Question on wire splicing

In reply to Kenny_McCormic:

In aircraft you would tie as close as possible to the splice because it's a higher mass point than the rest of the wire. If there's vibration or other cause of movement, that splice would be where it wants to focus. As I understand it anyway; I'm an EE and don't do wire routings or hold down placement drawings (save for the times my career took brief odd turns). But this is how I believe this is handled on the aircraft I'm involved with.

I'm a big believer in soldered and shrink wrapped connections for my stuff and use something similar to the Western Union/NASA splice shown earlier. Many people are mystified by soldering, it's actually very simple. The big thing to remember is that the work piece (wire conductor) is hot enough for the solder to flow into the wire.

The real reason OE harnesses are crimped and not soldered is speed of assembly more than anything else; if you stop and look critically at how a car is assembled it's all to make it faster to assemble, everything else is secondary. If any connection is done properly and not positioned in an area prone to flexing it should last the life of the vehicle; I have split open 25 year old rotary engine harnesses that are still shiny and new looking inside. I also have seen the effect of an improperly positioned splice, Toyota 22RE pickups and 4Runners from the mid 80's to the mid 90's had a 4 way splice in the engine harness that was at the lowest point of the engine harness 'drip loop'; water would get inside the harness and corrode the splice causing random/intermittent misfires. The fix was to split open the cover, cut the wires far enough back to get past the corrosion, solder and shrink wrap all the repair wires and then wrap the harness back up. You had to make sure the splice was no longer at the bottom of the drip loop.

Sometimes the only realistic repair is butt splices. The cheapie hard shelled ones sold at Wal Mart etc are complete garbage as are those Scotchloks. These will get you through in a pinch but are not my first choice. Butt splice:

Scotchlok:

These are truly junk. No weather seal, the connector 'guillotines' the conductor and that lowers the wire's amp capacity, that can create a 'hot spot' then possibly a fire.

Typically butt splices with shrink wrap are stupid overpriced, but Harbor Freight has them now. I got some, my only gripe is the shrink wrap is somewhat thicker than the Wurth etc high priced ones. http://www.harborfreight.com/10-pack-18-22-gauge-watertight-heat-shrink-butt-connectors-66595.html

I have also had good luck with uninsulated butt splices covered with shrink wrap.

Curmudgeon wrote: I have also had good luck with uninsulated butt splices covered with shrink wrap.

This is what I usually use. Quality crimps and a quality crimp tool. No failures to date.

TheV8Kid wrote: I recently gotten involved with the Formula SAE team at my school. They have lost several times in the past because of faulty wiring. They have asked me to do the wire splices, since I seem to have the most experience with it. My question is, how do you guys do your wiring? My father always taught me to use butt connectors, solder, and then a shrink sleeve similar to this picture: However, I have heard some people claim that soldering is not the way to go, and my team members had loosely agreed ,"No soldering this year." I think I can convince them to allow soldering, and I feel that soldering is the way to go, but I am open to other viewpoints. Also, I found that the NASA standard involves soldering the connection. http://makezine.com/2012/02/28/how-to-splice-wire-to-nasa-standards/ Anyways, I'm ranting, and I may have answered my own question. Tell me what you guys think. I'd love to have more opinions as this seems to have been the teams breaking point for the past several years.

"Faulty wiring" sounds like a catch-all phrase that might include "poor circuit design" and a host of other poorly executed electrical systems.

Did they really have several cases where the wires actually failed at a splice point? Yeah there are good, better and best practices to making a soldered joint, but it's not that hard to do, and even "good" should do the job for a reasonable time.

Since you are in an engineering program, I would recommend that you look up the Military Spec on how to do the job. Those guys are pretty fussy about their electronics and its something that I guarantee you will run into in your future.

pres589 wrote: How many sensors in an automotive application need to common their grounds to a single spot with other sensors? I would think that if the sensor ground isn't required by the ECU or similar logic box, the wise move would be to collocate the ground near as possible to the sensor.

Negatory! Don't just ground it wherever. You are trusting that there won't ever be a ground potential difference, statically or dynamically. That is a bad thing to trust when the electronics are sensitive to the hundredth of a volt!

Ground everything to the same thing, preferably on the same bolt. I call it "Ground Central Station" because I am somewhat of a weenie. The OEMs do it this way even though this means they have to add many feet more wire. They don't have problems. One ground point and done, not five dozen ground points here there and everywhere.

In reply to Knurled:

Lemme get back to you on this but I don't think it's how we design and integrate much at work. I don't work a lot with sensors, though, so this isn't off the top of my head kind of stuff. I do know that avionics LRU's aren't grounded that way but they're designed to not be so sensitive to this stuff.

If everything is well bonded to the same ground plane, that is, the chassis, then you shouldn't have ground potential difference. Right?

damn, I usually just splice the two ends with a lazy western-union style, solder it (often using a Bic lighter since I can never find my soldering gun and always have a lighter since I smoke), and heatshrink (if I remember to put it on) or else just tape it up. Never had any issues on the rallycross car, and that thing takes a beating in terms of shock and vibration.

And yes, I know that's not "doing it right," lol. But since I always have solder and a lighter in the car (part of my backup kit), I don't really sweat it. On our chumpcar, we put a bit more attention into the wiring since half my team are eletrical engineers :)

The only electrical problems I've had on any car in the last decade were due to bad/corroded grounds (ironically, most of them on the Subaru and the Toyota and none on the 30-year-old e30)

In reply to pres589:

In theory yes. In practice, it's not that simple, and best practice is to not assume perfect conditions.

Every automotive PCM installation that I've bothered to look into has used only one, maybe two ground points for everything related to a given system, and that point is generally on the engine itself. Generally speaking, this is the case. More and more, I'm seeing that they are going postal and sensor power and ground are all going to the PCM itself. So when I wire a standalone and certain items get their own grounds, I make sure that the ground points are all shared. It's only a little extra bit of wire and it eliminates any potential (sorry) hassle. It's actually a little less work this way, since I will gang several light-gauge wires into one heavier-gauge ring terminal.

Having grown up around aircraft maintenance and being an aircraft mech myself, I was taught to use aircraft splices. Quality splices are almost as good as never spliced one piece wire if done right. Of course no splices is best but realistically there is a need. Those first splices shown I've only seen used to put 3 wires together with heat shrink over it. Not done often. Solder is better than cheap splices. Was taught a lazy version of that western union solder splice. Terminal boards are good too but should be protected and using crimped eyelets are better for this.

And if you are splicing multiple wires in a bundle, stagger the splices so they are side by side.

NOHOME wrote: Did they really have several cases where the wires actually failed at a splice point? Yeah there are good, better and best practices to making a soldered joint, but it's not that hard to do, and even "good" should do the job for a reasonable time.

I was told they had multiple failures at splice points. They didn't really seem to have a good method for doing there splices. They said that in the past they were just twisting the wires together and soldering them. I'm not trying to be mean or anything, but I think they just lacked experience and did not have time to research and make sure it was done right.

What do you guys think of these?

Posi-lock

I've used these before, and they seemed to hold up very well.

In reply to Knurled:

What's probably coloring my ideas on this is how aircraft systems are designed to be somewhat stand-alone. Kind of like how you don't normally have multiple engines in a car, aircraft systems assume that they're the only one in the plane, and they basically ground where they need to and not worry about common ground points.

There's also a difference in how different companies do things. One airframer may do this different from another. I have a decent amount of experience at one of them that doesn't try to run a bunch of grounds to a common point. Other airframers that I've spent a bit of time with do try to do that kind of thing. For my position it really comes down to what the manufacturing engineers think is best practice. Also, sometimes there is a real benefit to having fewer feet of wire in the vehicle, so if you think you can get away with a collocated ground then it might be worth considering.

So, what we're seeing here is that anything can fail if badly executed

I understand that speed of assembly is part of the decision tree for manufacturers, but given that most production automotive harnesses will last a quarter of a decade without problems (other than particular design flaws such as Curmudgeon's Toyota splice) it's safe to say there's nothing wrong with a properly done crimp.

What would be interesting would be to see how F1 and Le Mans teams do it.

In reply to Keith Tanner:

I found This. From what I gathered from that the preferred method is crimping.

Nick_Comstock wrote: In reply to Keith Tanner: I found This. From what I gathered from that the preferred method is crimping.

I now feel totally inadequate!

This would make a great article for GRM.

Nick_Comstock wrote: In reply to Keith Tanner: I found This. From what I gathered from that the preferred method is crimping.

Oh wow, that's fantastic. Thank you.

This document was linked on that page. Check out pages 15-23 for how to properly crimp.

Here is the Army, Navy and Air Force manual for repairing wiring: TM 1-1500-323-24-1.

There is also a bunch of other cool stuff on that site.

Every time I read one of those "Mil Spec" manuals, I picture some poor sap under a hail of bullets going through the whole manual so as to meet military code on his field repair.

I be getting the duct tape!

Datsun1500 wrote: Am I the only one that noticed the wire in the nasa example changes direction?

CANNOT UNSEE

In reply to NOHOME:

The books are for when you are in your office with a cup of coffee, the 100 mile an hour tape is for the bullet storm.

Knurled wrote:

Datsun1500 wrote: Am I the only one that noticed the wire in the nasa example changes direction?

CANNOT UNSEE

Alien technology.

Rusted_Busted_Spit wrote: In reply to NOHOME: The books are for when you are in your office with a cup of coffee, the 100 mile an hour tape is for the bullet storm.

The books are to hopefully prevent the field fix.

Another reason to use chassis grounds on an aircraft: wires are heavy. Of course, a somewhat heavier aircraft in which the engine still functions is generally regarded as superior to one that is essentially a nose heavy glider.

I personally do a stay-kon (uninsulated butt crimp connector) and then solder. I do that for 3 reasons: 1) its easier than the western union to put them together, 2) its redundant, meaning that they are held together by both physical crimping and solder, and 3) it is very difficult for me to get the copper conductors hot enough to pull solder without damaging 1/2" of insulator, but I can set the crimp on the soldering iron and get solder to pull through from the other side with no damage to the insulation. It just adds more mass and heatsink in the middle which keeps the heat there instead of traveling out the copper as much.

Good shrink tube with a heat gun and done. It might not be NASA spec, but way more than adequate.