RS422

I don't know the exact details, but it is a differential method, not single-ended one like RS-232. You can't look at it quite the same way.

Jon

Differential output -- A is non-inverting, B is inverting. The nominal signal levels are +5V and 0V.

Cheers Chris

What exactly does that mean to say the nominal signal levels are +5V and 0V?

Isn't one signal always the negative of the other? So, if one is at

Chris wrote:

No, when one is +5 the other should ideally be zero. When it is at zero, the other should ideally be at +5.

They should, ideally cross each other at 2.5 volts, though this happens so fast, it is not really important. When one signal is more positive than the other, that is one logic state. When it is more negative than the other, that is the zero state. The exact voltage of either is unimportant, since all that matters is one relative to the other.

The output will go from 0v to 5v. To determine your value of high and low, you would have to compare your A and B outputs The relationship of the levels on the A and B output determines your logic level If A is negative of B you have a mark(off), or binary low. If A is positive of B you have a space(on), or binary high

No they are the 'logical inverse' of each other like any digital signal. Not 'inverted' !

Graham

And if your logic states stay stable for a long period of time a voltmeter works great. If on the other hand your logic is a train of pulses, a 2 channel oscope may be better.

Thanks for the help.

So, if A > B, it's a one, and if A < B, it's a zero?

Also, I remember reading that the noise cancellation properties of differential signaling were due to the currents being opposite. Hence, I thought one was the exact negative of the other. If the voltage is always positive though, then the current is always going in the same direction. So, I guess currents being opposite isn't that important for noise cancellation?

With conductors taking a similar path through electric and magnetic fields, similar noise voltage tend to get added to both signals. Taking the difference of the two resultant voltages subtracts these similat noise voltages with an approxmately zero effective error. Regardless of the amount of common voltage added to each side, the difference between them remains essentially + or - 5 volts, so the decoded state is unaffected.

This method has a big advantage over single ended systems (like RS 232) that subtract the signal from the local ground voltage to determine whether the signal is positive or negative.