Lighting and other non-DCC devices.

Not sure what you mean by that. There's never really a situation with "no signal". If a booster has no input from a command station then good design would be to shut down the track output.
If there are locos being controlled then the control packets are sent repeatedly to those locos.
If the command station really is quescent then it outputs a continuous preamble which is a string of '1' bits which results in a square wave at 8.6KHz.
MBQ

You can't have it both ways Guy!!! You say that "all waveforms can be described by a sum of sine waves" and then go on then say that you can have "pulse modulated DC"!! (DC cannot be described by sine waves unless you tale the case of zero frequency as a sine wave). If an FFT shows something other than *only* a DC component (zero frequency) the signal is AC!!!! All pulsed waveforms have a time varying component and are therefore AC.
Jeff

So what do you call it when someone does actually use stretched zeros? (if was you that introduced the restriction on not considering them). It is still DCC as per the spec; so does the description of the waveform suddenly change? No of course not.
(By the way different length 0's are allowed even when stretched bits are not in use.)
Jeff

The components, taken in isolation, are AC. When combined and added to a DC offset the resulting waveform can be DC. The *definition* of AC is that the current or voltage reverses or "laternates".
MBQ

If it has a large enough DC offset it is not AC.
I know where they are.
The resultant signal is also period.
Being periodic is a neccessary but not sufficient condition for a signal to be AC.
MBQ

Please provide an authoratative reference for your definition of AC as anything that varies over time.
MBQ

The problem is that you are only considering the very simplistic definition of AC that is taught at school. You have to look at more advanced text books to see a fuller picture. However, one slightly fuller definition onthe web is here:
Note the section that states:
" As for the computer you're using to read this, its signals are not ordinary sinusoidal AC, but, thanks to Fourier's theorem, any varying signal may be analysed in terms of its sinusoidal components. So AC signals are almost everywhere"
AC signals are *any* signals that vary with time or to put it another way can be defined as an expansion of a series of signals that may be described by V*sin(omega*t+phi), where omega is the angular frequency (2*pi*f), and phi is the phase angle. DC offset may have a term in the series where the frequency component is zero.
Also have a look at section 2 of:
part of which states the following and it goes on to expand further:
"A peculiarity, which is often introduced without comment, is that AC generators (of the analytical variety) are considered to produce sinusoidal outputs. Many practical generators (e.g.,mechanical alternators, radio transmitters) do indeed produce something approximating a voltage or current sine-wave; but the reason goes somewhat deeper than that. If we take, for example, a moving-coil microphone (which is a type of generator which produces electricity from air-pressure variations), we will find that its output in response to (say) the sound of the human voice, is extremely complicated. A technique known as 'Fourier analysis' however, shows that all waveforms can be built-up by adding-together sinusoidal waves of different frequencies; and physical investigation shows that these separate frequency components actually exist."
Jeff

Let me pose to you a little question.
If you apply a DCC signal or a 0-5V pulse data stream to a suitable transformer, what to you see on the secondary side??
Answer: the same as you put in, but there is no DC reference, so how in your view do you classify the output? Is the current reversing? If so in respect to what? Has it changed a 0-5V pulse that you consider to be DC to AC?? Of course it hasn't all that has happened is that any DC term if any has been removed from the mathematical series that described the waveform previously. AC went in AC came out.
Jeff

Perhaps "the same" is a slight exaggeration, the waveform will be distorted slightly depending on the bandwidth and hysteresis of the transformer.
Jeff

Sorry, but that's not good enough. It sounds like the A level physics teacher who taiught us about various op-amp circuits for adding, subtracting, etc and then told us that the new Commodore PET computers that the school had just bought had millions iof these circuits inside. I had no more reason to believe him than a throw-away comment in that article. All of the subsequent exammples clearly show waveforms with zero DC offset.
I don't dispute that you can decompose a signal imto it's components and use AC analysis on this AC components. The quote above adds nothing to the debate.
MBQ

Two very, very different scenarios which only goes to confirm your misunderstanding.
First DCC. We have already established that DCC is *AC*. The current reverses each half cycle. Put it into a transformer suitable for the frequency range and you will get an AC signal out. No problem there, that's how transformers work. You can m
Now a 0 to 5V squarewave.

Finger trouble on the previous post.
DCC is AC and you will get an AC output from a suitable transformer fed with DCC. You can make a very effective block occupancy detector for DCC using a current transformer.
For the 0 - 5V pulse case, that's more like the old transformer coupling of audio outputs. The DC offset is blocked turning the DC input into an AC output.
MBQ

An abortion :-) It's much easier to design for consistent timing.
MBQ

It is very interesting that you consistently ignore the question when the result does not fit with your theory.
Jeff

So is DC if a pulse train has a DC offset and AC it there is not...yeah right.
Jeff

Should read "So it is...."

Does any system actually do that? Digitrax certainly doesn't. This is flat out wrong. Decoders assume that the previous instructions stand until they get an update. If dcc commands stop going out on the rails, the trains will happily keep running unless you kill power. You've never had any runaway trains? It definitely happens.
Not sure what point you're trying to make. This is kind of a dumb message thread anyway. *

Give the man a coconut, zero frequency is indeed a special case. Now find a railway power supply that puts out a perfectly smooth DC waveform.
Bullshit. If current only flows in one direction out of the supply then the current does not alternate.
Bullshit. See above. Guy

As am I. For values of "school" that include the Faculty of Engineering and Applied Science at the University of Southampton, from which I graduated with a B. Eng (hons) in electrical engineering.
Alternating current is current which flows in alternate directions. Note the word current.
A direct current with a superposed alternating signal is still a direct current. If current always flows one way from the supply then to describe it as AC based on the resultant waveform after superposition of a control signal would be, to use a technical term, wrong.
Either that or RS Components, Farnell, Maplin and the rest have mis-labelled every single DC supply they sell. Guy