How do you cut-off any voltages greater than 5V before inputting to arduino ADC?

Hey guys,

I need a way to simply cut-off any voltages greater than 5V when it goes into arduino ADC. Will the attached configuration with simple diodes work.

diodes.png

How much do you expect from the source? If you replace diodes with resistors, or potentiometer?
What is the analog source in this case?

Cheers,
Kari

Attached is the complete circuit. The circuit amplifies the voice from the microphone. The thing is incase if the voltage goes above 5V i want to cut off hence preventing the arduino from overheating.

Thanks

Complete Circuit.png

Have you thought about using a 'crowbar' overvoltage protection diode on the input? By just tying the input between two resistors (or two diodes as you showed earlier) I think you may mute the signal but not actually prevent voltage spikes.

I haven't tried it, but the sort of thing I'm envisaging is putting your signal through a resistor and a zener diode to ground; the amplifier takes the signal at the junction between the resistor and diode. The idea being that the zener does not conduct at all until its breakdown voltage is reached, and when it conducts the series resister prevents it from taking excessive current.

Hi, I think your circuit as shown will work fine.. the 10K will keep any current flowing in the chip internal protect diodes very small.

I used schottky BAT85 diodes to simulate the circuit, and the diodes cut at 5.25V. Will this voltage still burn the ATmega328 microprocessor?

yaantey:
I used schottky BAT85 diodes to simulate the circuit, and the diodes cut at 5.25V. Will this voltage still burn the ATmega328 microprocessor?

The datasheet states that the maximum input voltage to any pin (except the reset pin) is Vcc +.5vdc. So if you are powering the chip's Vcc with 5.0vdc then you are fine with a voltage of +5.25vdc to a analog input pin. Above Vcc + .5vdc on any pin (expect reset pin) then the built in positive clamping diode for that pin will start to conduct and it takes very little current to burn out that protection diode. The purpose of the internal clamping diodes is to help protect the chip from ESD pulses which are low in current (energy really) but can be of very high voltages, which can destroy internal chip's semiconductor junctions. The pins also have a negative clamping diode that turn on if pin voltage goes more then negative then -.5vdc for the same protection purposes.

Lefty

Hey Lefty,

I agree the external diodes are a good idea here…

But, Reality Check?? With 10K series resistor, and 10V applied voltage, the current into the internal diodes would be less than 1 ma. Do you think that the “naive user” would get away with this??

I’ve heard some generalizations about protecting Arduino I/O with 200 ohms and Analog inputs with 10K. Is that defensible?

Thanks!

Could you set up an amplifier that amplifies stronger signals less than weaker ones? Possibly using an LM317T?

terryking228:
Hey Lefty,

I agree the external diodes are a good idea here...

But, Reality Check?? With 10K series resistor, and 10V applied voltage, the current into the internal diodes would be less than 1 ma. Do you think that the "naive user" would get away with this??

Probably work just fine without the extenal diodes due to the 10k resistor, heck I've read one AVR (or was it Pic?) application note where they showed wiring raw 120vac to a input pin to perform zero crossing detection, but of course they showed a series resistor being used of several megohms. Even though safe for the chip most don't recommend that as one has to wire the AC neutral to the chips ground, not the best practice in my opinion.

I've heard some generalizations about protecting Arduino I/O with 200 ohms and Analog inputs with 10K. Is that defensible?

Sure, or one could go full boat and copy the pin/chip/board protection components and methods used by Rugged Circuits' 'M1A1 Abrams tank' version of a Arduino board, I never seen a more comprehensive protection design before or sense, it's very educational to review the schematic of their board which they do publish.

I've found that the AVR I/O pins to be pretty rugged and forgiving, even with no active or passive protection components. The worst damage I ever performed is then during a brain fart I touched a I/O pin of a 644p with the +12vdc bus on a solderless breadboard. There was a spark, a snap, and a pop (but no magic smoke!) and after further investigating I learned that I lost the use of 3 I/O pins, but the rest of the chip and I/O pins still worked fine.

Lefty

Thanks!

I tested with with a 6.2V zener diode to clip off the voltage. I connected the zener diode in the configuration as shown in the link (http://www.electronics-tutorials.ws/diode/diode_7.html) under “single zener diode clipping” and used a function generator to input a AC signal and an oscilloscope to see the ouput. The voltage was cut-off at 7.25V. So, my question is can I use a lower rated zener diode in the same configuration to cut-off voltage at 5V?

Thanks

How about another op-amp running at 5v configured as a voltage follower (effective gain of 1). Wouldn't the op-amp just clip the signal if the input goes above 5v?

You only have a problem because you are running the op-amp from a 9v supply. Why not run it from the 5v supply instead? The LM358 is specified down to 3v. If you're concerned about losing dynamic range, choose an op-amp with rail-to-rail outputs instead of the LM358.