I'm reading a book about data transmission inside a PC.

I'm trying to understand a phrase, translated from german, saying this :

In a PC, data are transmitted almost exclusively using the so-called 'neutral current method', which could be for instance NRZ or RZ.

I'm not sure if 'neutral current method' is correct, as I didn't find anything relevant with google.

Does anyone know what this method is ?

  • \$\begingroup\$ NRZ = Non Return to Zero. That may help your Google'ing. It basically means that the signal on average is equally long high as it is low. I'm sure someone can further clarify better than I can. It is necessary to make the circuit reliably detect HIGH and LOW signals over long time. \$\endgroup\$
    – jippie
    Commented Apr 25, 2012 at 7:48
  • \$\begingroup\$ @jippie: "on average is equally long high as it is low". That's only true if the number of 1 bits is equal to the number of 0 bits. For other coding methods, like Manchester, this is always true, regardless of the 1/0 ratio. \$\endgroup\$ Commented Apr 25, 2012 at 8:23
  • \$\begingroup\$ Frankly, I don't see any relationship between "neutral current", a "PC", and "NRZ/RZ". I would say change the book :-) \$\endgroup\$
    – Telaclavo
    Commented Apr 25, 2012 at 10:14
  • \$\begingroup\$ Googling "neutrale ströme" gets lots of links to pages about Bosons & Neutrinos. Possibly relevant to a PC fom the 23rd century :) \$\endgroup\$
    – MikeJ-UK
    Commented Apr 25, 2012 at 10:56

2 Answers 2


I think the notion of "neutral current" is based on the fact that the voltage swings positive to negative around the ground reference (i.e. signal return path). In effect the current direction is different when the "signal" is high than when it is low. If the encoding is balanced (e.g. Manchester encoding), then on average the net average current flow is zero or "neutral."

  • \$\begingroup\$ This can be important if you have DC blocking caps at either end of the transmission line. If the average current is non-zero, charge will gradually build up on one of the DC blocking caps. Build up enough charge and you will let the "magic smoke" out. \$\endgroup\$
    – ajs410
    Commented May 9, 2012 at 21:35

I think its a bad translation.

"Neutral is used to indicate having no strongly marked or positive characteristics or feature."

In connection with RZ and NRZ baseband coding techniques, these are used in serial communication ports. The only thing remotely analogous to "neutral" for serial communication links are "balanced lines" which use differential voltages on controlled impedances to optimize signal quality over long distances. This applies to ethernet and RS485 & USB but not RS232.

..and thats my final answer for 1 million dollars

  • 2
    \$\begingroup\$ Balanced lines aren't always "current neutral" in the sense that the net current is zero (if that would be the meaning). In USB for instance both D+ and D- have positive voltages/currents. \$\endgroup\$
    – stevenvh
    Commented May 9, 2012 at 7:53
  • \$\begingroup\$ if you put a current probe around differential pairs... the current is null. \$\endgroup\$ Commented May 9, 2012 at 8:05
  • 1
    \$\begingroup\$ Yes, but that's because a current probe doesn't see the DC component. \$\endgroup\$
    – stevenvh
    Commented May 9, 2012 at 8:07
  • 1
    \$\begingroup\$ OK, I thought you were talking about a current transformer. But what you say isn't true. The Hall sensor (capital "H") will see a net current in both lines in the same direction: the voltage is positive for both (an AC signal superimposed to a DC level) and both lines are terminated to ground. \$\endgroup\$
    – stevenvh
    Commented May 9, 2012 at 8:23
  • 1
    \$\begingroup\$ agreed... again .. I think its a bad translation... very bad .... "sehr schlecht" \$\endgroup\$ Commented May 9, 2012 at 8:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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