Computers & Mobile Technology
How-To: Splice Wire to NASA Standards

Some commenters on my post about using a washer as a soldering aid noticed my sloppy splicing technique and were kind enough to educate me about the so-called “Western Union splice,” aka the “Lineman’s splice,” which is the preferred method for twisting solid-core wire leads together for inline electrical connections.

Developed during the heyday of the telegraph, the Lineman’s splice is designed for connections that will be under tension. It is commonly claimed that, properly made, a Lineman’s splice is stronger than the wires of which it is composed. In any case, it is a time-proven method, and, coolest of all, one of NASA’s Required Workmanship Standards. To wit, in a NASA-approved Lineman’s splice:

  1. The conductors shall be pre-tinned.
  2. There shall be at least 3 turns around each conductor and the wraps shall be tight with no gaps between adjacent turns.
  3. The wraps shall not overlap and the ends of the wrap shall be trimmed flush prior to soldering to prevent protruding ends.
  4. Conductors shall not overlap the insulation of the other wire.

Though the Lineman’s splice was originally used without solder, today soldering is common. And NASA insists on it:

  1. Solder shall wet all elements of the connection.
  2. The solder shall fillet between connection elements over the complete periphery
    of the connection.

This material comes from page 84 of NASA-STD 8739.4 (PDF), which is a great reference if you’re interested in best practices for interconnecting cables and wires. [Thanks, Alex Barclay!]

118 thoughts on “How-To: Splice Wire to NASA Standards

  1. I read that NASA document years ago: it got me started using clear heat shrink tubing. Just have to remind myself to slide it on the wire before soldering.

    1. I’ve made that mistake myself.

      I started using this method when doing auto electronics, but before I used heat shrink. It made insulating the wires with electrical tape much easier. Before, my tape would always slide off the end of the splice exposing the wires. But with this method, you can wrap each wire, and then bundle the whole group together and wrap the whole bundle neatly.

      I also found that if you are good at stripping the insulation off of the middle of a wire, this is a good method for making taps as well. Also you can splice more than 2 wires together by coiling 2 wires at the same time (I doubt that is up to NASA standards though).

  2. What a badly-drawn picture. The wrappings on each side somehow reverse direction. They can’t end that way.

    1. Really? You get all this good information and despite the visual typo, you still understood what was going on and despite all of the authors work, you had to nit pick? Just…wow.

  3. Very M.C. Esher-esque.

    For those of us who are visual learners, these images are confusing. It would be difficult for those wire ends to terminate the way they do in the images. It looks like they reverse direction halfway through the twists.

    1. Right. It’s like an optical illusion. The end of the right hand side of the twisted wire goes opposite direction. Interesting discovery!

  4. A video of someone with skill doing this would help to understand how you do the wrappings efficiently and cleanly. How does this work with stranded wire or is that a different standard?

    1. There’s a whole category of when to use solid or stranded wire. I think you could use this splice if you had to with stranded, particularly if you make sure to twist the strands in the right direction (so they don’t loosen under tension).

      That said, I wouldn’t recommend using stranded in this application, for several reasons, but the one that leaps to the top is that a soldered connection subject to any sort of twisting or vibration will fatigue the strands at the end of the solder, leading to failure. Solid core is more resistant to that, if properly sized with reason.

      You can tin a stranded cable into a solid unit to avoid that at the splice, but you still have a solder interface farther away. So that only works if you’ve got either strain relief or no real strain in the first place (neither seems likely if you’re needing to use this splice). It’s less the tension, but the inevitable vibration that will slowly kill it.

      But then, I’m thinking in the mindset of the “it must not fail” school of engineering here. So YMMV.

    2. It may not be a standard, but what I’ve always done with stranded wire was to spread the strands on both ends a bit, slip the two ‘tufts’ together (end to end) until the ends of the strands of one are almost touching the insulation of the other (trying to make sure that all of the strands are interspersed evenly), twist the whole thing until it is solid (pre-twist the wires in the opposite direction before joining to make the soldered connection lay flat), then solder into a solid unit. I’ve found that, in a pinch, if you strip off a lot of insulation, you can often get away with not using solder. Also, if you slice along the insulation instead of cutting it off, then wrap it back around the soldered joint, it adds mechanical support.

    3. The way to do it with stranded would be to separate all the strands so you can tin them individually, then twist them back together, do the splice, and when you finish with the final solder it will be the same as a solid wire splice, with no copper exposed anywhere in the connection to oxidize.

  5. At one of my old jobs I would have to do this type of splice on occasion. Handy tip: keep a thimble in your toolbox — it helps you avoid accidentally stabbing your fingers with the end of the wire.

  6. Nice standards document! I’m not going to apply all of it to any of my projects(overkill, even for me), but some elements of it are very informative of what to be thinking about when doing a wiring harness (in my case likely an aircraft wiring harness which is a bit more critical than my other projects).

  7. This is based on a splice invented by Western Union to splice telegraph wires. Recall that in the 1800s there was the bare minimum of tools available. The splice was designed to not only be secure electrically but also mechanically. In addition the same cathodic affects that prevent corrosion in Wire-wrap assembly also protect this splice from corrosion.

  8. I still hate doing it with smaller wires…anything smaller than 20awg is just too much for my toe-fingers

  9. One of the first things I learned (or was at least taught) in soldering, especially in wire work. Was to always make a good solid mechanical connection first. The solder is there only for electrical connection.

    I have also found that this method puts less strain on the wire than when they are just twisted together, and is less likely to break.

    I have to admit, I still use less dependable connections when testing that are easier to disassemble.

  10. The example is good and it demonstrates a very strong splice. Thanks for making a note of it.

    There is a problem in your diagram at the far right end of the wire on the final twist. The direction of the wrap defies logic and physics unless you are like me, an admirer of art of M.C. Escher. Nevertheless, the concept is clear. It is fine as it is for us g-jobbers.

    Does that bug in the drawing really matter that much? Consider that a NASA or military SOP (standard operating procedure) may include or reference a picture or diagram similar to the one above for use by factory assembly staff and test/repair technicians. If the diagram above were included, that M.C. Escher twist on the final turn at the right end would become an expensive issue during an audit should verification or certification of a splice be required. QA inspectors in the DLA (Defense Logistics Agency) are required to compare an observation made during an inspection with a requirement of a standard. And as crazy as it sounds, that splice would be sited as defective and formally documented as a condition of noncompliance. Even if that inspector is intelligent (some aren’t) and realizes that the drawing has a bug, there is still a discrepancy that they will document. It is sad. I’ve been there. I’ve seen it happen. (Then there are all the stop-work forms, change authorizations, engineering analyses, change notices, re-certifications and so on once the problem is addressed . . .)

    (Man am I glad I don’t work in defense electronics anymore. :- )

    Anyway, your article will help others to make better repairs, even if they aren’t Escher fans. Thanks for publishing it.

  11. Is it just me, or are the ends in the 2nd picture the wrong way around?

    From the first picture and looking from center out on the 2nd picture, you see the wire coiling over the front from bottom to top. But the end is in the BACK from bottom to top.

  12. It’s easier to make this splice if you cross the wires at a 90 degree angle, with the insulated parts below and the stripped ends pointing up. Leave 2/3 of the stripped part above where it crosses. As the stripped part wraps around it gets much shorter so you need more length to wrap around the straight part. I had to do two of these today to repair a damaged power cord on a commercial wetvac. Solder and heat-shrink finished the job.

  13. I recall someone (a post doc I worked with?) saying that the only way to get any work done on the scientific satellite their group was making was to come in at night while the NASA auditors weren’t around.

    “Is that the correct Philips-head screwdriver for that screw?”

    Might just have been a grad student war story, though.

  14. I learned to solder from a NASA instructor but not for certification. Learning to solder wires together tool like 6 weeks. Yes, this is one of the joints we learned

    The interesting part is why. Early on NASA was very unsuccessful in launching rockets. The failures were traced to “drum roll” …bad solder joints.

  15. Hmmm. Something about that middle illustration. That’s some fancy wire twistin’.

    1. Yes the right side starts out as “Overhand” but the end appears a bit “Underhanded” :-)

  16. That has also been military standards, at least in the avionics field, for a very long time. Have to say that getting one to look as pretty as these illustrations would be a neat trick. The wire never wants to cooperate.

    1. Part of A-School we were taught to splice and solder. I remember that wrap clearly to this day. It was the most challenging part of that class (that and avoiding getting hot blobs of solder in your eye – that happened to me).

      Actually, I think it might have been AFTA where we learned that – sort of like a secondary more advanced A school (Advanced First Term Avionics).

  17. I’ve made many Western Union splices, none of which looked like the illustrations. Of course I only work in three dimensions not four.

    1. Ahahahaha – I’ve so DONE that! Looking at your beautiful solder job – nice and clean with not cold spots – lo and behold no shrink tubing. DOH!

  18. i am familiar with this splice, known to me as a western union splice . I used this splice many times over the years. But what I found interesting is that while I was in the Navy in the 60’s I was flying on P2V-7s patrol planes (anti-sub). We relied heavily on HF radios and Morse code. We had several antennas and one of those was a long wire trailing antenna, which was a very long piece of solid copper wire with a weight on the end, wound on a motor driven spool. When we used this antenna we would unlock the spool and the weight and gravity would pull the wire out behind the airplane. The weights and wires would sometimes break and require splicing. The method of splicing was to use a western union style of splice to splice two wires together and they added another piece of wire at the splice point and it was also spliced western union style into the splice. This was just a mechanical splice, not soldered. They seemed to be good splices because one time we were landing and I forgot to rewind the antenna so we landed with the long wire antenna weight bouncing down the runway and the weight never came off.

  19. Ok the the wire starts looping one way and ends the other way round. Seriously bugging me.

    1. Three long twists of the wires around each other, then take each end and do the three to four tight wraps of wire. Snip of the excess.

  20. Learned this in Electronics 1-2 in high-school. We were required to do examples of various splices and the proper amount of twists per splice. This was the hardest, but I managed to get mine looking exactly like this and can still do this today.

    Thank you Mr. Trujillo.

  21. CAD Drawings do not always imitate the real world as evidenced in these pictures. The idea is otherwise portrayed flawlessly.

    In movies this is called suspension of disbelief !

  22. NASA may have adopted it but they DID NOT invent it. I was a USAF Manufacturing Engineer and STS Operations Officer. This splice was invented for telegraph wires and the formal name is the “Western Union splice”. It has amazing tensile strength, and is nearly impossible to pull apart – the wire itself will usually fail first. The splice was then adopted by the fencing people when innumerable miles of wire were strung to contain livestock. An amazing patented little tool by the JL Behmer Corp. (http://www.jlbehmer.com/) makes the splice fast and easy and creates a perfect and tight splice (after a little practice. It is a must have if you mainteain wire fences (my wife uses to maintain 7 1/2 miles of fence). As part of the FAA tests to become an aircraft mechanic – you must demonstrate this splice and then test it to destruction – the wire should fail before the splice. sent by Terry Galbreath

  23. I’ve posted a scan of the Digital Equipment Corporation’s Workmanship Standards Manual to scribd. It’s got all kinds of chassis and PCB assembly tips in it, probably some cribbed from NASA.

  24. I was Certified for mobile electric wiring back in the mid 90’s and Soldering a joint was a no no. The correct way was to us a crimp connection believe it or not?

    1. That’s an ongoing discussion/debate about reliability of joints in a high vibration environment.

      I’ve seen it argued both ways, and the industry has wandered around both directions as well.
      I believe Ford currently requires splices to be twisted together perpendicular to the wire run, and then folded down against the wiring before being covered with heat shrink or tape, but I may be mis-remembering the procedure.

  25. Although this looks like a good rendition of the splice it has a significant visual flaw in the representation that could be confusing. According to the turn of the twist starting from the center of the splice the wire should be turning up from the bottom of the center wire on the right end and coming over the top on the left end. The visual makes it look as if the wire ends changed directions in the middle of the twist. It’s a small detail, but, it creates a little confusion.

  26. this method was used first bu the people that installed the telegraph system. I myself am one of the few electrical masters in Michigan who was trained under the different methods used when the knob and tube wiring were installed. many buildings in Detroit still have wiring operating from 1910 we had two basic methods used for concealed ans another used for surface. the lesson learned from this method of connection is when it works dont change it When I joined the navy and learned electronics this method was taught along with how to solder. I have seen connectons over one hundred twenty years with a solid connection without showing any resistance because I tested it

    Grampa~~~~~~Mullan Elect.

  27. Father in-law worked at NASA (radio Engineer) taught me to use simple Lap Splice. Maybe something with RF ????? I have seen the lineman’s splice but never use it.

Comments are closed.

Tagged

I am descended from 5,000 generations of tool-using primates. Also, I went to college and stuff. I am a long-time contributor to MAKE magazine and makezine.com. My work has also appeared in ReadyMade, c't – Magazin für Computertechnik, and The Wall Street Journal.

View more articles by Sean Michael Ragan