DCC Extras

In message , Mickey writes

This is one of the Frequently Asked Questions (FAQ) on the ZTC site.

Assuming you are operating OO gauge then you will need to fit a power booster, when your track gets to approximately 300 to 350 feet in total length, to avoid data corruption due to the capacitive loading of the track work. This figure includes all sidings and run around loops etc. Data corruption usually manifests itself in a variety of ways all related to erratic running. Locomotives are also slow to respond to commands sent to them and in severe cases can even run away for no apparent reason.

My garden railway has 1200 feet of track, so according to this formula I would need 4 boosters.

Also, remember that the decoder in a locomotive uses about 25 milliamps when the loco is not running, so if you have a lot of parked locos they will drain the available power somewhat. When they are running they can take much more current than this, so in order to calculate how much output you need you should add up the current requirement from the maximum number of locos you are going to be running at once together with those that are on the track but not running, and divide that by the maximum output of your boosters, which will tell you how many boosters you will need.

consider what you are running and what lights and accessories are coming from the same power source. This will determine whether you should wire in a new power district. Hope this helps, Steve

Cheers for the replies,

Mickey

The limitation before requiring boosters is the number of trains. Your DCC unit will be limited to 2 or 3 amps while your trains will each draw 0,25 to 1 amp each, even more if you run lighted coaches. That could mean anywhere between 12 and 2 trains maximum. If you're likely to have for example 3 heavy lighted trains in your main station then you will need to add boosters and subdivide blocks but if you operate less than 12 branchline locos on your layout you won't need boosters even if your tracks run all the way around the local park.

Gregory Procter wrote:-

Don't go giving John Sullivan ideas :o)

(kim)

It depends on a lot of things - scale, bus wire size, number of locomotives, and their current draw.

N-Trak in the US is suggesting (via an N-Trak Reccomended Practice: See

that a layout have a booster FEED to the busses in the middle of every 80 feet of track. Depending on your exact physical configuration, this MAY not mean a booster for every section. If the appropriate point can be reached by a bus feed with little voltage drop, an additional booster may not be necessary. Particularly with a home layout, simply feeding power from another channel of a power manager device such as a Digitrax PM42 may be sufficient.

To this end, our N-Trak DCC RP strongly suggests 12 AWG wiring for the bus. Yes, this is big wire. Most folks are using the 12 AWG zip cord now sold in home hardware stores for low voltage landscape lighting applications. At 12 volts, voltage drop over long runs of smaller wire can be enough to make DCC unreliable, and this doesn't even take into account the losses from connectors. (Remember that the total run of wire is TWICE the distance from the booster - the distance out to the furthest point AND BACK to the booster.

Good quality N scale locomotives from outfits like Kato and Atlas can generally be relied on to draw .25 amp or less. (your mileage on Grafar/Humbrol/Bachmann/whatever may vary!) This means a 5 amp booster can run 20 locomotives at a time, or fewer if you're a fan of lighted passenger cars. BTW, those passenger cars are one of the major design considerations in that 80 foot (What, roughly 25 meters?) limit on power districts.

More specifics about the layout you have or have planned would help.

He'd lose weight running around the local park! Does he need to lose weight?

Regards, Greg.P.

Thanks everyone, very helpful to me.

Mickey

Phil: As a long-time user of Zero-1 and now DCC, my observations:

1/ Although in theory, and in normal operation, there is only '1 circuit' needed for dcc (/other command control systems), there is an adavantage in being able to isolate sections for fault finding...

..if converting an existing, sectioned DC layout, to DCC, simply connect the DCC controller in place of 1 existing DC controller, and switch it on to (ALL) sections, at (ALL) times, unless there is a requirement (eg to run more than 1 dc loco) to have the sections switched elsewhere.

(Oviously, any section of track must be switched to DCC, or OFF, or a DC controller, but never both at any time - I originally bought 30 x 4 -way mechanically interlocked (RadioBAndstyle egLW MW FM SW) switches to select between any of 4 DC controllers - 1 per 'section')

If a short circuit occurs anywhere on the track, the ability to 'turn off' sections will aid identification and rectification. An obvious error would be a derailment or overunning a point set the wrong way (**see below).

2/ Reversing loops are NOT a problem, and are simply handled automatically, by the 'appropriate box': In our LGB, MTS layout, this is a small module located by the section, which will automatically reverse the connection from the adjacent track / distribution supply, as required. For most OO implementations, the 'booster units' act as both a power supply, and auto-reverser if required.

For my imminent rebuild, I am intending to use maybe 3 'power districts': due partly to length (I was having to modify Zero-1 point modules as I added the 25th-30-th modules, due to loading on 1 bus / circuit) and partly for fault-finding / isolation, and also because I will NOW be including reverse loop which I previously avided like the plague! (And also coach lighting, and 'quiescent current' from lots of loco modules just parked on the track - currenty about 0.5Amps for locos / lit coaches only)

A 'Power District' is the term used to describe the area / amount of track supplied from a single controller/booster/dcc base unit. Typical output limits are 3A or 5A per unit (district) (some 10A available)

(>> **see below) If running points off the system, and a short circuit occurs in a 'power district', then point modules/decoders, fed from that power district will be unable to operate.. as they will have no power... therefore the faulty pint would have to be changed by hand!!!)

If running points off the system, the POWER to operate the points is usually taken from a simple ac supply / cd discharge supply -- wheras on Zero-1 this was ALSO from the track (and why I ran separate busses for them)

rgds Phil

In message , Gregory Procter writes

Yes he does!

In message , Joe Ellis writes

Are you forgetting that with AC the rails act as plates in a capacitor? See the quote from ZTC's FAQ I posted earlier.

No. Are you forgetting that the DCC signal ISN'T AC, but rather bipolar DC?

Frankly, I don't see what I said in my initial post that even bears on that issue. Perhaps it's because you don't know what a "power manager" is? One type is the Digitrax PM42. (see

select "Power Management" from the menu on the right, then "Digitrax Power Management System-PM42" from the list of products). This lets you run as many as 4 power districts from a single booster, at a rather large savings over buying 4 boosters! As I said before, how you approach this problem depends on a number of factors, including the physical size and configuration of the layout in question. It is VERY possible to run a layout with 500 feet of mainline track on a single booster, IF the configuration of the layout is suitable and the bus wiring is up to it.

The referenced article above formed the basis for the DCC layout design at the N Scale East show in Chantilly, Virginia, USA this past August. (It's the SHORT version - the full DCC planning document ran to about 75 pages in Adobe Acrobat format!) In that layout of well over five HUNDRED N-track modules, about 5/8 of them were on the DCC red line route. (See

for details)

The decision has already been made that for the next event, the _entire_ red line route (a full circuit of the layout) will run on DCC.

That track around the perimeter of the park could be the answer! I'd suggest 12 volts on the track and a hand held IR controller operating a receiver in the loco tender might be more suitable than DCC.

Which IS AC, albeit with a DC offset when measured from "ground". As far as capacitance is concerned, it makes no difference.

So it's AC, but it's not. Whatever.

Phil: Offset? - only when in a braking section, or running a non-equipped loco at other than zero speed (where the 'offset' is created by controlling the ration of 1s and 0s sent so that the lowpass filtered (dc) element becomes non-zero, and the loco moves. (The motor windings acting as the filter, and doubling up as a loudspeaker - hence the buzz**)

**This seems to be how the old (DC) codar controllers added a steam whistle to a loco back in the seventies.. a superimposed hf 'sound'.

The DCC Track waveform is a SHAPED*** datawaveform: it is therefore an ac signal, and in normal use, has NO DC content (except for loco 0 as above).

The ALTERNATION is at the datarate: the transitions occurring at either of

2 spacings, representing a LOGIC ZERO or a LOGIC ONE - basic datarate approx 10kHz.

This is A/ FULL WAVE RECTIFIED by the decoder, to provide DC power onboard which is then regulated, as required, to control the motor speed (USUALLY by pulse-width drive at frequencies from 50Hzish (old buzzy modules) to

32kHz (latest sound modules from ESU Loksound) - ehatever drive frequency the decoder uses is totally independant of the datawaveform (and that is where it is VERY different from the Zero-1 method, which was therefore FIXED at a local mains frequency 50Hz or 60Hz (which would have been a big problem in Japan, which uses both!!!)

This is B/ Read by the Microcontroller onboard the decoder (which was powered bty the rectified track signal), and used to control the functions and (pulse-width) drive to the motor.

*** Shaped: It is not 'TTL Logic' is shape, but rise-limited / slewed / filtered to remove very high frequency components which would otherwise interfere with other equipment. It is not a Sinewave, either, but this is not a requirement for an 'ac' waveform 8-)

OFFSET - Braking - this is one method used, and supported by some more recent decoders, to initiate an automatic local (in section) braking mode, with preprogrammed stopping distances etc - created by simply adding a diode drop in one direction on the local track waveform - the assymetry is then detected by the decoder, and applies the brake.

Phil Spiegelhalter wrote:-

I couldn't help noticing at the time that the Zero-1 system was introduced very shortly after series of articles on model railway control were published in Wireless World. Does anyone know if the Hornby system was based on the information in those articles or was it a completely different principle altogether?

(kim)

At least with Digitrax and Lenz systems, it's not the "ratio" of ones to zeros that's changed (which whould change the signal on the rails!), it is achieved by "stretching" or lengthening the "0" signal on one rail or the other so that a "perceived" voltage difference is applied to the the DC motor. When "equal" zeros are present, the DC motor vibrates rapidly at a very high frequency in equal amounts. (this is the "buzz) As the "zeros" are "stretched", a bias develops in one direction, and the motor begins to turn. The more the zeros are stretched, the faster the motor turns. The pitch of the "buzz" also drops, and the volume decreases, as well. However, this also slows down the transmission of data, and it can be a _significant_ slowdown with several DCC locomotives on the rails and an analog locomotive run at or near full throttle. This can be observed first hand by setting up several MUd locomotives, then turning up an analog locomotive to full throttle. (You don't need to actually have one on the track). Turn the lights on, and reverse the direction of the consist. You can see the delay as the directional lights change on the locomotives. Turn down the analog throttle and compare the response time. This is why most clubs do not allow operation of analog locomotive on DCC lines at public shows or during operations.

Once again, it is NOT an AC signal. A true AC signal runs, for example, from +12v to -12v on a single rail, in a sine wave. The DCC signal, though, puts out 0 to +12v on one rail, and 0 to -12v on the other, using a digital square wave. It's a seemingly minor but very significant differance.

Phil: Wrong! - It is a bidirectional current signal, which with 2 wires is the only way it can work - and that makes it 'ac' - place a diode in the way to rectify it to dc, and it will stop working.

The only voltage measurement relevant is that between the two rails - and these are alterantively positive and negative with respect to the other rail,.... that AC.

Within a decoder, they are full wave rectified to produce 2 more voltage rails - but those are internal to the decoder, and not relevant to this discussion.

the polarity is reversing at abot 10kHz.