Voltage drop

[kevinlms]

Some people might find it quite relevant, but it won't upset me in the slightest if you choose to ignore it.

Sorry, I don't understand your point.

 

I would never just rely on a fifty foot loop, without any bus and droppers and just relying on the rail and joiners. I wouldn't technically be ignoring nickel silver rail resistance, but overcoming it by default (and rail joiner resistance - which of course is MUCH higher) by using proper wiring techniques.

 

 

Edit to add.

 

David C Broad, highlighted the problem area earlier in the thread

 

"I find in practical terms there is no voltage drop on plain track indoors but there is considerable rail joiner drop and it increases (dis) proportionately with load. A four loco lash up on 2 X 1 amps controllers will set dodgy rail joiners glowing"

 

Note that unsurprisingly, the problem is with resistance at the rail ends (joining method) and has nothing to do with the actual rail.

Edited July 1, 2017 by kevinlms

[34theletterbetweenB&D]

I will use the good old trial and test it formulae and see how it goes...

 But what you will be testing is the new assembly from new or near new components. If the layout is to be in long term use, that's when the 'belt and braces' of running an all copper bus around each section with regular soldered connections to the rails, brings benefits. Every metal to metal contact (rail joiner to rail) has oxidation occurring on the contacting surfaces, increasing the resistance as the years go by; and while you could solder them up that makes future repairs or alterations something of a pain as well as preventing any expansion movement.

 

Now, if you are a 'build it, use it for ten minutes, tear it down and build another layout' type, ignore all the above. (I notice this polarisation in opinion between two distinct layout building groups, the other being 'build layout, use it 'forever'.)

[kevinlms]

 But what you will be testing is the new assembly from new or near new components. If the layout is to be in long term use, that's when the 'belt and braces' of running an all copper bus around each section with regular soldered connections to the rails, brings benefits. Every metal to metal contact (rail joiner to rail) has oxidation occurring on the contacting surfaces, increasing the resistance as the years go by; and while you could solder them up that makes future repairs or alterations something of a pain as well as preventing any expansion movement.

 

Now, if you are a 'build it, use it for ten minutes, tear it down and build another layout' type, ignore all the above. (I notice this polarisation in opinion between two distinct layout building groups, the other being 'build layout, use it 'forever'.)

Basing my opinion on the track plan supplied, the layout doesn't appear to be a temporary measure. So yes, short cuts are going to bite in the future.

 

 

FWIW, does anyone know how Peco wire their layouts at Pecorama?

 

I would be most surprised if they relied on Insulfrog points, relied on rail joiners and relied on the point blade pressure against the stockrail for electrical continuity. But perhaps they cheat (I have never seen Pecorama) and build a layout so that it looks like a standard layout, but the points on the public side are actually wired up solid and can't be changed. Thus taking away, almost all of the potential (sorry) problems.

[tigerburnie]

Interesting views and I am looking at them all, but as a former maintenance electrician(higher voltages most of the time), I'm sure I can come up with a cunning plan. The thing I did not think about was the joiners and their flimsy material they are made from, looking at the main loop from either end of the station, a supply link to each metre length is straight forward. The station is going to be a forest of switches and wires, unless I wire in a small PLC and knock up a program to run the whole thing by computer..............................

[Crosland]

 as a former maintenance electrician(higher voltages most of the time), 

 

Just remember that the voltage drop in a 240 V supply at 1 Amp is 20 times less, in %age terms, compared to the same current in the same cable in a 12 V system.

[DCB]

You don't have that many possible parallel movements so the complexity will depend on whether it is live frog or dead frog.  I wired a pretty similar station, 1 less road each direction but lots of headshunts and 4 parallel movements possible and it does need lots of switches, mainly because it is live frog.

Your diagram is fine for dead frog. If you go live frog you will need isolators on each through road or setting points "in opposition" will short it out.

Where to put the isolators is difficult as metal wheels bridge them.  DONT USE COMMON RETURN, on this type of layout with live frogs it causes horrendous complications.  Actually, you can't use common return with live frogs as you will need to isolate one side on the Up side of the layout and the other side on the lower.  My current MO is to use two pole rotary switches so any section can be fed by any controller, 

Indoors I also count fishplates and point blades, 6 sets of either from a feed is my limit.  That is a lot longer at 20 feet maximum and shorter at 18" minimum than other people recommend....

I hope you have some nice long 1/8" drills for all those long holes through the 3/4" chipboard baseboard...  (Unless you run the wires on the surface and ballast over them)

[tigerburnie]

Lots of good sound advice, thanks all, I am now thinking of looping each track join to remove any connector problems from either overheating or ingress of rust/ adhesive from ballasting etc. The main loops will be straight forward from a bus wire, the station will initially be fed from each end, but taking on board the comments about relying on points blades/springs for continuity, then switched feeds are likely for each passing loop. This will hopefully help with the wear and tear that I had not thought about at all.

[AndyID]

Code 75 bullhead rail has a cross sectional area of approximately 1 square millimeter. (Code 100 FB will be a bit bigger, but not that much bigger.) The voltage drop at 0.5 amps due to the resistance of the nickel-silver rail alone (no allowance for rail-joiners etc) is 0.136 volts per meter, or 1.36 volts over ten meters. That's quite a lot.

 

16/0.2 copper wire has an effective cross sectional area of 0.5 square millimeters but copper is around 17 times more conductive than nickel-silver so length-for-length the drop over the wire will be about eight times less - say 0.17 volts over ten meters.

[AndyID]

 DONT USE COMMON RETURN, on this type of layout with live frogs it causes horrendous complications.  Actually, you can't use common return with live frogs as you will need to isolate one side on the Up side of the layout and the other side on the lower.

 

HI David,

 

I've used common return on layouts like that, including many live frogs, and never had a problem. In the diagram Tiger posted the common would either be connected to all the nearest rails or all the furthest rails regardless of the direction of "forward" travel. If the common is connected to some nearest and some furthest rails there will be horrendous problems.

 

Regards,

 

Andy

[tigerburnie]

I used to make up quite a lot of "prototypes" when working in industry for test rigs and found that working in phases and adding in small changes, you could quickly revert if something goes base upwards. When changing programs on plc's we had a test edit facility, pity we can't have that in hard wire(in fact that's just given me an idea).

[kevinlms]

Code 75 bullhead rail has a cross sectional area of approximately 1 square millimeter. (Code 100 FB will be a bit bigger, but not that much bigger.) The voltage drop at 0.5 amps due to the resistance of the nickel-silver rail alone (no allowance for rail-joiners etc) is 0.136 volts per meter, or 1.36 volts over ten meters. That's quite a lot.

 

16/0.2 copper wire has an effective cross sectional area of 0.5 square millimeters but copper is around 17 times more conductive than nickel-silver so length-for-length the drop over the wire will be about eight times less - say 0.17 volts over ten meters.

Funny, this posting isn't far removed from something I said earlier. By putting a copper bus in parallel with a nickel silver rail, will negate any potential voltage drop inherent, due to NS's higher resistance. Perhaps I used a wrong word, that someone objected to?

[AndyID]

I've noticed there have been quite a few comments about voltage drop due to the contact resistance between rail and rail joiners, but does anyone have actual data on this, or is the evidence mainly anecdotal? I'm not suggesting it's a good idea to rely on joiners to carry the current but I would think initially at least (before the contact surfaces start to oxidize) the contact resistance would be quite low.

[Free At Last]

I never rely on rail joiners for conductivity. I was doing some checking after altering some wiring and before connecting all the droppers back up to the bus. The track has been down a while, fixed with pva and the rail sides are painted ready for ballasting. Not going to investigate any further as connecting it back to the bus will solve the problem. It is on a dead end loco spur.

 

[cliff park]

I had exactly that problem, and was really confused because I always put in droppers and bypass rail joiners. Except I'd missed one, and of course it failed after about 3 years. It was so well ballasted and painted I honestly thought at first that the rail had failed in the middle of a length !!

[smokebox]

I had a voltage loss between two ordinary pieces of curved set track. Temporary jumper wires proved it was the track joiners at fault and two more connections to the bus provided a permanent fix. This was pretty new track which hadn't been ballasted. ( Still not ballasted BTW)

[AndyID]

I've heard that some people actually fill some joiners with solder, not that I would ever do anything like that myself you understand

 

That's alright for short rail lengths but obviously it's not a good idea if the joints have to accommodate much in the way of expansion and contraction.

[AndyID]

It's a bit toasty here at the moment, so rather than do some much needed gardening I did a little experiment to try to find out how much rail joiners really contribute to voltage drop.

 

To get a solid measurement I pumped 1.5 A down a 3 ft length of Code 70 FB nickel-silver rail. The end-to-end drop was 0.192 volts which means its resistance is 0.128 ohms.

 

Then I pumped 1.5 A across two short sections of rail attached to a rail-joiner. The drop across the rail-joiner was 0.0105 volts for a resistance of 0.007 ohms.

 

I thought the drop across the joiner would probably be less than the drop along one yard of rail, but I really didn't expect it to be that much less. This suggests that if you are seeing voltage drop problems the likely culprit is the rail itself rather than the joiners. Of course that all goes out the window if the joiners are not tight on the rail and/or there is oxide between the joiners and the rail, so the best solution is to use lots of droppers connecting the rails to the power feeds.

 

We all have our particular preferences when wiring-up our layouts, but I prefer to use a substantial copper common return. It's large enough that the voltage drop along its length is negligible. That means I've essentially cut the drop problem in half. Also, I have to make fewer connections.

 

BTW, if anyone has seen different results from similar experiments please post them.

[kevinlms]

I've noticed there have been quite a few comments about voltage drop due to the contact resistance between rail and rail joiners, but does anyone have actual data on this, or is the evidence mainly anecdotal? I'm not suggesting it's a good idea to rely on joiners to carry the current but I would think initially at least (before the contact surfaces start to oxidize) the contact resistance would be quite low.

I can see why your asking this. The contact resistance would be virtually zero on new track and clean tight fitting rail joiners. However, with loose joiners, dirty sides of rails etc and the dreaded PVA getting into places where it isn't wanted, the resistance would be quite high - probably 10k plus. Certainly enough to stop trains.

 

Due to unknown variables, there can't be any usable data. Sorry, but not everything can be quantified, with a formula.

 

Advice was given earlier, that if the layout is short term, it may not matter, but if the layout is going to be around for years, then precautions to prevent such a situation occurring sometime in the future are worthwhile. I entirely agree with this.

[AndyID]

I can see why your asking this. The contact resistance would be virtually zero on new track and clean tight fitting rail joiners. However, with loose joiners, dirty sides of rails etc and the dreaded PVA getting into places where it isn't wanted, the resistance would be quite high - probably 10k plus. Certainly enough to stop trains.

 

Due to unknown variables, there can't be any usable data. Sorry, but not everything can be quantified, with a formula.

 

Advice was given earlier, that if the layout is short term, it may not matter, but if the layout is going to be around for years, then precautions to prevent such a situation occurring sometime in the future are worthwhile. I entirely agree with this.

 

You may have missed my latest post.

[kevinlms]

You may have missed my latest post.

I did see it just before I posted, its hard to do very low resistance tests, without the correct equipment (which I assume you have). What readings do you get with your preferred bus in parallel?

 

I also note "Free At Last'"s post. Although I'm not sure that the correct Ohms scale was selected.

[locomad]

My layout is about 40 feet oval, before I rigged up some thick copper wire, I tested a few locos going round, further away from feed controllers the slower they got. Remined me of playing scaletric the cars slowed the further they got from the feeds.

 

I used 2.5mm thick copper wire type used in house wiring striped down, even used the bare earth for points return, easy to solder droppers too. Notice now no voltage drop feeds about intervals about 10 feet.

[AndyID]

I did see it just before I posted, its hard to do very low resistance tests, without the correct equipment (which I assume you have). What readings do you get with your preferred bus in parallel?

 

I also note "Free At Last'"s post. Although I'm not sure that the correct Ohms scale was selected.

 

The equipment was extremely sophisticated and I have to admit it was very hard to conduct the experiment. It must have taken me all of ten minutes.

 

The current source consisted of an old PC power brick which has a regulated 12 volt output and an automobile brake-light bulb, but I do have a decent DVM to measure voltages and current.

[tigerburnie]

I've started soldering joining loops across the rail joiners to create "blocks", have to say I am struggling to do this without melting the parts of the sleepers that hold the rail, it's only localised, but this is making having close sleeper spacings difficult.

[cliff park]

The best way I have found is to solder the droppers on before laying the track. Slide the sleepers as far back as you can, solder wire then restore the sleepers. It is even possible this way to solder the dropper on the flat bottom of the rail, drill a hole in the baseboard in the right place, or as near as practical, then thread the wire through as you lower the track into position. Even without ballast they are then totally invisible. The wire must be near the end of the track so as not to obstruct the sleepers.

[tigerburnie]

Thanks Cliff,it's not droppers, but loops from one rail to another effectively by passing the joiner with regards to power supply, I'll try and get some photos.