IDC failures after 10 years?

[jrbernier]

  I have been a proponent/user of IDC’s for some time.  Our club has been using them for over 10 years.  We have used either 3M Scotch Loc or Posi-Tap connectors to feed power around our 2 level 27’ by 27’ layout.  The bus runs are either #10 or #12 wire, with short #18 feeders.  The 3M IDC’s were installed using the proper tool.

  Over the past years we have had some booster shutdowns where the 8 amp Digitrax boosters have shutdown before the PM42 Power Districts.  After some initial P/D, we thought we isolated the problem to 3 PM42’s.  We swapped them out with 3 PM42’s in staging and the problem appeared to follow the PM42’s.  I ordered 3 new PM42’s and we replaced the suspect ones.

  While we were doing all this we mapped out all of our Power District runs and we found that we had some over 55’, and one that was 100’ long.  This is much further than the suggested 25’-35’ most sources suggest.  Moving one of our 3 boosters will resolve this.  But we also did some checking with RRampMeter and a 1156 auto tail light lamp for a ‘load’ at various places on the bus runs and at track level.  This provides a steady 2.2 amp load.

  The main bus(#10) held up pretty good when using the AWG ‘loss’ specs.  We did lose about .5 volts through the PM42’s(relay contacts?), but we still had at good voltage at the ‘output’ of the PM42’s.  Measurement at track level was something like 12.2 volts – We were losing something like .8 volts through the IDC’s!  The ‘quarter’ test works on each track, but we sometimes are tripping the booster before the PM42 can trip.  I popped off the cover of one of the IDC’s and made sure the ‘blades’ were firmly seated(they were).  Another member then broke open the IDC and soldered the connection from the bus to the feeder – the voltage increased by .8 volts, and the PM42 tripped every time before the booster.

 

  Needless to say, we are now going to ‘map’ everyone of the 62 Power Districts and I suspect we will be soldering all of the connections.  Not the summer project I am looking forward to, but something that needs to be done.

  Has anyone else run into 'age' issues with IDC's?  The layout is in a climate controlled business office, so there are really no enviromental issues like high humidity or wild temperature swings.

Jim Bernier

[Bayfield Transfer Railway]

I was first introduced to IDCs in the early 70s by a couple of Bell Telephone techs.  They swore they were the bestest thing ever, had been tested for years, etc etc.  And Bell was VERY fussy about their hardware.

One thing I DID notice is that those old IDCs for finer wire were filled with silicon jel.  I wonder if that's to keep air off the junction between blade and wire?

[gmpullman]

Jim

As you are aware there's quite a bit of controversy over the success/failure of making wire connections...

http://cs.trains.com/mrr/f/744/t/232258.aspx

The posting there from the late Bob Hartle seems to follow your experiences... 

Even looking at forums of other common users of these IDCs such as boaters, RV-ers, motorcycle enthusiasts wire joining using IDCs is a pretty hot topic there as well.

Looking at the link I referenced above, I am in agreement with the last poster, Mike Lehman, in that I use wire nut connectors (also sometimes called "Scotch-loks" too) and I am very satisfied with their performance after 20 years on my layout. I only buy the 3M or Ideal brands.

My first encounter with "suitcase" type IDCs was in the late 1970s when my dad rented a camper trailer for one of our family vacations. The rental place installed a trailer hitch and wiring harness. My dad watched the installer crimp these "newfangled" wire connectors on to the auto's brake light wires and he questioned their dependability. "Never had a problem" was the installer's reply.

http://multimedia.3m.com/mws/media/313619O/3mtm-scotchloktm-family-brochure.pdf

Today, the Scotchlok^® name seems to only apply to a specific line of small gauge IDCs. I have boxes of twisted wire connectors and even crimp wire connectors that also have the "Scotchlok^®" brand name on them.

About a year later my dad got pulled over for not having working brake lights! Trying to figure out what the problem was, he finally got to the wiring and the connector had cut through the wire and corroded to the point that the wire fell apart.

Sure, that's a poor environment in a car trunk but it was still fairly dry in there but remembering that experience has turned me off of using those "suitcase" IDCs. One thing I am very careful about is not nicking the wire when stripping it and these tap connectors rely on nicking the wire in order to make contact. Now, computer ribbon cable, telco gel-filled snap connectors and some of the other specialty types are OK for their purpose but for me the wirenut twist-on connector is perfect.

I daisy-chain my #12 stranded buss and make a #12 to #14 tap every ten feet or so, then off that #14 I'll wrap maybe 4 or 5 #18s that then go to each individual feeder. I always allow enough slack so if I need to make changes I'll have enough wire to play with. If I have to remove a wirenut I will usually re-strip all the conductors for a clean joint.

I agree 100% with Randy that soldering is the best joint you can make and of course all my feeders are soldered to the rail and anytime I have the ability to"pre-wire" an assembly at the bench it gets soldered but under the layout it is either a twist-on wirenut (of the proper size) OR a quality crimp connector (Sta-Kon^® brand) and terminal strip.

I have played around with the Ideal brand of "In-Sure^®" push-in wire connector. I use lots of these when replacing fluorescent lamp ballasts, quick and easy but you still have to strip the wire. They are supposed to be good with stranded wire but I will tin the stranded first.

Another nice thing about wirenuts is that IF you happen to have to diagnose a short, it is pretty easy to twist off the nut and "ring-out" the individual wires. Usually the nut will come off clean but if not, as I mention above, I'll just re-strip and make a new joint. That's why I allow a few inches or more of extra wire.

I have seen photos in MR and other places where the buss wires are stretched across a module or benchwork with very little slack, then those IDCs are squished on and I wonder how you could ever make a repair without splicing in more wire. Just make an allowance for such situations by leaving a few loops of wire every once in a while.

This is just my experience, others may have different views...

Good Luck, Ed

[Enzoamps]

I am speaking in general, and have no knowledge of your specific installation.  But i have been using IDCs professionally for over 40 years in the electronics industry.  I have found them to be reliable long term in most cases.  At least when properly apaplied.  Where we do see problems is where the connectors are under specified.  In other words if an engineer sees a 5 amp rating on a connector and runs 5 amps through it, it will eventually run warm and get resistive.  Good practice on connections handling current (as opposed to signal lines that carry very little current) is to use multiple pins.  Two or three adjacent pins can be wired in parallel for increased current capacity.  If you are drawing 2 or more amps through a connector pin, perhaps it is heating the pin and over time that weakens the connections, causing the contact resistance to grow.

In fact, in some equipment, due to reliability issues, service bulletins are issued telling us to hard wire the connections of high current wires.  Basically we cut off the connector and solder the wires directly to the circuit.

Any chance this might apply in your case?

[ATLANTIC CENTRAL]

A few thoughts from an electrician:

The simple fact of production tolerance suggests the a suitcase connector MUST be sized to "damage" the wire it is sliding on to, not a good idea in my view.

We never use them in building or machine wiring.

We do use a number of "push on" designs, and they work well, and again they do require stripping the wire.

"wire nuts" - It all depends on the size wires you are joining, but for general building wire sizes the best "wire nut" is by far the Buchanan B-CAP. Only three sizes cover an increadably large range and they hold the best.

http://www.idealind.com/products/wire_termination/twist-on/b-cap.jsp

Personnaly, I consider wire nuts not neat enough for control wiring like model train work.

Terminal strips are the order of the day:

This is one of about 7 major control centers on a layout I wired with an advanced DC cab control system. Some ofthe wiring you see is a 12 gauge throttle bus for four different wireless DC throttles. Rather than use a continious run of wire, the throttle bus "jumps" from location to location with terminal blocks on the relay boards - easier to install, easier to trouble shoot. This system has been in operation for some 7-8 years - zero failures. 

Sheldon 

[Soo Line fan]

Jim,

A trip down memory lane. The first time I saw these were at a parts store owned by a family friend. At the time, they were marketed as "temporary power taps" by 3M.

This was back in the day when cars came with crappy radios, no trailer tow harness and alarms were anything but standard. Lots of opportunities for adding on accessories gave me first hand knowledge on how bad they were.

Nowadays I get to write class materials, service bulletins and repair procedures and as IDCs cannot pass SAE testing, they are never approved for any type of automotive applications. So much for automotive usage.

No doubt many layouts have the same issues as you described. Loading the circuit and then measuring the drop is the correct way to check the robustness of your wiring.

However, as the owners do not know the proper way to measure voltage drop, they are unaware a problem exists. To them, if the so called quarter test works and the trains run, then NTF.

I could not think of a worse application than adding a bunch of IDCs to a long buss run. Each one nicks or cuts a few wire strands.

As additional IDCs are added, more strands get damaged but never the same ones. So by the end of the buss run a 12 ga wire is actually 14 or 16 ga.

Using solder, terminal strips or wire nuts, while taking a little more time, will yield much better long term results. Comparing the actual contact area with any of the three methods I mentioned to IDCs and the also the tension they retain is another factor to consider.

 

 

[dknelson]

Soo Line fan

I could not think of a worse application than adding a bunch of IDCs to a long buss run. Each one nicks or cuts a few wire strands.

I can see how stranded wire is nicked or cut almost by logical necessity -- are the same problems experienced with solid wire, or doesn't anybody use that for layout wiring any longer?

Dave Nelson

[ATLANTIC CENTRAL]

dknelson

 

 

Soo Line fan

I could not think of a worse application than adding a bunch of IDCs to a long buss run. Each one nicks or cuts a few wire strands.

 

 

 

I can see how stranded wire is nicked or cut almost by logical necessity -- are the same problems experienced with solid wire, or doesn't anybody use that for layout wiring any longer?

Dave Nelson

 

Dave, again with my perspective as an experianced industrial electrician, I would never use solid wire in a situation like a model railroad - too hard to work with.

But yes, the same problem is experianced with soild wire, no different than if a novice nicks a wire while striping insulation to install a switch or outlet in your home. Once the cross section of the wire is reduced, the resistance is increased and the current capacity is decreased.

In additon to this aspect, the suitcase connector has never convinced me that it has/makes adequate contact surface area for a low resistance connection.

Traditional twisted splices with wirenuts have contact areas equal to multiple times the cross section of the wire, insuring a low resistance connection. Same is true of terminal strips.

Sheldon

[rrinker]

 This nicking of the wire and small connection area are the main reason I have avoided using IDCs. I DO use solid wire for my feeder drops, but the bus wires are stranded, simply because when you get to heavier wire sizes, it's much easier to pull stranded around the benchwork. The solid feeders work because it's easier to poke the stiff wire through the benchwork holels, there's no chance of a stray strand alon the rails to cause problems, and the bus end of the feeder stays tightly wrapped around the bus line which makes it easier to complete the solder connection. In fact, may last two layouts ran weeks with most of the feeders simply wrapped around the bus wires, until I got around to soldering all joints.

 I use crimp on connectors at terminal strips, but even with the proper type of crimp tool, I thn go back and solder them for additional reinforcement. Maybe that does nothing, but it makes me feel better and only takes an extra minute. I also use the full round connectors, which means I have to fully remove the screw to attach them, unlike the spade type where you can just slip it under a loosened screw. gain, just gives me extra confidence that once the wire is attached, it's going to stay there.

               --Randy

[Soo Line fan]

dknelson

I can see how stranded wire is nicked or cut almost by logical necessity -- are the same problems experienced with solid wire, or doesn't anybody use that for layout wiring any longer?

Dave Nelson

Per solid wire, it becomes notched. Now measure the overall thickness and the cross section will be smaller. Voltage drops are a cumulative or stack effect. As each one occurs, they stack up. So ten 0.100 mv loses stack up to a total drop of 1 volt.

Stranded is used where vibration occurs or where it has to be routed through many bends.  All things equal, I would use stranded. But if you can get solid for free and cost is a factor, it can be used.

[hon30critter]

In my humble opinion, IDCs are like the Sirens of ancient Greek mythology. They sound very tempting but if you go there you may have a real wreck on your hands.

Randy - I'm with you! I tend to overdo things a bit too. There is nothing wrong with being safe, and despite my ever stiffening joints, I can still kick myself in the butt real good when I have taken a short cut that doesn't work out.

Sheldon - Terminal strips are obviously the way to go. Your wiring is a great example of how to do it right.

Dave

[NP2626]

Since when have Scotch-Loks, IDCs, Suitcase connectors been labeled "Wire Nuts"?  Isn't a nut considered something that gets screwed onto the end of something?  Wire nuts have been well defined for decades as a nut that screws onto twisted wires to secure them.  I guess the fact that some have decided a “suitcase connector” will be called a “wire nut” is unimportant.   However, it is interesting how we always seem to bastardize the English Language. 

Without some type of sealant to keep the atmosphere from reaching the joint, I guess the fact that corrosion might cause a problem over time, isn’t all that surprising.  At the time I did wiring on my layout, using Scotch-Loks never crossed my mind.  I figured soldering to be the best method.

[gmpullman]

NP2626

Since when have Scotch-Loks, IDCs, Suitcase connectors been labeled "Wire Nuts"?

I'm not sure where, exactly, anyone refered to IDCs as Wire Nuts but I might make a suggestion as to where all the confusion over the terminology comes from and much of the blame, I believe, falls on the 3M Company for throwing the brand name Scotchlok on just about every electrical connector they ever made.

 

I have a box of Scotchlok "Electrical Spring Connectors" Type G on my bench right now. I also have a few boxes of Scotchlok parallel splice crimp connectors...

And the perennial favorite, Scotchlok "Suitcase" connector which 3M prefers to call a"Run and Tap Connector".

http://solutions.3m.com/wps/portal/3M/en_US/EMDCI/Home/Products/ProductCatalog/~/3M-Scotchlok-Electrical-Run-and-Tap-Connectors?N=5430181+3294245698&rt=rud

Brand names and trade names get genericized so it adds to a great deal of confusion over what exactly a Scotchlok is and what a suitcase is or what a Sta-Kon is or what a Zip-Tie is (I call them Ty-Raps, a Thomas & Betts trade name)

Now, those 3M folks make a BUNCH of tape products including electrical insulating tape (Super "33+" is my favorite) so If someone says wrap that joint in Scotch tape what are you going to get?

From Wikipedia:

The use of the term Scotch in the name was a pejorative meaning "stingy" in the 1920s and 1930s. The brandname Scotch came about around 1925 while Richard Drew was testing his first masking tape to determine how much adhesive he needed to add. The bodyshop painter became frustrated with the sample masking tape and exclaimed, "Take this tape back to those Scotch bosses of yours and tell them to put more adhesive on it!" The name was soon applied to the entire line of 3M tapes.

Now how about that Scotchguard, Scotchbrite, Scotchwrap, Scotchlite, Scotchcal, Scotchprint and, yes, I have some Scotch audio recording tape.

Lots of confusing names for the same things or, worse yet, lots of similar names for very different things. This can cause some real greif as in pharmaceuticals!

Now pardon me while I make another Scotch and soda 

Enjoy! Ed

[ATLANTIC CENTRAL]

And, "Wire Nut" is a registered trade mark of the Ideal Corp., for just one style of twist on connector. Ironicly, the "Wire Nut" design is now the least prefered design among trade professionals - the Buchanan B-CAP is much superior.

Sheldon

[NP2626]

Scotch is a brand name for products produced by Minnesota Minning and Manufacturing, 3M.

[carl425]

Soo Line fan

As additional IDCs are added, more strands get damaged but never the same ones. So by the end of the buss run a 12 ga wire is actually 14 or 16 ga.

How can this be correct?  If one strand breaks, immediately past the brake it is in contact with it's neighbor strands. Why wouldn't it just pick up the current from its neighbors and carry it down the wire?

There was another comment about a nick in a solid wire reducing it's current carrying capacity.  This would only true for the length of the nick.  We know that short runs of small wire are ok because the voltage drop is a function of the wire size and the length of the wire.  A "nick" in a 12 ga wire probably brings it down to an effective 11.999 ga for .01".  If you do the math, I'm sure you'll find that you can calculate the voltage drop, but you couldn't measure it with the equipment available to the average model railroader.

I have always heard that the problem with nicks is that they frequently turn into breaks.

Soo Line fan

Using solder, terminal strips or wire nuts, while taking a little more time, will yield much better long term results. Comparing the actual contact area with any of the three methods I mentioned to IDCs and the also the tension they retain is another factor to consider.

I agree.  I solder everything.

[ATLANTIC CENTRAL]

carl425

 

 

Soo Line fan

As additional IDCs are added, more strands get damaged but never the same ones. So by the end of the buss run a 12 ga wire is actually 14 or 16 ga.

 

 

How can this be correct?  If one strand breaks, immediately past the brake it is in contact with it's neighbor strands. Why wouldn't it just pick up the current from its neighbors and carry it down the wire?

There was another comment about a nick in a solid wire reducing it's current carrying capacity.  This would only true for the length of the nick.  We know that short runs of small wire is ok because the voltage drop is a function of the wire size and the length of the wire.  A "nick" in a 12 ga wire probably brings it down to an effective 11.999 ga for .001".  If you do the math, I'm sure you'll find that you can calculate the voltage drop, but you couldn't measure it with the equipment available to the average model railroader.

 

 

Soo Line fan

Using solder, terminal strips or wire nuts, while taking a little more time, will yield much better long term results. Comparing the actual contact area with any of the three methods I mentioned to IDCs and the also the tension they retain is another factor to consider.

 

 

I agree.  I solder everything.

 

Carl,

Electricity is just like water, once you build a dam/cause a restriction, you loose that flow down stream.

Each successive damaged area further reduces the flow of current and voltage.

It is not the first small nick that causes the problem it is the cumulative effect of many.

Sheldon 

[carl425]

ATLANTIC CENTRAL

Electricity is just like water, once you build a dam/cause a restriction, you loose that flow down stream.

Even if this is true, it would not be cumulative (at least not for water).  If you have a 4" pipe that reduces to 3" for a short distance then opens back up to 4", the flow through the system is reduced to the capacity of a 3" pipe.  Additional reductions from 4" to 3" down the line would not further reduce the flow.

[ATLANTIC CENTRAL]

carl425

 

 

ATLANTIC CENTRAL

Electricity is just like water, once you build a dam/cause a restriction, you loose that flow down stream.

 

 

Even if this is true, it would not be cumulative (at least not for water).  If you have a 4" pipe that reduces to 3" for a short distance then opens back up to 4", the flow through the system is reduced to the capacity of a 3" pipe.  Additional reductions from 4" to 3" down the line would not further reduce the flow.

 

Yes it is, while the pipe does fill back up in terms of volume, some of the pressure is lost permanently at the point of restriction. Each new restriction is a father loss of pressure. With water, some pressure will be "used" filling the pipe back up after the restriction, each new restriction will repeat the cycle. It is true that with water, if you had a constant maximum pressure, it would start to act like a continous 3" pipe. But in a variable situation, if the water was being "throttled" like electrcity, the restrctions would be cumulative until you reach "full speed".   

With electricity that "pressure" is voltage drop.

We are talking very small numbers, and complex formulas of resistance here, but obviously the OP's problem is real and measurable.

Sheldon

[dknelson]

I really do dread the experience of soldering while under the layout looking up -- and yeah suitcase connectors sure looked and look like a tempting alternative.  I am reminded of an article decades ago in RMC (or just possibly the NMRA Bulletin) -- enough decades ago that I cannot seem to find it -- where the guy did all his wiring connections on terminal strips screwed to things that looked a bit like ping pong paddles or pizza slabs with a flat surface and a handle, and the stranded wire was all left loose enough that he could (gently) pull out the paddle and redo or review the connections and the wiring in the best light and working from above not below.  He also thoroughly mapped out and labled everything on that "paddle."  The more I think about it the more I think he made his own terminal strips using parallel lines of brass screws.  Like I said it was an article from a long time ago ....  does it ring any bells with anyone?  I'd like to re-read it if I could find it.

Dave Nelson

[ATLANTIC CENTRAL]

dknelson

I really do dread the experience of soldering while under the layout looking up -- and yeah suitcase connectors sure looked and look like a tempting alternative.  I am reminded of an article decades ago in RMC (or just possibly the NMRA Bulletin) -- enough decades ago that I cannot seem to find it -- where the guy did all his wiring connections on terminal strips screwed to things that looked a bit like ping pong paddles or pizza slabs with a flat surface and a handle, and the stranded wire was all left loose enough that he could (gently) pull out the paddle and redo or review the connections and the wiring in the best light and working from above not below.  He also thoroughly mapped out and labled everything on that "paddle."  The more I think about it the more I think he made his own terminal strips using parallel lines of brass screws.  Like I said it was an article from a long time ago ....  does it ring any bells with anyone?  I'd like to re-read it if I could find it.

Dave Nelson

 

Dave,

This is why the terminal strip and the crimped connector are such a good choice - no soldering up side down.

The relay panel I posted a picture of earlier was built and wired on the work bench then installed under the layout, requiring only the "input/output' conncetions to be made under the layout - sitting comfortably on a low bench on the floor looking at the panel in front of you - not over your head.

Wire is cheap - terminal points should be brought to easy to reach locations.

Sheldon

[Soo Line fan]

Carl

I have no idea where you are getting your information or your background. However it is incorrect.  Any dmm is capable of measuring voltage drop. 

Per my other comments you can search for tons of online training for basic electricty current flow voltage drop etc and validate it. 

[carl425]

ATLANTIC CENTRAL

We are talking very small numbers, and complex formulas of resistance here

Agreed.

ATLANTIC CENTRAL

but obviously the OP's problem is real and measurable.

The OP said that the bus "held up pretty good" and his voltage drop was at the IDC.  So it was the poor connection of the IDC, not the nicks they made in the bus that was his problem.

[carl425]

Soo Line fan

Any dmm is capable of measuring voltage drop.

My DMM measures resistance down to .1 ohm.  Not fine enough to tell the difference between a foot of 12 ga vs a foot of 16 ga.

 

 

[Soo Line fan]

Try measuring voltage drop the correct way. Use mv instead of ohms.

[carl425]

Soo Line fan

Try measuring voltage drop the correct way. Use mv instead of ohms.

 

Then I would need to set up a load - too much trouble.

But back to your post about multiple IDC's breaking multiple strands, what I don't understand is how you can ignore the fact the strands inside the wire touch each other and once you get past the break in one strand, all the strands are carrying current again.

[Soo Line fan]

Once the electron flow has a loss it does not recover from it on the other side of the loss. Thats why its called a loss. Its gone for good. Current can not recover losses unless the source of loss is found and corrected. Current is not self healing.

[carl425]

Soo Line fan

Once the electron flow has a loss it does not recover from it on the other side of the loss. Thats why its called a loss. Its gone for good. Current can not recover losses unless the source of loss is found and corrected. Current is not self healing.

One last try to communicate.  Tell me where I am wrong.

1. We are concerned about voltage drop in the wire.
2. Voltage drop comes from resistance in the wire
3. Different sizes of wire have different levels of resistance per foot
4. If we have a connection made from multiple wires of different sizes, we get the total resistance of the connection by multiplying the per foot resistance for each size wire by the length of each size and adding the products together.
5. We calculate the voltage drop by multiplying the current by the total resistance (E = IR)
6. Therefore a connection of 50' of 12 ga + 1" of 16 ga + 50' of 12 ga results in less drop than 50' of 12 ga and 50' of 16 ga.

My point is that when you break a strand in a 12 ga wire, for the length of the break there is more drop than the same length of unbroken 12 ga, but once you're past the break, it's 12 ga again.  It's not like the whole length now has one less strand.

Now I give up

 

[Soo Line fan]

The problem is you are hung up on the single strand concept. Go back and read where i mentioned it was a cumulative or stack effect. As the bus is nicked in more places as additional feeders are added the total loss is increased.

[carl425]

Soo Line fan

As the bus is nicked in more places as additional feeders are added the total loss is increased.

Yes, it will be cumulative.  At a VERY low rate.

Let's assume a power district bus with a 5 amp booster running at full capacity. We'll also assume that a nick reduces the effective gauge of the wire from 12 to 16 for the length of the nick.  At 5 amps, 16 ga wire drops your voltage .04/foot vs .016 for 12 ga. so you get .024 extra drop per foot at 16 ga or .002/inch.

The "blade" on the IDC looks to be about .03". So if the length of the nick is .03" times .002 volt drop per inch = .00006 volt drop per nick.

So you would need 16,667 IDC's to create a 1-volt drop.