Lighting and other non-DCC devices.

I think you'll find it reduces over time, initially in a fairly linear way and later definitely non-linear. Guy

Not bullshit at all, the DC component may only cause current to flow in one direction, but the transitions will cause currents at frequencies above DC to flow in both directions, they may, or may not, be smaller than the dc current but they do exist and do flow in both directions.

Jeff

Yes, bullshit. Either that or there is effectively no such thing as a DC power supply. Guy

Depends if you recharge it between operating sessions or not.

I suppose you could take that view if the DC level changes or you switch the DC voltage on and off. At the time of the transition there will be AC components generated. However, it would be silly to classify a DC power supply as AC merely because of short AC transients caused by varying the voltage or switching on and off, but in the case of DCC the AC components are there *all of the time* and make up the characteristics of the signal, ie the pulses. Without the AC parts the DCC waveform could only be just a constant DC level.

Jeff

Decoders see a DCC signal not an AC signal otherwise they would not be able to decode the commands. For non DCC locos on a DCC layout they see PWM by stretching one's or zero's depending on direction to move the loco. So you could argue its almost DC in that instance and almost AC with no DC locos.

Note that Hornbys Zero 1 system of the 1980's was very much as you describe an AC supply with control data overlaid where as DCC is the supply and signal combined. What this means for the modeller is that if the supply voltage gets to the decoder so does the commands and everything works. Whereas the Hornby system was not as reliable as the supply voltage might arrive but the commands sometimes got lost.

Not really, no. It's a lot smoother than rectified DC but the voltage will still fluctuate, and will go down as the battery discharges. Guy

You have described AC in a way that will necessarily also encompass pretty much every DC power supply on the market. This has put you in your normal position of being the only one marching in step.

Enough of your sophistry. Guy

On

You very conveniently snipped the next part of the post that said "it would be silly to classify a DC power supply as AC merely because of short AC transients caused by varying the voltage or switching on and off".

With a conventional DC power supply (not DCC or PWM) any AC components make up a small part of the output. With DCC if you remove the AC components all that you are left with are 2 DC levels +V and -V which cancel each other out to zero!! You cannot switch between to 2 DC levels without the AC components, they are there all of the time and are real and form the major characteristic of the signal i.e the data.

Jeff

Yes, because that was the point I originally made and which you were apparently disputing. Since you'd accepted my point but were continuing to argue that you're right and everybody else wrong I did not see much point continuing. Guy

I think the problem here is one of perception, the problem seem to be that this is being viewed from the point of view of a motor control system rather that the general view of how to classify a signal. If the DCC waveform, or any other pulse waveform for that matter, was being used in any other electronic application as a means of sending data down a wire, there is no way that it would be considered as being DC, even considering the dual purpose of the signal (ie conveying data as well as a power source after rectification).

Getting back to whether you can classify slowly changing DC or slow on and off switching of DC as an AC signal, it depends on your point of view and what information you are trying to recover. If you view it over a long enough time period then you can recover AC information, the long integration period being a way to improve the signal to noise ratio of your measurement. So to one person a slowly changing level or slow on/off is a DC signal, but to another it is AC data. In a model rail application it is obviously viewed as DC.

Jeff

Read the first review of Hornby Zero 1 system in MRC, was well ahead of its time, quite amazing really.

Cheers, Simon

Which is what I said in the part you snipped.

Not correct read the spec. In order to keep the average level at zero, if +ve stretched bits are in use then the period of the -ve part of the

0 data bits are extended so as to average the DC back to zero between stretched bits, that is why 0's are allowed to have different length -ve and +ve transitions. However, to the motor the fast -ve bits are are not seen as well due to the Low pass characteristics of the motor, so the motor sees a net +ve (or everything the other way around if you are reversing and using -ve stretched bits).

Jeff

It was not me, I have never said that DCC is DC, that is how this whole argument started.

Jeff

Totally wrong.

Look at the picture in the spec you seem not to have read for yourself. The positive and negative portions of the stretched zero are markedly different. That's how a DC bias in imparted to the DCC waveform to drive an unchipped loco. A "stretched 0" is one which *by definition* has unequal mark/space ratio.

MBQ

No, it is pwm. The apparant frequency variation is an artifact of the enco= ding mechanism, there is no way to be pwm without that happening.

Put it this way. Transmit 101010 and 010101 and the frequency will be the = same. Decoding the frequency will not tell you what the digital content is= . If it were frequency modulated you could decode into the frequency domai= n and know the exact digital content.

I had hoped this thread was long dead.

encoding mechanism, there is no way to be pwm without that happening.

Sorry, but the clasic definition of PWM is a fixed period, i.e., fixed pulse repetition frequency, with varying mark space ratio. There are planty pf text books that wll tell you this if you refuse to accept it from me.

DCC, however, varies the period but keeps the mark space ration constant, other than when using stretched zeroes to control an unchipped loco.

MBQ