What type of digital signaling is more efficient, digital DC or digital AC?

Bararontok

What type of digital signaling is more efficient, digital DC or digital AC?

With digital DC, the square-wave signal is only between a (+) value and 0.

With digital AC, the square-wave signal is between a (+) & (-) value.

Bararontok

The thread originator is requesting an answer to the question.

berkeman

Without more context to the question, the answer is that they are the same.

skeptic2

Can you define efficient?

Bararontok

Which digital signalling method listed in post #1:

-Uses electronics that is less complex and prone to malfunction
-Is less error prone
-Uses less electrical energy
-Can operate at higher data rates

skeptic2

[QUOTE=Bararontok;3990445]Which digital signalling method listed in post #1:

DC - Most logic devices are unipolar.

One measure of immunity from errors involves the amount of difference between a 1-level and a 0-level. If the difference between the levels for AC and DC transmission is the same, the immunity from errors should be the same.

I'm assuming you mean uses less electrical energy for the same bit error rate. I'm not sure there's a difference.

I can't think of any reason there would be a difference. In my experience more efficient signalling methods involve more efficient coding algorithms rather than relying on AC or DC.

DC has the disadvantage of not being able to pass through capacitors or transformers. This may be important if bandpass filters are used to reduce noise.

Bararontok

What if the DC is a fluctuating DC? A constant DC cannot pass through a capacitor in series because once the capacitor is fully charged, it can no longer permit current flow, but in a fluctuating DC, the capacitor has time to discharge its stored energy when the power is cut off.

In the case of a transformer, is a fluctuation in current enough for the primary inductor to emit radiation that can be absorbed by the secondary transformer or is a reversal of polarity necessary?

skeptic2

With both capacitors and transformers the fluctuation of the DC, which will include both data and noise, will pass through, but the DC bias will not. At that point we can no longer call it a DC transmission method.

Bararontok

That is correct since the DC bias comes from a constant DC power supply while the fluctuating DC signal is a separate component.

rcgldr

If the orignal post means digital signals used in common integrated circuits, then most logic just uses some positive voltage and zero voltage for the two states. Decades ago, computers like the HP 2100 mini-computer used +5, -2 volts for it's TTL logic on it's circuit boards.

Interfaces like USB send synchrounous (fixed clock rate) data via a differential pair of signals, +D and -D, where -D is an inversion of +D, each with a 3 volt range. There are also ground and +5 volt pins, which the host drives to send power to some USB devices. Wiki article:

http://en.wikipedia.org/wiki/Usb

Interfaces like RS-232 can range from +/- 3 volts to +/- 15 volts, and for the original PC, +/- 12 volts was used for RS-232. There is a ground pin used as a reference for 0 volts. Wik article:

http://en.wikipedia.org/wiki/RS_232

HowlerMonkey

Won't the "phase angles" of a sampled A/C become random unless your sample rate is matched to the A/C frequency?