share the specs of the diode D1 and D2 I should use for 3.3 v Launchpad? Also, I am curious, that will adding this circuitry, reduce or alter my readings? since, I can't find the max voltage that can come out of the sensor, I have no idea how to calculate values for voltage dividers etc. \$\endgroup\$ – Doherty Apr 24 '18 at 12:53
  • \$\begingroup\$ Sorry,but could you also, please help me analyse this schematic by sparkfun? learn.sparkfun.com/tutorials/… I am amazed if Sparkfun made a mistake. \$\endgroup\$ – Doherty Apr 24 '18 at 12:56
  • \$\begingroup\$ R3 is only for current limitation when D1 or D2 opens, can be lowered down to 100 ohm if you have noise. D1 and D2 is better to be schottky diodes like BAT 41 that have a lower forward voltage because at least with D2 there is an ESD diode in parallel inside your chip that we dont want to open when clipping. \$\endgroup\$ – Dorian Apr 24 '18 at 13:09
  • \$\begingroup\$ R1 and R2 are making the DC bias, when no signal from the piezo sensor the ADC input voltage will be half of the above VDD. I hope you understand , that line you should connect to a voltage, if you use ADC reference voltage you will use the maximum ADC range since the steady voltage at the ADC input will be at the half of the maximum. D2 limits the voltages above VDD. C2 is a decoupling capacitor, you might already have one on the board, is for cutting the current spikes to VDD \$\endgroup\$ – Dorian Apr 24 '18 at 13:20
  • \$\begingroup\$ @Doherty I wouldn't bet to much on Sparkfun's design, as you see their signal is also limited to -0.6V after connecting to Arduino board even after "dumping" that means that the negative currents ar still dangerously going through the ESD protection diode inside chip. The eqiuvalent impedance for my circuit would be 0.5 M ohm, if their measurement are accurate this will bring the voltage to approx 2.5 V max, appropriate for ADC range. \$\endgroup\$ – Dorian Apr 24 '18 at 13:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.