DCC Distribution comments, please
Blair & Rasa
I have a fairly ambitious layout a-building, and it's time I considered DCCtoggle quoted messageShow quoted text
power distribution. I have posted a PDF to the files section of this group,
named "ACR 20040402b DCC.pdf", to give you an idea of dimensions,
constraints, etc. What I'm looking for is creative ideas for power
distribution within the room, presuming a single level layout. I'm trying
to keep busses to a max of 30' from a DCS100, but that rule isn't firm. I
believe the room could be served by two DCS100 (5A), located at A and B, in
terms of my expected traffic loading - up to 15 HO diesels running at any
one time, with another 30 in staging tracks). Each DCS100 will feed at
least one PM42. However, I'm concerned about total feed length around the
LHS. I may be better served with three DCS100, one each at A, D, and E.
(the link shown from A to B represents the availability of a below-floor
duct, but to extend power buss across this is silly - total length consumed
is about 12', to what end?)
I'm sure I haven't given enough info, but please just ask away.
From: Ulrich Albrecht [mailto:albreuf@...]
Sent: March 31, 2005 12:49
Subject: Re: [WiringForDCC] Digest Number 8
I did not want to scare anyone off DCC, but would actually encourage
everyone to take this route. I would only recommend that everything is
working fine under DC so that the conversion experience does not become a
pain. All too often, one blames DCC for problems that are actually old
My main concern is that the wiring needs to be able to ahndle the loads.
Yes locos use the same current under DCC as under DC, but on a DC layout
running six engines this load is distributed between 3-4 power packs,
while under DCC all may go through one wire at a certain point. This may
result in poor performance due to voltage drop or overheating of the wire.
On 31 Mar 2005 WiringForDCC@... wrote:
Yahoo! Groups Links
Thanks for posting the detailed drawing. I will think about your
layout some more over the weekend. I'm sure some of the other people
will have a few comments for you. Here are a few of my initial
You seem to have a good grasp of the "30' rule," so I will assume you
are familar with my website.
Have you considered punching a hole in the wall to run your bus
through it? That would be one way to shorten one of your buses.
In deciding where your boosters will go, you need to think about the
loading on the boosters. Are you likely to have one or more places on
your layout that will have a concentration of running locos? That may
not change any of your plans, but it is something you need to
consider. I suspect you will need a DCS100 and 2 boosters (or
DCS100's operating in booster-only mode).
Blair & Rasa
Yes, I'm familiar with your website. The hole in the wall for wire is
possible, but as the offending wall is 10" concrete block, I'll only do it
if it saves me significant effort in wiring. As an electronics tech, the
wiring doesn't intimidate me, I just want to check my ideas against the
thoughts of those who have done this more than I have.
As for the DCS100 vs Booster debate, money is a concern as it is for most,
but I also see advantages to having equipment commonality, as opposed to
multiple different items, so it's a tradeoff.
I should give you some data about the layout as it is presently envisaged;
I'll post a schematic later this PM.
Blair & Rasa
Schema posted last night, and previous file moved (one note - the schema
posted uses letters to designate connections; they are not related to the
letters used on the floor plan). In the schema, be advised that the yards
depicted vary from 8 to 12 tracks in width. The schema also doesn't show a
couple of important details; the CNR and CPR will either both have, or both
have access to, a reversing loop to allow trains in staging to be turned for
reuse in the same, or subsequent, sessions. The loops will be embedded in
the peninsula that starts at E. In addition, the future existence of a
helix at D and engine turning Y tracks at A and B were omitted.
I will be wiring the DCC bus with 14/2 Romex, stripped and twisted; each run
from the DCS100 will either run direct to the layout uninterrupted, or pass
through a PM42 or equivalent power distribution centre; other than a few
T-joints, which can have the main bus uninterrupted, there will be no breaks
or other connections to introduce higher impedance in the busses. However,
I am likely to have feeder stretches up to about 24" (actually 48" of wire).
Feeders will run to every piece of track and switch, I'm not fond of
soldering joiners unless I really need to for physical alignment.
Considering all of the above, I think my best bet is to locate boosters at
D(feeds BC, helix) & E (feeds BFG) , with the command station to the right
of A (feeding left of A thru to H). The duct (A-B) will be used to feed
loconet signals from A to B to D/E, but not a DCC bus. Of the three
locations, A will be the most accessible, so it is a good spot to put UR91,
Does anyone have any experience with the PS2012 from Digitrax? It looks
good, but I wonder about, in my situation, running a long DC bus from A to D
and E to get the power to these boosters. It may make a lot more sense to
have localized DC supplies than to buss that around the layout.
Turning to the loconet topic, the layout will be DT400R/UT4 operated, with
perhaps a Zephyr for one or more of the yards (a friend's layout is Zephyr
controlled, and he would like to utilize it on my layout so his young
daughters can participate. my guess is once they use a UT4, this issue will
go away, but...) In order to control the yards, it would seem to me that I
want loconet plugs, at a minimum, at each yard. However, are there other
good reasons for wanting them along the right-of-way? I can't see it. I
suppose one long run around the layout from yard to yard could always be
broken into later to install UP5 panels, but if I can, I'd like to avoid
Funny, the process of writing this down for the list and thinking it
through leaves me with a workable plan, I think. If anyone can see
pitfalls, or better approaches, please fire away .
Regarding PS2012: I prefer to have a power source with each
booster/command station. Given the cost of individual power sources
and the PS2012, there really isn't a cost advantage either way. The
PS20212 surprises me as a product because I thought Digitrax was one
of the manufacturers that advised against using a common power source
for their boosters. Note that the Digitrax command station and
boosters can be powered by AC sources. DC is not required.
Loconet plugs: The only reason to have more Loconet plugs is if
someone comes over to operate that doesn't have wireless capability.
Loconet: There is no practical restriction as to how long your
Loconet can be. You can run it anywhere, anyway you want to.
Feeder length: Feeder lengths of 24" is not advised, but if you can
pass the short test (see http://www.wiringfordcc.com/track.htm#a16)
you can do it. You might want to consider branching off your main bus
with more 14 AWG sub buses to get shorter feeder lenghts.
Blair & Rasa
See individual responses below. By the way, thank you both for the effort
put into your website, which I've been following for at least two years now,
and for this new group. Just pulling this together so I could ask some
questions has served to both focus my thinking on the bus issues, and
address some misgivings I had.
Regarding PS2012: I prefer to have a power source with eachThey've said in their manual that this is because each booster should have
it's own current protection device, which seems to be why they also include
a Y-adapter (including current limiters) to feed multiple boosters from the
PS2012. Anyway, I'm in your camp for other reasons; home-kitting
half-a-dozen 5-Amp boxes with transformers, fuzes, switches, idiot lights,
and suitable connection means is a no-brainer, and then I know what I've
got, and it's modular; not to mention ease of repair.
Note that the Digitrax command station andyep. And transformers are cheap. I've never seen a schematic for a DCC
booster, but I presume the first input element is a bridge rectifier (this
permits use of either polarity of DC as well as AC for feed - maximizes
robustness); if so, putting a bridge in the transformer chassis would be
redundant unless the transformer chosen is high enough in voltage that we
want to shed a few watts OUTSIDE the booster. BUT, do boosters run cooler
with DC rather than AC input? Does anyone know? It doesn't seem reasonable
from a transformer-diode-filter cap type of power supply point of view, but
I presume they're using switching technology given the size of the box for a
DCS100, so I don't know if there are differences in efficiency between
sinusoid AC and smooth DC input.
Loconet plugs: The only reason to have more Loconet plugs is ifOkay, hadn't considered that, but it's my layout. If I'm willing to
relegate visitors to yard ops unless they bring a radio throttle, that's my
Loconet: There is no practical restriction as to how long your
Feeder length: Feeder lengths of 24" is not advised, but if you canOkay, theory discussion follows. Is it better to put a joint in the
high-current main bus, or a bit more length in the
feeders? Thinking about it, the resistance of a copper-solder-copper joint
has just got to be up there in the same league as several feet of smaller
gauge wire. So unless I'm off-base, and I haven't actually checked this
with the a milli-ohmmeter, it would actually not really matter which way I
went. The primary concern is the coin test, and that's up to me to try. As
far as I'm concerned, any branch in the main bus is at least as much of
concern as multiple feeders, because the main bus joint will see a larger
proportion of the total current being delivered to the loads.
Deep River, Ontario, Canada.
Where there's still 18" of snow in the bush, and a good month for model
railroading before the yardwork begins.
Even if you branch off your main bus with more 14 AWG, one joint
isn't going to hurt you. My main bus passes through several screw
terminals before the power arrives to the track. Bottom line, if
you can pass the quarter short-circuit test, you are good to go.
I wouldn't think whether you used a DC or an AC source would
matter. If a device can operate on AC, then if you use a DC supply,
it will still pass through the same circuitry inside the booster as
if you used AC - generating the same amount of heat. We will have
to see if anyone else knows any differently. What I think is
important is that your source, whether AC or DC, not put a voltage
that is significantly higher than the anticipated track voltage.
The differential will certainly be dissapated as wasted heat in the
booster. If you use the PS2012 for example, don't set it to G-scale
and then run your boosters set to HO.
I think you are fine with your plan on Loconet jacks. Even if you
have friends come over, I think most people have gone wireless.
Everyone I know in my area has done so. Besides, if you have enough
throttles to provide to your operators, you don't have to worry
about people coming over with a non-wireless throttle.
Having the jacks near the yards and such is good for two reasons.
1. So you can assign engines. 2. There appears to be a slight
delay when using wireless. So some of us do our switching plugged
in to a jack.
Blair & Rasa
Agreed. However, I see 12 VDC and 12 VRMS as potentially very different
inputs, depending upon how the booster deals with the input. If a rectifier
bridge and filter cap, even if small, are used, then the 12VRMS will result
in 15-16 VDC in, vs 12VDC from a 12 V power pack. So I think we need to
know more about what the booster does with AC vs DC input power. The real
proof is if we run 5A of load on the DCC bus with different voltage inputs,
AC and DC, and measure how long it takes for the DCS100 to shut down in
disgust. (big grin - try this with caution, folks, it's not smart to let
the magic smoke out of the DCS100).
However, I'm not generally in favor of running into walls just to make sure
they are solid, nor would I run a booster at full rated load just to check
dissipation. If the rail voltage we eventually have is high enough for
full-throttle satisfaction, so be it.