Too dry and you get lots of static - bad for all our modern electronic devices, including DCC systems and decoders. Here it varies - winter time it's 10-20% humidty, summer it's 80+. Winter time is static death, every time you touch something metal you cringe expecting the shock. Poor cats and dogs, you go to pet them and they get shocked. In my previous house I ran a dehumidifier in summer and it ran pretty much continuously (hooked up a drain hose to the condensate pump for the AC so I didn;t have to lug buckets of water), just to keep things around 80. In winter, running a space heater kept the temp high enough to get around 40% humidity, and there weren't any problems.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
Sounds a bit like Florida weather! I remember being sent to the redneck riveria near Pensicola for some environmental work and my hotel window was totally fogged over when the A/C was running inside due to the condensation outside. In Indiana the humidity would be so high there was a haze of moisture in the air, you felt like you could cut it with a knife. I recall my father talking about 100% humidty in the philipines, guam in the south pacific when he was in the air force. But isn't 100% humidy when the air is so saturated that it condenses out and rains? At least that is what I recall from my meterology class in college. All this talk of humidity makes me long for California where the air is nice and dry in the summer, like 10-15%!!!!
Rio Grande. The Action Road - Focus 1977-1983
Crandell and Riogrande,
Although off topic, I agree that the effects of humidity are more significant than thermal expansion. Crandell mentioned humidity of 75% over the summer. Well, in Sydney, Australia over the weekend we had 100% humidity on Saturday and 94% on Sunday. In fact, the average annual relative humidity in Sydney is a whopping 69% at 9:00am - source: http://www.bom.gov.au/climate/averages/tables/cw_066062.shtml
I have a dehumidifier in my train room running 24/7. It takes about 8 - 9 litres (20 pints) of water out of the air each day !!! I would not believe it myself if I hadn't collected some for the neighbour's fish tank.
As I'd had a mould problem in the room before it ever became a train room, I also decided to paint my base boards on all sides before starting my layout. I thought a matt finish might be sufficient, but it wasn't. My base boards are now fully coated in black gloss paint. Without the dehumidifier, my track was getting dirty very quickly. Nickel sure doesn't like humidity.
I now manage to keep the humidity down to around 45% most of the time. The equipment is costing me around $10 per week to run. But it is saving me hours of track cleaning time.
Cheers
Nige'
Nige, someone figured it out several years ago and found that 100 contiguous feet of NS Code 100 rail in HO would expand a whole 1/4" over a 30 deg F rise in temp. Nothing much, really, three or four sloppy gaps will look after that around a layout.
However, it isn't that kind of expansion and contraction that affects most layouts. Instead, it is the linear expansion of the wood framing due to changes in humidity. If one builds during a summer where ambient is near 75% indoors (not much air treatment going on), and then winter heating reduces ambient in the house to 40%, the wood will shrink. It takes the roadbed with it, and in turn the roadbed wants to compress all the joints in the rails stuck to it. When compression room is taken up to the fullest extent, the pressure will often result in broken or bowed rail segments. Happened to me once.
The wisdom is to solder the joiners on curves when using flex track to keep from developing kinks, to have two or three small gaps here and there, and to otherwise leave some joiners in the tangents free to slide...unsoldered. That practice seems to work for many of us.
Crandell
Nigel, Thanks for the look at physical properties and reaction to temperatures!
My biggest problem was in a garage layout I had in Indiana where the temperatures could be below zero F in winter and as high as mid 90 F in the summer, with a good deal of humidity. So those were on the more extreme end of things. The layout I had was HO and 16 x 19 feet in a hollow L shape. When I laid the track, temperatures were moderate, and at worst I saw the track push out on the curves due to expansion, and in winter the track pulled apart at the solder joints due to contraction. While my situation was rather extreme, I did learn it is a good idea to leave a thin gap when laying track and to allow the track to float. If you have contitions which are maintained at a fairly steady temperature and humidity, then soldering the rail joints may not cause much trouble.
A lot of good points in those last three posts. So I guess there are more of us soldering to rail joiners than we thought. However, it did raise a couple of questions:
@Harold - I'm curious to know why you run your buss wire around the outside of the layout? My wiring diagram would be more described as a star shape, with several power districts radiating out from the PM42 power management to BDL168 occupancy detectors. From each BDL a series of buss wires radiate out for individual blocks. It sounds to me that you have one big loop of buss wire, right?
@Randy - Yeah, if I used flex I'd probably go for every rail joiner too. Except for turnouts - I've double gapped the branch ends of those as I plan to rely on the power routing of the turn out measure the current for turnout position feedback ... but that's a future project.
@Riogrande - Yup ... expansion is a good point, particularly on big layouts. It sure doesn't affect me on my 6' x 10' N-gauge, but someone mentioned having over 2000' of track ... if you've got a layout that is 20' - 30' long then, depending on the temperature range, things will shrink or grow. But is that a problem?
The rate of expansion in Nickel (around room temperature) is 13.1 microns per metre per degree Kelvin. Sorry, I've got to work in metric, but I'll convert the answer to imperial at the end. Let's say you've to a nice large garage that you've converted to "the train room", with an "L" shaped layout that is 6m (19'6") along it's longest side. In winter, the room gets down to 5 degrees C overnight, but in summer it sometimes gets up to 35 degrees C ... This is a typical scenario in Sydney, Australia. The variation over the year is 30 degrees C = 30 degrees Kelvin. Doing the math: 30 x 13.1 x 6 = 2358 microns, or 2.358 mm or nearly 1/10th of an inch. As a percentage it is around 0.04%.
As Riogrande says, with a small gap between pieces of track, this expansion can easily be absorbed. Over 19' you're going to have around 6-7 lengths of track. Thus, a gap of 1/50th of an inch, or 0.5mm, is just about right - again as Riogrande suggests. I can't imagine expansion causing a problem here.
But if all that track has been soldered ... something's gotta give. Curves in the track will help resolve the problem, but a straight length of track around 18' long with soldered joints is going to give you a problem. Stuffs going to move a little. Will it cause a problem? Who knows?
Here is another argument for soldering feeders to joiners and not always soldering the joiners to the rail. Two words: expansion and contraction. I built a garage layout which was very suseptible to rail expansion and contraction and rail which was soldered at joints expanded and broke at the joints in many places. If you face this issue to a significant degree, then allowing the rail to "breath" expand and contract by first allowing a tiny gap between each rail (perhaps the width of a business card) and then not soldering the joints, but allowing them to "float" and slide in and out as the rail expands and contracts. If you solder a feeder to every joiner, this should give full power and signal through out the layout, as each joint of flex track (if you use that) will have power and signal entering the rail from both ends.
When I converted to DCC I removed all the old block wiring, ran the bus wires around the outside perimeter of the layout and reconnected the feeders. Now as simple as that sounds, here is my advice.
1. Use the same two colors for all the bus wires and feeders. I used red and black and made absolutely sure they were spaced 4 inches apart, never ran through the same holes in the benchwork or in any metal hooks.
2. I didn't use suitcase connectors - i soldered everything to the bus wires, ran wires from the bus to terminal strips, ran wires from the strips to the track and soldered the feeders to the existing rail joiners. If you can find a source for these strips, obtain them. There was an electronics company in town that went out of business and I was able to go in the there and take hundreds of strips, connectors and spools of wire.
3. Label everything, keep it in writing, and test every connection as you go for shorts. Also, if you have to gap rails, and eventually you will, make sure the gaps aren't going to close up on you. I have used insulated joiners and pieces of an old credit card that I glued in place and then shaped.
Now I have added two major extensions to the layout since installing the DCC system and have basicaly followed this same process without any problems.
If you haven't purchased your DCC system, do some research. Whatever system people may recommend can be a matter of personal preference. There are differences in the systems - some significant - so you will want to do a little homework. I won't even tell you that I prefer Digitrax Super Chief....... Also, if you are going to use sounds, which I don't, I think there are differences in some sound decoders.
Anyway, all the advice you have already received is spot on. What I am sending along maybe is a rehash, but it's just my way of doing everything. Best of luck.
There's never time to do it right, but always time to do it over.....
All my feeders are soldered to joiners. And I don;t solder every joiner - especially not those on turnouts. I DO solder joiners on two connecte sections of flex to make curves, so that there is less chance of a kink at the joint. I DO have feeders soldered to EVERY joiner - so a turnout ends up with 3 sets of feeders. I just make up a whole bunch ahead of any track laying and evey time I come to the end of a piece of track, I just grab a set of my joiners with wire attached and keep on going.
Second layout done this way, and no issues whatsoever, even after painting the rails - including the joiners.
NW PA Reader Also, earlier in the discussion someone wrote about soldering feeder wires to joiners instead of to the track. Is this plausible as it sounds like it would make it a bit easier to hide the wires going under the table?
Also, earlier in the discussion someone wrote about soldering feeder wires to joiners instead of to the track. Is this plausible as it sounds like it would make it a bit easier to hide the wires going under the table?
Ed,
Folks can poo-poo my comment all they like but I assure you, there is no reason not to rely on rail joiners for soldering track feeders. Also, you do not need to solder the joint.
The issue with rail joiners is the cross secional area of the material conducting the current and DCC signal from one rail to the next. Over a great distance, particularly with sectional track, this can influence the DCC signal quality. Rail joiners plenty of surface area in contact with each rail - usually more than twice the cross sectional area of the rail. It's the thin bit of rail joiner material spanning the two rails that can cause a problem.
Not my photo, but take a look at this:
http://2.bp.blogspot.com/-IrsQZmoIu-I/TxjNZ-ZDyVI/AAAAAAAAB1M/yH90z2J8zXI/s1600/IMG_7373.JPG
You must appreciate that soldering to the joiner greatly increases it's cross sectional area and therefore improves the transmission of the DCC signal across the joint. My tip is to not solder to the underside as this fellow has done, because you can't monitor it. Always solder to one side of the joiner so that you can inspect it over time.
I work in N-gauge using Tomix sectional track. A long straight piece is about 11" long. I use AWG18 or AWG22 single core wire for feeders. AWG18 is about half the width of an N-gauge joiner. I set out my feeders so that the signal never has to pass through more than 3 - 4 joiners. Absolutely no problem so far, with 120' of track and feeders every 5 - 6 joiners on long blocks.
Now, if you want to solder your joints then I agree you can't get a better result in terms of signal transmission and conductivity. If you want to solder to the rails instead of joiners, by all means be my guest. I'm not great at soldering. In testing during the early days I kept melting the plastic rail ties near the joiners. Maybe this is more of an issue with sectional track.
Therefore, I went with soldering the joiners. It is quicker and easier. Using solid core wire you can bend the stripped end into a neat "L" shape and solder the leg on to the side of the joiner. It looks neat, sharp, and can disappear with a lick of rust coloured paint.
Also, if there is a piece of track that has bad signal strength or conductivity, you can simply remove the rail joiners and run solder along the side to increase its cross sectional area, then put the joiners right back where you got them. I haven't had to do it, but a friend did with a DC layout.
From my experience soldering the rail joiners works just as well as other methods. Anyone else who has done it will confirm this.
Actually, I do use feeders soldered to joiners, but then I solder the joiners to the rails, so the feeder is actually soldered to two sections of track and is nowhere that can contact a flange.
Dave
Lackawanna Route of the Phoebe Snow
Actually, a great many of us find joiners to be reliable if, and only if, they are soldered. You'll find a lot of support here and on other sites for that procedure. And since you solder the joiners to the rails to seal that joint (prevents oxidation), it also provides an excellent bridge for continuity which the joiners don't do reliably left by themselves.
Secondly, since you are going to solder the joiners anyway, why not flatten the last 1/8" of a 22 gauge feeder wire, with its other end already soldered to the bus (if you need a bus) insert it into a gap between the inner surface of the joiner and its host rail, and solder it all up at the same time? Presto...a secure mechanical joint with continuity assured and oxidation prevented. Could it be simpler?
NW PA Reader Also, earlier in the discussion someone wrote about soldering feeder wires to joiners instead of to the track. Is this plausible as it sounds like it would make it a bit easier to hide the wires going under the table? Ed You will find that to be a minority opinion. You should never rely on rail joiners to supply power to a rail. Peter
Ed
You will find that to be a minority opinion. You should never rely on rail joiners to supply power to a rail.
Peter
I got my Digitrax Zephyr several years back and just hooked it into my old layouts block wiring and ran the trains without a problem. I didn't put in buss lines until I got ready to tear the layout apart and rebuild it.
Dr. Frankendiesel aka Scott Running BearSpace Mouse for president!15 year veteran fire fighterCollector of Apple //e'sRunning Bear EnterprisesHistory Channel Club life member.beatus homo qui invenit sapientiam
Phoebe wrote
"I'm glad I already have my Digitrax Super Chief installed and operating. If I was a new guy I would be scared off by the descriptions I'm reading here about how complicated and expensive it would be with all those expensive boosters and power districts."
Well, I'm that new guy and that's exactly how I felt reading some stuff about wiring for DCC. Seems like every other line is "DON'T DO THIS (OR THAT OR THAT) OR YOU WILL BE FRYING DECODERS AND SETTING YOUR HOUSE ON FIRE." Well I suppose with any electrical wiring you could set your house on fire, but it sure makes it seem pretty intimidating. Getting to understand that it doesn't have to be that involved makes it seem like an easier undertaking.
Hey gondola,
Yes, I did something similar except I used an LED instead of a bulb and my old 1970's Hornby controller (still going strong). The reason I used an LED circuit (LED + 1k resistor + schottky diode) is that the LED will only light up if the polarity of the track is correct. This is one step beyond simple short circuit detection.
As it turns out, like Randy, my buss wires are colour coded - Red and Black. My track feeders use the same colours and I simply set up a system of black = the outside rail and went from there.
From memory, the only wiring faults I had were pinning errors at the plug connectors. I simply de-pinned and re-pinned the connectors.
On a separate note, I'm impressed that you have 2100' of track. Sheesh. I've never worked mine out ... about 120' of N-gauge ... there you go.
DO get a few good books on layout building, track laying, and wiring.
DON'T take everything you read anywhere on the Internet or Internet Forums as the best of advice.
Elmer.
The above is my opinion, from an active and experienced Model Railroader in N scale and HO since 1961.
(Modeling Freelance, Eastern US, HO scale, in 1962, with NCE DCC for locomotive control and a stand alone LocoNet for block detection and signals.) http://waynes-trains.com/ at home, and N scale at the Club.
Gotta say I agree with Phoebe here. I have built two layouts with DCC now and have yet to hook a feeder up crossed or anything like that. But I do use two different color wires for the feeders and for the bus wires - and they happen to match. Red and white feeders, red and white bus. And since both were around the walls sort of thing, I kept the Red feeder to the Rear sonistently, and so only have to make sure only red feeders connect to red bus wire, and white feeders connecto to white bus wires. I built a bepper box early on, to use in the manner described - and it promptly ended up in the bottom of a junk box somewhere. After two moves now I'm not even sure whereit might be - I do know I did take the battery out to use in a jump throttle I built for m Zephyr, so I don;t have a leaky battery making a mess of other potentially useful junk.
I also made myself a decoder tester. And after testing the tester, have NEVER needed or used it. I should try to find that and recover the LEDs I used, at least. I think I did take out the load resistor, to use in the F7 dummy I put sound in, as a motor load so it would program. I've yet to have a bad decoder, and yet to burn one up after it's been installed - but since I do have a system with a low current program track, I DO always check out a new install on the program track before placing it on the full power main. If it doesn;t work on the program track - something is wired wrong, but the low current will keep the decoder from frying. Also, most of my installs bypass any factory circuit boards, which seem to be where all the problem is in failed installs. If I strip the loco back to some wires from each truck, wires from the motor, and wires to the headlight, I KNOW I am getting the correct loco wire to the correct decoder wire and there is no sneaky 'oops' on some circuit board that shorts together a track wire and motor wire but only when the loco runs in reverse or some other silliness.
I guess there's the DCC common sense - always use color coded wires. Always put the same color feeder on the same side of the layout, always test decoders ont he program track before applying full power, avoid factory circuit boards if possible. I dunno, it's worked for me for over 10 years now.
I'm glad I already have my Digitrax Super Chief installed and operating. If I was a new guy I would be scared off by the descriptions I'm reading here about how complicated and expensive it would be with all those expensive boosters and power districts. Only a few of the larger layouts need all that stuff. If it is not a large layout, it doesn't need to be any more complicated than a DC layout other than the need to install decoders in all the locomotives. Complexity, if desired, can be added later.
New people need to understand the difference between what you NEED to do and what you CAN do with DCC.
Good Morning here Nige'. I soldered all my track feeds to the rails and as I did my next section of mainline buss I connected a old transformer to the buss BEFORE it was connected to the existing one and wired up a 12 volt bulb to 2 alligator clips. I clipped it to my first solder joints and as I added the track wires to the buss it was lit and if I crossed one it would go out. I didn't have to crawl out from under the layout until I was done. It saved me a lot of time searching for that one single oop's and after over 2100 feet of track it was well worth the time it saved.
Colour code everything.
Solder your track feeders to the rail joiners, instead of the track - that way you won't melt sleepers, etc.
Even if you're not using Block Detection equipment, wire up as if you were. Include plug connectors so that you can disconnect individual zones for troubleshooting.
Don't wire up and connect the whole layout at once. Plan your power districts, zones and blocks and wire up in stages. Once you have your first zone wired, test it with your command station before connecting another zone.
Use slow motion turnouts - the biggest mistake I made was going for snap switches to save a few bucks. They're noisy and you end up spending more on stationary decoders to handle them. For example, to control 8 turnouts I need 2 x Digitrax DS64s. But with slow motion turnouts I could have used a single SE8C, and gained signal control for free.
Don't lay track and wire the entire layout without a short detector. Quite a few have done that and found out they had a short, somewhere. Get a buzzer from Radio Shack and use a nine volt battery to make a short detector. Keep it clipped to the pair of rails. As you move along, the buzzer will let you know if a short exist when you add the wires.
Others will have comments on laying track and wiring in a planned method. Reverse loops and cross overs are couple places where shorts can happen.
Rich
If you ever fall over in public, pick yourself up and say “sorry it’s been a while since I inhabited a body.” And just walk away.
By conversion, do you mean you'll initially run DC and then convert in time?
You'll have to convert your motive power to include decoders, you know that? Other than that, Nick, Just wire up the layout without all the blocks. As you get more than a total of about 100 linear feet of rails, no matter how they are configured, you would probably want to make a new power district. This figure supposes that you occasionally run several trains with several MU'd locomotives with sound.
If you need a district, it means a booster. The booster is linked to the main command station, but I don't know more than that because my Digitrax Clunk runs my medium-sized layout nicely. I believe a booster with Digitrax would need a data link cable for command signals to be amplified plus its own power supply. That means you now have two power supplies, one for the booster and one for its commander.
My Super Empire Builder's DB150 and its 5 amp supply run a total of about 180' of track: twin mains, a spur, a 25' descent to staging, and about 40 feet of yard tracks. My bus isn't long because it goes two ways from the DB150 for about 20 feet each. I don't run many engines at once, so that's well in keeping with sensible power management. If you run several consists of locomotives, my suggestion earlier may be better for you.