Read signals made of 8V square waves on an Arduino

Question

I am working on putting together an Arduino board that would read signals sent between two modules in a car. The signals are sent as 8V square waves.

An example of a message sent this way would be a sequence that is 12mS long consisting of a 8V high for 6mS then a 0V low for 9mS repeated back to back 14 times.

The problem I have is that if I put more than 5V into a digital pin on an Arduino I will blow the board. This gets worse because with this being in a car it is completely possible that the digital pin might accidentally see full automotive voltage of roughly 15V.

My thought was to use a voltage divider that would bring 15V down to 5V this way even if the digital pin saw the full 15 voltages it wouldn't hurt the Arduino. I would also look at Zener or other method for over-voltage protection.

My concern, though, is the timing is in milliseconds and I don't want the timing to be off or I won't be able to decode messages accurately. What I am really looking for is way to read these signal in real-time just at a lower voltage. Will this work or do I need to go a different direction?

Answer 1

You really should want some hysteresis as well as signal conditioning to your Arduino input levels. I'll get to that near the end.

I don't know what's powering your Arduino. If you are supplying it through an inexpensive \$5\:\textrm{V}\$ converter that works off the automotive voltage rail, you may find that it protect your Arduino from periodic load dumps.

Assuming you have a clean and safe \$5\:\textrm{V}\$ supply for the Ardiuno, then your main worry is if your Arduino I/O pin gets exposed to an external voltage (relative to the ground, of course) that is above the \$5\:\textrm{V}\$ supply rail. In such cases, the protection diodes will kick in and start leaking current into the supply rail. On most processors I've used before, I find that observing a maximum of \$2\:\textrm{mA}\$ meets the absolute maximum specifications. However, you may need to look up your specific processor to get the actual value for it. Since you didn't mention the exact Arduino board, I can't look it up for you. But the following circuit should be entirely safe to use and is easy to make.

^{simulate this circuit – Schematic created using CircuitLab}

The above circuit uses two very cheap BJTs to create the hysteresis and the desired output signal. The output to the I/O pin is low when your signal is low and is high when the signal rises over about \$3.9\:\textrm{V}\$ (rising direction.) In the falling direction, the signal must go below about \$2\:\textrm{V}\$ for the output to return low. So the circuit includes a hysteresis band that is almost \$2\:\textrm{V}\$ wide, centered over about \$3\:\textrm{V}\$. So, good noise tolerance.

\$R_6\$ (and \$R_5\$) provides plenty of protection, as well. You could add a zener. But I don't think it's needed. In the case of supplying \$15\:\textrm{V}\$ to the signal input, the \$Q_1\$ base will experience about \$400\:\mu\textrm{A}\$ of base current. Which should be just fine. Even with \$100\:\textrm{V}\$ at the signal input \$Q_1\$'s base still sees only \$4\:\textrm{mA}\$. And either way, the I/O pin is protected. (Even if \$Q_1\$ completely fried all pins short for some reason.)

It also should be plenty fast enough for your needs.

Answer 2

You can use a voltage shifter or voltage translator. You need to provide the battery voltage though. CD4010B-Q1 CMOS Hex Buffer/Converter

Voltage Level Translation Guide

^{simulate this circuit – Schematic created using CircuitLab}

You can also use an optocoupler (May be with or without the driver transistor depending on the available current of your input signal.)

^{simulate this circuit}

Both of these solutions will give you propagation delays in the vicinity of 100nS using commercial grade components. So you can use them safely in application of a few kilohertz.