Digitrax DCC

I am building a large (HO scale) basement layout, powered by Digitrax DCC.

I was anxious to start running trains and although I have my DCC buss (12 ga stranded wire) under the layout, I have yet to install any jumpers under the new yard.  The closest DCC jumper wire is in the peninsula (foreground) and is at least 30 feet from the yard.  The sound locomotives operate in the yard just fine. My voltage meter indicates about 1/2 volt power loss over this distance.  Power is being conducted solely through rail joiners!  I will eventually install more jumper wires, as using rail joiners to conduct DCC power is not a good  long term solution.

I have been a satisfied DCC user for over fifteen years.  Regarding DC versus DCC; if your layout runs well on DC it will run well on DCC.  Clean Track - Good Electrical Buss.

Before converting to Digitrax DCC I had a large basement layout and used two MRC20 walk around power packs; and a small MRC power pack for the yard only. Like my current layout I had a 12 ga buss and  plenty of jumpers. I used common wire DC and used toggle switches to control the blocks.  In planning to expand control I would need to replace all of the toggle switches with rotary switches, and pull additional buss wires under the layout  for each new power pack.  Although I loved the MRC20 power packs, the cost  to move to DCC was a "no brainer".

I ran another small OPS session last week. Although we only had three operators, the capability to work two locomotive crews in close proximity was great. No one had to worry about electrical blocks.

I'm not sure exactly what

I'm not sure exactly what you're asking. Do you have a DCC layout?

Anyway, I've been running trains on DC since 1989 and switched to DCC in 2006. Heck no I'll never go back.

DCC is so much better in so many ways that i would never consider switching back to DC ever again.

And my layout is small. 11X7 Shelf.

Ring Engineering bets on Filtered DC

If you check the sales pitch for the RailPro system (RailPro vs DCC), the claim is made that the filtered DC power supply of the RailPro system is much better than DCC for keeping rail-to-wheel contact.  It seems that Ring Engineering has made a large bet on DC by developing the RailPro system.  And while Ring Engineering does not advocate the use of inboard batteries, battery power would be the ultimate solution to dirty track problems.

Keep alive capacitor circuits

In theory, keep alive capacitor circuits seem like they ought to solve the problem of intermittent power pickup, right?

In practice, they work, but they're not that great.

I have Lenz's Gold decoder keep alive in a loco. Yes, the loco does keep running when it hits dead frogs or dirty spots in the track, unlike how it ran on a decoder without a keep alive circuit.

But the keep alive capacitor takes almost as much room as a speaker does and it's only good for about 1 second before it's completely discharged.

In practice, a loco with this keep-alive circuit in it will slow when it hits dirty track or a dead spot like a dead frog, then it will speed up again. In other words, the loco does keep running, but it's erratic in how it runs - normal speed, then slow, speed back up, run for a while like normal, then slow down again, speed back up, slow down ... and so it goes.

If a normal loco runs like that, you put it on the workbench for repairs. Erratic speed performance is not enjoyable performance.

So my conclusion is the keep-alive capacitor idea does not truly solve power pickup problems. The loco does NOT keep running like nothing happened. Its performance is not ideal by any stretch. The real solution to this problem in my mind is batteries. They don't take that much more room than a decent sized capacitor and they most definitely keep the loco running a lot longer (hour plus instead of a few seconds).

The real solution

It sounds to me that the real solution to this problem is to fix the trackwork.

Fix the track

Well, that depends. Dead frogs do not necessarily represent broken trackwork, nor is dirty trackwork (or dirty wheels) considered "broken trackwork".

A solution for the frogs is to put something like TAM Valley frog juicers on the dead frogs, at about $7 per frog. As for the dirty track or wheels, that can be an ongoing maintenance issue. Various solutions exist to help improve the dirty track problem, but it doesn't seem to stay "fixed" forever. You will have to clean the track or wheels again eventually.

The keep alive capacitor idea comes up as a solution to these problems and I'm pointing out it's not that great a solution.

If you want a tech solution that improves the track pickup issues neatly, then your best bet is to pursue battery power, not keep alive capacitors.

Power

For some years, our club used pulse power DC throttles made up from similar throttles described in MRR magazine. Most locos had flywheels, can motors and ran very well at slow speeds. We switched to DCC as this became a better way of preventing running into some else block while gabbing.

We adopted to not be lazy and kept the track clean, right from the beginning. Most members had had a home layout also.

Rich

Technically

Running the theory  DCC will have better connectivity over the same dirt because the higher voltage will punch through before the lower DC voltage. 

As I said on an earlier post on the other thread.  I have not seen any difference between DCC and DC locomotives on dirty track or dirty internal power routing over a lot of years as long as the DCC locomotives are not sound equipped.  As an addition to the earlier comment, I have found that locomotives that have sit around for awhile have more problems than those who get run more often.

The current through the motor is the same in DC and DCC locomotives.  The system power is roughly the same until sound is added.  The sound is a big addition in power and will discharge any input capacitance much faster making the hold up time much shorter causing both the sound and motor amp to drop out sooner.

Note the eye cannot detect changes at very fast rates--the flicker of a 60Hz powered light bulb or an engine run on 120Hz pulsing dc is not detected.  At the same time, the ear can detect 14KHz or above, a frequency more than 100 times as great.

The ear is detecting the clicking in the sound system in locomotives as they go across the dirt.

I submit the ear is detecting the dropout, not the eye and equivalent functioning DC and DCC locomotives operate very close to the same on the same track/wheel conditions.

From another angle, the higher DCC voltage will punch through a thicker layer of dirt than a DC locomotive running the same speed indicating the opposite is true.

Having said all of the above, it is much more annoying when a sound equipped locomotive runs over dirty track even if the engine seems to be running smoothly so I can see why one would think the way they do.

Terry

DC vs, DCC

It seems to me that whenever I see a thread about DC being a better power source than DCC, it's really saying that straight power from the tracks is better than also sending a signal through the tracks to communicate with a decoder. It's the signal that seems to degrade most with dirty track, causing issues with sound especially. 

The only place I can remember seeing these types of discussions, is in threads about radio control of decoders. If the only thing coming through the track is power, and the decoder gets its communication through a different channel, (radio for example) a quick voltage drop is less likely to be noticeable.

Does this mean that anyone who has build a DCC layout would ever go back to the days of DC block control? No, I think they are simply trying to point out benefits to using a secondary channel for control.

DCC is Still DC Power...

There's a huge misnomer here that suggests DCC is not DC, when indeed, DCC is simply a more direct manner of controlling the flow of DC current.  Whether you use batteries, DCC, or no command and control interface whatsoever, you are still using DC - unless you're with those guys using AC, of course!  And yes, while there are a multitude of modifications to DC power, it's still DC.  We're concerned with getting the DC power to the components we want to work when we want them to be working.

My observation is that with DC, where there isn't a decoder in the way, it simply seems to me locomotives run better with less finickyness.  I see brass on DC power running without a catch - perhaps not at the lower voltages, and there may be a higher starting voltage, but overall, they don't hiccup, they don't stutter, they don't stall - though yes, there are some models that indeed DO!  But I've seen one too many DC layouts where the owner has not cleaned the rails in 15 years, and yet, the trains still run smoothly.  The locomotive wheels are worn down to the brass under the nickel plating, but they still function.

DCC communication introduces a square wave signal on top of the DC track current.  The dirtier the contacts are, the more intermittent this signal becomes - the squares become humps, then indistinguishable altogether - any one with experience with electronics knows this is the case.   The cleaner the signal, the better the communication; if we separate the signal from the rail power, we should get better control.  As decoders become more advanced, and we see more 2-way decoders emerge on the market, there's going to be a lot of information in those rails.

It then makes complete sense to lift the signal out of the rails and put them in the air where they don't degrade near as much as they do in the rails.

Now I am not quick to completely throw away all virtues of DC power that comes from an AC transformer [aka the wall].  I have way too much experience with batteries and their shortcomings.  At first, they are a joy, but in time, their performance falls off.  First, they stop working before you're done with your session,and then, the longer you use them, the shorter your usable sessions become until finally, the battery is kaput.  New batteries are what, $20, $30, $40??  Replace that every 2-5 years, ouch.  Nie, I say, we shouldn't consider the battery as a primary source of power.  As a secondary source, perhaps, but not the primary.  Indeed, if you look around my workshop, you'll see all new power tools have cords - and that is not an accident!

The keep alive circuit is a very important comopnent in preserving communication between the decoder and the transmitter.  Obviously Sound introduces an issue because it eats up so much current.  Perhaps what we need to do is revisit the decoder circuit altogether...

1) Establish the keep alive circuit so that the decoder only maintains motor functions and lights in the event of a power drop.  Hence, when power cuts out, so too does the speaker, but we're not dead in the water as is the most real issue.

2) Rewrite the sound file programs so that when we have such intermittant shorts, the sound file does not start over a tthe beginning, but rather, the decoder keeps track of where in the "song" we are, and thus we don't experience a "skipping record," persay, but rather, the "static radio" effect.

3) Obviously we're now talking about a much more complex decoder - something in the order of a micro processor - and a program file dump [memory]. We build this microprocessor/memory into the wireless interface board [the one that allows us to plug and play any pre-existing DCC decoder], perhaps load an OS into that static memory space, and we have a very functional machine under our fingertips.

#3 opens a number of doors to other potential operations that would be delightful to have for any model railroader, whether they operate their 10-operator club-sized layout alone or not.

This whole conversation would hopefully see track cleaning evolve back down to a simple wipe of the railheads now and then,  no magic formulas, oils, toxic cleaners, etc and so forth.  No sandpaper, polishing, treatments...perhaps a vacuum car, now and then...

That's my take on the issue.  Your mileage may vary.

quick voltage drop

The function, what ever it is, drops out with the voltage drop as the controlled circuitry, such as a motor, can no longer function. 

At the frequencies used,  the dirt is no impediment to the control signal.  It's capacitive coupled with the dirt acting as the capacitor. 

It's the DC power needed to drive the sound amplifier which is the problem.  Not enough power, no sound even when the control signal is there.  There are a lot of applications that do not need a continuous control signal to work properly--it just does what it was last told to do until told to do something else.

Note that operations suffer when the wire diameter is too small.  This indicates too much voltage across the wire and a lack of voltage at the controller and controlled things like lights, speakers and motors.  The control signal has mostly no effect.  I was driving a simple dcc decoder through about 50 feet of track.  The locomotive slowed down at the far end.  Still functioned properly, but slowed down.  So did the dc locomotive unless I increased the speed using the throttle.  Same manufacturer, same model.  The only difference was the decoder in one.

One other thing that points to a power rather than control problem--the control signal does not have to have the same amplitude as the controlled device.  With proper design, I can control a 120vac device with a 5vdc signal. 

Terry

DCC is not DC

Just a quick note..  DCC is not DC on the track.  DCC uses AC, varying between positive and negative 16V (roughly), with the length of the individual waves defining the bits.  The net effect to a motor is ) volts, which is why a model not equipped with a decoder will sit and hum, as its armature is repeatedly pulled back and forth by the AC wave.

I would also like to point out that even the slightest momentum in the mechanism will overcome momentary lss of connection, in either DC or DCC, but there is no comparable momentum of sound.  Therefore, our ears hear even the slightest interruption in sound, while our eyes may never notice a nanosecond loss of power.

...

When the visible loss of momentum is precieved, we're long past the point of audible loss...

Not true

"DCC communication introduces a square wave signal on top of the DC track current.  The dirtier the contacts are, the more intermittent this signal becomes - the squares become humps, then indistinguishable altogether - any one with experience with electronics knows this is the case. " 

This electronics person disagrees with this whole statement. Ohm's law and its reactive equivalents say this will not be the case.

Terry

Wanna try again?

DCC allows an input to the command stations/power boosters that can be either ac or DC. At the track level, the voltage that your engines see, is...

NMRA:

This is why the whole DC v DCC myth is silly; DCC is in essence simply a more controlled use of DC, a square wave riding the DC voltage.

There is in existence a couple companies running AC power on their track; by and large they're European.

Okay

Measurement just made at the rails: 12.4 vac, 0.002vdc.  Checked both polarities.  Locomotive on track with sound running. 

I got called on the same thing a while back and was referred to a specific reference that said the voltage on the track was ac.  Measurement confirmed this.  I do not remember the reference.

There is a dc component between the ground of the booster and each rail which is the same for each rail indicating there is no dc on the track.

The waveform is technically not a square wave, but a rectangular wave.  No data can be transferred with a square wave.  Data is sent by varying the width of the "low" and "high" portions of the waveform.  The peak value of this data waveform is the peak of the voltage.  It is not riding on a dc level.  I have looked at it on an oscilloscope. 

Marx, Lionel and American Flyer used AC on the track.  I don't know if they make ac products anymore, but their stuff sure turns up a swap meets.

Terry

wave shape

Check the first two DCC standards in the NMRA DCC section

Note particularly the data in waveshape is full height, the waveshape drawing is referenced to the center which indicate an ac waveform and the references to voltage are "bipolar" indicating something other than dc

Terry

...

Look, you can argue against the specs provided by the manufacturers all night if you wish, but the output from the DCC systems provided by Digitrax and NCE are listed as VDC outputs.  If they were AC outputs, they'd be listed as VAC.  It is VDC that our DCC systems provide to the rails.

DCC Voltage is a high frequency square wave made within the DC power source in the DC power supply.  It is _not_ AC power as it is commonly suggested.  AC is Alternating current, a sine wave, but one hump is in one plane and the second hump is on a plane 90 degrees to the first plane.  DC is Direct Current, in one plane; with DCC you have direct current "alternating" between Vmax and Vmin, but it is _not_ AC voltage.

In order to properly read your DCC voltage, you will need a multimeter with true RMS capability or an oscilloscope. Using the meter you have onhand, you'll be able to read the voltage value in terms of VAC, whereas you're measuring the potential difference between the two extremes of the wave [0 and +13], but it's still high frequency VDC that you're measuring.

More myths we have to undo...

...see, this complication by arguing the rectangular waves versus square waves does not address the issue of  dirty contact. The issue with poor contacts relates to the shape of these waves themselves, whereas the more "dirt" you have between the contacts, the more the square wave distorts on the edges, hence, instead of nice right angles at the beginning and end of the wave, the ends get rounded off.  Do this enough and the wave [be it square or rectangular] is no longer readable by the decoder, hence, the decoder no longer receives the signal and it stops functioning, even though power is still present.  The solution has been, find a way to micro clean the rails and add even more contacts to every single wheel we possibly can.

Moving the signal to the air means we no longer have to contend with the noise produced by what amounts to a microlayer of electric dust on the rails.  I Never had the issues we have in DCC with even my old Bachmann junk, running on sectional track; in that case, my issue was plastic wheelsets depositing a very sightable amount of crude on the railheads.  Mind you, these older models only have pickups on one truck.  Now, we have the cleanest track I could ever imagine,and yet we have more contact issues than ever before.  My hypothesis: get the signal out of the rail and put it in the air where it stays clean.

starting at the top

1.  I cannot find a reference or specification to the output of the boosters ie track voltage for either manufacturer.  I do find references to the system input.  In the Digitrax case it is AC or DC.  Your reference to VDC is the output of the power supply, not the output to the track.

2,  AC does not have to be a sine wave.  A sine wave is a special circumstance.  I can make any complex waveform into ac in any system using capacitors or transformers--sound comes to mind.  AC is in a single plane just like DC.  It is referenced to 0 v and goes positive and negative from there.  I can have either ac or pulsating dc depending on the reference.

3.  The meter reads ac in true rms. If the voltage were dc then I would read an average dc value on the meter and would get a negative reading of the same value by reversing the leads.  I get the same both ways. 

4.  Check the NMRA specs you will notice the wave form drawings show different widths therefore the waveform cannot be square.  None of the drawings show data riding on a dc level.

5.  With the data waveform being the same height as the power then Ohm's law comes into play and the decoder fails when the voltage level at the decoder drops below the minimum for operation.  The dirt is a high value resistor.  Read the NMRA specs.

6.  One can add a spike or ringing to the front edge of a waveform just as easy as rounding it off.

7.  Do you know how much electronic noise is in the air?  The various clicks and hash heard on radios when some electrical apparatus turns on or off.  Your dryer maybe, your computer, or worse, your neighbors computer or stove.  How about a microwave?  The ether is full of garbage and not as clean as most think

Other comments. 

The NMRA specs allow both AC and pulsed DC data paths.  The decoders don't care.  Run one with the front to the left--it runs.  Turn the engine around so the front is to the right--it runs just as well on the opposite polarity.  A waveform on a dc level would have some problems with this

Checking output voltage on a Digitrax booster is done by connecting a dc meter between either rail and ground and multiplying by 2.  That's pretty close to what I read on my meter when measuring ac between both rails.  Pretty much says ac to me.

Terry

Picture time

How about a picture from the oscilloscope. Would that clear up this disagreement?  

Bernd

picture

It would definitely help.  I've personally looked at the waveforms, but I no longer have access to an o'scope since I retired.  All I have are the waveform drawings from the NMRA specs and a Fluke multimeter.

The measurememt details, where are the probe(s) connected and where is the probe ground connected are important.  Different connection points will show different waveforms.  The o'scope trace to show all the detail is not the simplest measurement.

Terry

Bringing out the textbook...

When I could not find the output from the boosters, I looked at the command stations - whereas boosters are command stations without the command part - and I found both Digitrax and NCE have an output of 13VDC; not 13VAC.  I went so far as to post both links for you, along with the information as it appears in the literature.  There is not confusion here.  The boosters as advertised may take an input of either AC or DC, but your suggestion that since you could not find any literature describing the output and therefore making a conclusion without the information is like me saying that since I can't find any literature about what the output is on my laptop power supply is, the output must be the input I put into it or whatever i want to believe it is. Either conclusion would be entirely wrong. 

So I need better reference material to explain the difference between AC power [which DCC powersupplies can use as input] and Varying DC power [what DCC uses as output at the rails.]

I found this site, it's most helpful: http://www.kpsec.freeuk.com/index.htm It has all the information I need, everything[and more] from my formal instruction that I have forgotten, so I will refer to it from here on out.

Step one, Whats the difference between AC and DC power: http://www.kpsec.freeuk.com/acdc.htm

AC power:

 DC power:

These are empirical definitions, there's nothing to debate here.

The short of this is, AC power consists of two magnitudes and two directions.  DC power can consist of two magnitudes, but it ONLY has one direction.  DCC uses Variable DC.

When you measure DCC voltage with the meter set to DC, you are measuring the part that remains constant - the DCC "signal" - hence you read 0.02 - which is why you can hear DC locomotives set on a DCC layout "Hum."  The Varying part of the voltage has to be measured with the multimeter set to AC, but it is not AC; it is DC, acting like AC [insomuch that it is not constant].  Your meter set in AC measures the difference between the minimum and the maximum and provides to you the difference as a number in volts.

By the way, I found this image and caption quite interesting, considering decoders are electronic circuits:

We know DCC is digital information; we know we have a circuit that needs steady power and the means for transmitting information at the same time.  Hence, we know that our command station is supplying a constant voltage.  But how does it vary? 

Moving on...

You're arguing the semantics of the difference between a square wave and a rectangular wave where such definitions really make no difference.  All square waves are rectangular waves; the "square" mostly applies to the shape of the wave.  Digital information is transmitted as square waves, be they square or rectangular as you wish to call them; and you're quite incorrect about one thing altogether; DCC comminication is not a variation in the vertical position of the peak; it is a hortizontal change in the distance between the edges - it is Digital information!

But the issue here is not what we call these waves, the issue is how dirt affects the shape of the wave, effectively changing our message.  It distorts the beginning and the end of the wave, thus confusing the decoder; when the decoder is confused, it simply shuts off; or perhaps you could have a sudden rocket start, though that may be a different issue altogether - best we not confuse it!

Here's the information as presented by this site.  Mind you, the site is not discussing Train Digital Command Control, it's discussing electronics in general.  The general principles still apply..

http://www.kpsec.freeuk.com/555timer.htm

http://www.kpsec.freeuk.com/counting.htm

http://www.kpsec.freeuk.com/analogue.htm

The big issue is best expressed by the digital expression of the clock, but we have to work up to it.

 We then introduce Duty cycles - the idea that varies the length of our wave, thus producing a different output.

A short discussion of triggers - imagine for a moment how dirt may affect such a circuit.

All together now:

What we have in DCC is very similar to this, though probably different.  We enter a value in our our throttle, the Command station interprets the input numbers as a packet and transmits it to the decoders onthe track [all decoders receive all packets - they're all attached to the same track].  This packet which is literally nothing more than a series of numbers with a header and a footer telling the decoders which decoder the information is meant for and when that decoder should start reading the message and then stop.  In the decoder,  every output is controlled by a CV; the numbers in the packets are digitally addressed to each CV; the data string is interpreted by the decoder and it sets each CV as the message instructs it to set them. 

When we have dirt on the track, the information we send gets distorted - and even though we're actually sending every message three times when we send it, that microsecond these messages are transmitted in is shorter than the 100 milliseconds the locomotive takes to clear that bit of dirt is has encountered on all 4, 6, 8 or 12 wheels you have on the ground.

The decoder fails because it no longer has communication contact with the command station, even though there is power beneath it.  It's not a matter of the power being lost, but the decoder not knowing what to do with the power it finds.  In such cases, the decoder defaults to "off" [as if it were sitting idle in the power yard] and the locomotive stalls until the decoder can find the signal again.  If the decoder set to "On" as a default, then all of your locomotives would be full blast on all the time, so we can pretty well deduce the default value is "off."

You can add a spike to the beginning and the end of the wave, but now you're just adding more problems to an already troubled mix.  That spike could be interpreted in a number of ways, depending on the interference [dirt] that is encountered.  Best we just eliminate the dirt altogether.  Most people suggest we should slave our lives away to essentially micro-polishing all our metal interfaces and then nickel plating everything too, when those same brass wheels ran for decades without a single issue back on old reliable DC power.  I suggest we eliminate the issue by bypassing the dirt altogether and go into the air - a much simpler solution.

Not only do we have the rail to wheel interface to contend with, but we also have the axle/wheel to "chassis" pickup.  And my electronics - be they the radio in my truck, the Motorola XTS radios, my cell phone, my laptop reading my wireless router from 100 yards away, even just the simple 900 Mhz house phones, these devices have all proven to me beyond a doubt that the air is Far cleaner, even with all this electronic noise, than our contact interfaces.   You can tell me there's all sort of "dust" in the old dirty air, but "dust" would be a welcome imp in exchange for this devil "mud" we contend with on the rails!  It's like you're telling me I should be happy with my water that is brown and soupy because the water that is tinted dusty or even just clear [to the eye] still has "dirt" in it...either of the latter would be a welcome exchange int he right direction!!!

The Cleanest interface would indeed be a solid connection between the command station and the locomotive parts, hence why I dare say the "ultimate DCC" system is one that puts the Command station Inside the Locomotive where you would upload the entire route to the Decoder as a computer program that then manages the entire session for that train.  This wouldn't be very exciting for those who want to drive the train, but it is the most stable format and it would be very useful for AI controlled trains on a layout where there is not enough engineers to run all scheduled trains.

Now you may ask, do I have any practical proof that this will solve the problem.  And I do, I dare believe.  There is this one locomotive at the club that is a number of years old, it doesn't have the best pickups;Ii hate running it alone.  Couple it with a second engine, though, and it's nice power.  I have found this true with just about all locomotives, they simply run better in tandem than they do alone.  And this, I dare believe, is because any time either decoder loses contact [that 100 millisecond moment] the other unit is still in contact, and it effectively pulls the "dead" unit back into contact with the command station.  Since the unit has good flywheels and sufficient internals, it coasts for quite a distance when it shuts off, so it doesn't affect the second unit all that much when it hits dirt - though a dead spot may lead to a momentary "drag."  It could be the power itself losing contact, but it seems to me it's the communication that dies first.

And I just had another dangerous thought; anyone remember those old crystal radios?  They didn't need batteries to operate, seeing as how the radiowaves had all the energy necessary to make the radio operate.  We could do something similar for our decoders, hence the decoders themselves would not need power to function; they'd need power to run the various systems attached to them, but they would not need it to communicate.

My hypothesis; go to the air, clean up the issue altogether, and we can stop this micro-polish silliness - and save all our elbows for more meaningful work besides cleaning track!!!

Electronics 101

Benny,

I take my hat off to you. An excellent explanation of how DCC works.

I learned about DCC way back when Keith, I can't remember how you spell his last name, produced his first CTC-16 system. I bought the bare boards and built up the system. I had the biggest problem with the receivers. I had bought 16 boards. The first one I assembled didn't work, it was a disaster. I did look at the command station output with o'scope and the signals did exactly as you describe in your post. What turned me off from DCC from that time on was operating on a guys layout that had this system in place. Seemed like there were a lot of problems form what I remember.

And you speak the truth about going to the air. I believe the signals should be sent over a 2.4Ghz frequency such as the fly-boys use. They don't seem to have any problems with "dirt in the air". Spread spectrum works for them. This system can have more channels than DCC has at the present time.

Bernd

DCC - AC? Yes, by your own definition

Benny

While you argue persuasively, your initial definition torpedos you:

We are talking about what is present on the rails, no?  Consider: when observed with an oscilloscope with two channels, you will find that one rail is at Vmax, while the other is at Vmin at any instant in time; they alternate between the two levels.  Their absolute value relative to ground, or anything else but the other rail, is irrelevant.  What the decoder input sees, the voltage signal on the rails, is the voltage Vmax-Vmin.  At one instant in time, that value is positive from rail A to rail B.  At some later instance, it is negative from rail A to rail B.  If the decoder is consuming current, that current will alternate in polarity from the viewpoint of the decoder.  This alternation is the crux of the above-quoted AC definition - alternating current.  I will grant you that if you choose the right reference point, you might show that the two output waveforms are always positive (Vmax) or 0 (Vmin) relative to that reference point, but what I care about is what is applied on the rails as received by the decoder, no?

Every DCC decoder I am aware of has a rectifier-capacitor circuit for provision of filtered DC,and ultimately a regulator to provide smooth, stable power for the microcontroller at its core. Generally, the filtered unregulated DC is then switched appropriately by the microcontroller to provide DC with appropriate pulse width and polarity to turn the motor as requested in the  encoded commands.  Regulating the voltage to the motor as well as the CPU would really increase the complexity of the circuit, so it is not done (or rarely, anyway).

As for whether manufacturers oversimplify the output and call it DC, probably in order to make it feel more comfortable for those transitioning from pure DC or pulsed DC systems ("my prized loco is DC, my DCC system is AC, I must have to replace all my engines"), and possibly to differentiate from the Marx crowd ('cause everyone knows Marx sets and DC sets are incompatible), its' irrelevant.  It looks, smells, feels, and tastes like AC, and to boot it fits your definition of AC.  Varying pulse width, granted, but alternating current for sure.

Operating a DC loco on address 0 is accomplished by varying the pulse width of the two rail waveforms to introduce an average DC component.  Without that, the unit doesn't move, another good sign that the rails must have a purely AC waveform.

Blair Smith