Tying Outside Rails Together

I think the only thing this does is give you better connections to the outside rails.  There's nothing about TMCC/Legacy that would dictate tying the outside rails together.  Our club layout has one rail hot, and TMCC/Legacy runs flawlessly on it.  The only issue I see with one rail hot is small motorized units with four wheels and traction tires only have one wheel for the outside track connection, that sometimes creates an issue.

Both outer Gargraves rails are jumpered together and connections are soldered. I run only MTH power units with DCS. Never any problems or operational issues even with old Lionel in conventional mode.

 A solid way to go I guess?

Our entire layout, all Atlas solid nickel silver track, has the outside rails tied together at every single power feed location. We run DCS, Legacy, and TMCC at any and all times, even together. Works fine.

 It seems any  engine with traction tires or a 2/3 rail engine is going to have stalling  problems around switches,crossover etc unless both outside rails are tied together.  This doesn't mean we   can't use insulated track sections, we  just have to be careful where we  put the insulated track pins.

Hot Water posted:

Our entire layout, all Atlas solid nickel silver track, has the outside rails tied together at every single power feed location. We run DCS, Legacy, and TMCC at any and all times, even together. Works fine.

The question was do you have to tie them together, I don't think anyone disputes that things will run fine with the outside rails common. 

gunrunnerjohn posted:

Hot Water posted:

Our entire layout, all Atlas solid nickel silver track, has the outside rails tied together at every single power feed location. We run DCS, Legacy, and TMCC at any and all times, even together. Works fine.

The question was do you have to tie them together, I don't think anyone disputes that things will run fine with the outside rails common. 

Oh, sorry. I didn't realize that you moderated here also.

I can't think of any reason why it would matter as far as the command system.  As GRJ pointed out, having the outside rails connected will give better operation for every system, conventional included, as small units and traction tires make for poor electrical connections.  Other than relatively small sections used for triggering various devices, I don't know why anyone would not connect the outer rails. 

So do they HAVE to be connected together?  Nope, but why wouldn't you?

JGL

I use Atlas track and only one outside rail is connected to common (same continuous rail). I have both DCS and Legacy and it works just fine here. My layout is not all that large (6'x16'). They layout was wired with DCS in mind, following Barry's book as closely as possible. However, I believe many here recommend connecting the outside rails together, if they are not already connected.

Hot Water posted:

gunrunnerjohn posted:

Hot Water posted:

Our entire layout, all Atlas solid nickel silver track, has the outside rails tied together at every single power feed location. We run DCS, Legacy, and TMCC at any and all times, even together. Works fine.

The question was do you have to tie them together, I don't think anyone disputes that things will run fine with the outside rails common. 

Oh, sorry. I didn't realize that you moderated here also.

I just point out when people fail to read the topic and post a non-answer.

Larger layouts need the outside rails tied together.  If you want to run signals, do it right with IR.   My statements are not from something I read or something I heard somewhere from someone.  I am talking first hand (hands on) knowledge.  I had the opportunity to have Lionel uses my layout for testing before TMCC was released and I had DCS on my layout a year before it was released.  I have been involved with some very big layouts that SMARTT and Rich Roman built.  

In the matter of attaching one or both outside rails to layout common, I'd say that there is no one answer in the sense that it must be one way or the other.  There are two considerations.  One is the operation of the digital (plus analog for the DCS FM sound channel) communications used by TMCC, Legacy, Lionchief, and DCS.  The second is safety; that is related to the 120v power system and its ground conductors.  In communications work in the real world, safety is first-- because in practice certain States here would not allow lifting of ground wire (to alleviate signal problems)... this led to expensive satellite ground stations being refused occupancy permits.  DoD adopted a matching policy.  I was involved in the engineering solutions to such problems.  Although by chance, I was a pioneer (although there were a few others) despite my motto, "Break no new ground."

That said, remember that the rails of a toy train layout are to be fed by a Toy Train Transformer (UL 697) that isolates both the hot and common outputs from all conductors of the home (or exhibit hall) power system.  That is, all  rails are isolated, or "dead-front" as you might say.  In the home, layouts in a concrete-floored cellar will present the risk of having an extended metallic surface, in the form of the rails.  Well, an unfinished concrete floor will usually mean bare steel columns and beams, which are bonded to power system ground under the latest codes, and therefore are an extended grounded metallic surface; the unfinished concrete floor is an extend surface usually having a potentially dangerously low resistance to ground.  These risks are greatly reduced in recent codes by ground-fault detectors (GFPPs, more often less specifically called GFCIs-- PP means personnel protection or 5 ma trip (10 ma is possible but unusual)) in the cellar receptacle circuits.

All that said, one has to realize that perhaps the continued isolation of the layout rails should be checked at certain intervals.  This can be somewhat less than intuitive for the average person, but knowing that TMCC has a disconnectable capacitive connection to power system ground (passing rather less than 5 ma under the impression of 120v at 60 cycles), that person might test for unintended voltage to ground (with a less sensitive moving coil voltmeter) on outside and center rail, alternately, for 120v connection; then use the ohmmeter connection to test for unintended connection to ground in a similar fashion.

Assuming that anyone actually does this    (meant for humor, not approval of ignoring this)-- one will realize that isolating outside rails increases the difficulty of checking for the continued isolation of all rail segments.  This would explain the position that Lionel takes, as noted above by Marty Fitzhenry.  I find this understandable, although I would attribute it more to a concern for safety by Lionel than something always indispensable to the operation of TMCC and Legacy.

Actually, all this concern for safety has mostly to do with the use of power tools in a basement.  Ungrounded old tools that become live, or doubly-insulated tools with insulation failure, are the documented sources of bad outcomes in the home unfinished cellar lacking GFPP.  I know of no documented cases involving UL 697 transformers.  But, unsafe tools require a grounded surface for a bad outcome.  That is where one wants to be layout-aware.  Or not use power tools around a layout (that was meant for humor ... I have only one gremlin choice today).

Well, there is safety put first in communication systems problems (I have outside documentation--1999--of that approach, which I used 20 years before that ...bookmark lost, as usual ).

Now, for those who must isolate an outside rail, I would say it would be good to know the following tidbits, assuming your cellar GFCIs are in place:

o The sump pump outlet usually is not on a GFCI circuit.  In past, a duplex receptacle was permitted; now only a dedicated single receptacle is permitted.  Do not plug in power tools or transformers.

o The lights are not usually on a GFCI circuit; in fact, this may be not permitted, in case of emergency at night (cannot recall).  Do not screw in your plug adapter (do they still make these) for your power tools or transformers.

o Know that GFCI's need monthly testing.  This is because they fail after a few years.  Accordingly use a second GFCI in line with your power tools and transformers (I have such an extension cord for my power tools, in addition to the outlet GFCI's, which is a standard construction industry recommendation).  BTW, how do you test the insulation integrity of a red-hot soldering gun? Answer: With one hand, because it is trigger operated (I'm getting punchy).

o In case you are involved with an AFCI (arc-fault interrupting breaker), know that those advertised as dual devices providing GFCI, these are NOT GFPP.  Their GFCI device shall not be more sensitive than 20 ma (milliamps) as it is intended to reduce fires in appliances but avoid nuisance trips. Of the only four approved manufacturers, GFCI trip can be as high as 40 ma; 20 is dangerous to people, and 40 is very seriously dangerous.  AFCI's protect against fire, not against deadly shock.  The argument is that reduction in fires will save more lives.  It is at present unknown if downstream GFCI's can be used successfully with them.  It may be case by case whether an AFCI breaker can be closed in each situation.  Be aware.

o As to the toy transformers, one will of course want to be aware that the RS232 connector on the MTH Z4000 has the house power ground on both the signal ground (usually pin 7 in the 9 pin sub-D but may differ here) and the shield braid if any.  Well, this was how the computer industry did it... a major source of trouble in satellite ground stations.  I didn't even have to think about the cause when the the first OGR board guy killed his TMCC signal when trying to use his DCS controller to issue TMCC commands.

Now, if one must isolate an outside rail for either signals or signalling, here are some reasons which might be necessary in some cases:

o If one runs DCS (including in conjunction with TMCC), one might have maxed out ones TIU's at five (20 blocks), and still have one of those 20 blocks too long at 175 feet to have signal all the way to the end.  Ones longest working blocks might be 135 feet, one center-fed and one end-fed, all wired by an electrician.  This might have been possible only because ones Gargraves track had the entire layout with one outer rail isolated to operate the indications of the trackside signals.  I refer of course to that famous MTH test track (built in part with their advice, anyway) located in NE Washington and oft visited by the OGR bus.

[The communication principle here is that the characteristic impedance of the transmission path is doubled (from, IIRC, 60 ohms to 120 ohms), greatly extending the range of a usable DCS signal.  I know this is counterintuitive, but Bell System built a record-setting 600-ohm open wire (#10s @10") circuit for unamplified telephone 900 miles between New York and St Louis.  Less is sometimes more... although some say the newly introduced passive loading coils at intervals were made by a guy who was also pushing magical light bulbs.]

o If one has a layout with, say, 3000 real feet of Gargraves track, one may have literally too much capacitance from rail to ground.  In this case, taking one outer  rail out of the TMCC antenna system may be a necessary change to get Legacy to work.  Although it may be thought that this is an impossibly large layout at today's prices, there are people with a supply of older, cheap relayer rail (as they say) and an ingenious method of both laying it curved the first time without kinks, which doubles as a system to recurve the old curves into new curves, still without kinks.  Well, at least there is one such person and this is a real example (on this board).

o I would suggest in these cases, the isolated outside rails each have a pigtail brought to a central row of small slide switches, which could be closed to layout common whenever a test of isolation from power system ground is to be conducted.

>Hopefully I've succeeded in being on all sides this question.  Remember, it is the local building inspector who has the definitive word, if it's in the code, which only the transformer and the wall warts are.  The layout is not, but connections to the ground wires of the house power system are (these wires can be dangerous at times).  My exclusive authority does not begin until the equipment passes above 400 amperes (at least in DC, at the time when the code was expanded from 225 to 400 ampere equipment there).  If you need 600 amps track power, I could speak more freely.  [Avoid 600 amps on the home layout.]

--Frank

JohnGaltLine posted:

Other than relatively small sections used for triggering various devices, I don't know why anyone would not connect the outer rails. 

JGL

On my mainline, I don't, because I use that second outside rail for detection over almost its entire length. Now on non-mainline track, I put a ground feeder on alternating sections of alternate outside rails. So with a little electrical contact help from the track pins, both outside rails are tied together.

Now you know.

Elliot wrote "On my mainline, I don't, because I use that second outside rail for detection over almost its entire length. "  What type of detection are you using that works over the entire length?

Thanks

Bill

Well, Lionel likes it that it way. Pick up any piece of track that that Lionel manufactured and the outside rails are connected electrically.

If you run a Lionel product on non-Lionel track and have unsatisfactory results, that's on you.

Same goes for track connections. Make them any way that you want to connect them.

It seems like all have missed my point. I was not suggesting to leave one of the rails as a dead rail. My point; is it really necessary to tie the outside rails together at every power drop? At each power drop I plan to alternate which outside rail is connected. In that way both outside rails are still connected to common.

Keith, the short answer is no.  I'd say, given that you're not trying to use the rail for signaling, that it's likely a good idea to tie them both together at each power drop. 

Marty Fitzhenry posted:

Larger layouts need the outside rails tied together.  If you want to run signals, do it right with IR.   My statements are not from something I read or something I heard somewhere from someone.  I am talking first hand (hands on) knowledge.  I had the opportunity to have Lionel uses my layout for testing before TMCC was released and I had DCS on my layout a year before it was released.  I have been involved with some very big layouts that SMARTT and Rich Roman built.  

Larger layouts certainly do NOT need the outside rails tied together.  Elliot (post above) and I both use one of the outside rails for signal detection.  At switches, I do wire both outside rails to common so that there is always a common-powered rail through each route of the switch.  The vast majority of my 550 foot main line and over 400 feet of staging tracks have only one rail wired to common.  I heartily disagree that IR detection is "doing it right" with respect to signals.  Some of my signals are set to display a stop indication if any one of 3 track blocks is occupied anywhere along their lengths.  You can't do this with IR.

My statements are also based on first hand knowledge since I wired the signal system myself.

ogaugenut posted:

Elliot wrote "On my mainline, I don't, because I use that second outside rail for detection over almost its entire length. "  What type of detection are you using that works over the entire length?

Thanks

Bill

Bill, it's just the classic "ground jump" of the insulated rail. I use GarGraves track, so the rails are insulated out of the box. This is one of the great advantages of 3 rail trains, and most people throw it away.

I'll add a DC ground signal to the AC ground, then when the train bridges the rails, the DC will go back through a diode and capacitor, then tell the computer where the train is.

This is just one of the detection panels, and all the wires hanging down by the legs are coming from the insulated rails. I'm in the process of connecting the wires to the panel.

Elliot

How long are the detection sections or blocks?  Presumably you are cutting the Gargraves rail or leaving pins out selectively.  Does the next detection block start immediately after the end of the first?  Are there any areas between blocks without detection?

Thanks

Bill 

For what it's worth;

My layout is 30' x 10' and all Atlas track.  I've connected both outer rails for at every power drop on my layout (approximately 36 power drops) and run 2 black wires (along with 1 red wire) back to the terminal block.  It added about 200' of black wire but has been invaluable insurance.  

My thought was; If I wanted to add anything using an isolated rail, all I need to do is separate a section of outside rail and use the wire I have already run to that section.  

The bonus is that I've electrically tied together both rails through the layout.

Gentlemen,

    If you want your Legacy signal to be strong use both outside rails, if you are running DCS use the out board Black Channels and then run another Legacy drop directly from the 990 Base directly to the opposite outside rail, I do this even with my FasTrack and it works perfectly.  MartyF is absolutely right, IMO this may not be absolutely necessary on some layouts, however I trust his experience & knowledge and when he makes a recommendation of this kind, I take that Engineering knowledge and run with it.  He has never led me wrong.  

PCRR/Dave

I tried tying the outside rails together but had a dead short??

Tom

Very nice, I had a similar plan with an around the room layout, but there were too many obstacles, stair case and windows.

Of course, you are kidding about tying the rails together.

Right!  three railers have all the fun.  What you are looking at is 13' 6'  X 29'.   Stairs down to the top right under the benhwork and no windows. I do miss the three rail wiring options.