14

One of the things that's always been drilled into me is "loose wires cause fires," and that's something I always remain conscious of when changing out devices. But is there such a thing as "too tight" when it comes to tightening the screws on standard 15/20A receptacles and the like? What problems can it cause if one tightens things a little too far?

(There have been a handful of times in the past where I cammed out the screw head, stripped the threads, or bent some part of the structure by going a little too far. In all those cases, I discarded the device and tried again with a fresh one. But is there a torque danger zone before any of these things happen?)

smitelli
  • 601
  • 1
  • 6
  • 13

3 Answers3

28

Recent (last 10 or 20 years, maybe a bit more) code says to use torque wrenches/screwdrivers for all connections now.

You found some problems with going too tight. In addition, you also squeeze the wire too much. Most electrical wire is either copper or aluminum. Most screws and screw holes are made of harder materials. If soft wire is tighten/squeezed too much, it can be reduced in size, causing an increase of resistance (heating). Weakened/pinched wire may cause the wire to break off, including after you've closed the box.

All devices should have their torque values listed now, so no reason to be too loose or too tight.

End Anti-Semitic Hate
  • 2,253
  • 8
  • 36
  • 61
crip659
  • 24,171
  • 1
  • 30
  • 61
  • 8
    I fail to see what is wrong with this answer, other than being an inconvenient truth. – Harper - Reinstate Monica Aug 14 '22 at 02:56
  • 2
    Strange downvotes. Maybe the individuals involved can explain themselves. – Aloysius Defenestrate Aug 14 '22 at 03:33
  • 3
    If I make a mistake, I do like to know how or why, so I can learn. Do know that the exact code should have been mention, but I don't know the code that well. – crip659 Aug 14 '22 at 17:31
  • I'm skeptical of all torque specs as actually practical and useful. Sure, I understand some math in Autocad tells you X number, but in practical experience, torque specs are almost always worthless. Just make it tight and the item still works as expected. Soooo, since I think this opinion is common, "torque specs" is not a satisfying answer. I haven't downvoted, but I can see why. What evidence is there, other than obvious "you broke it" situations, that you can over tighten and cause issues? –  Aug 15 '22 at 01:49
  • 1
    I did most of my electrical work before torque specs, GFCI/AFCI and quite a few other things were thought of. I do know that this stuff is all tested before they tell us we must do it. Will admit if I had to change one switch/outlet/light fixture, I might not run to the store half an hour away to get a torque screwdriver just for that one device, but would if I had to change a few. Quite often, too tight will leave a screwhead hanging by a thread/tiny amount, and in time fall off. – crip659 Aug 15 '22 at 02:02
  • 1
    @Harper-ReinstateMonica I didn't downvote but this question shares no insight about the issues associated with overtightened connections. – MonkeyZeus Aug 15 '22 at 12:28
  • @MonkeyZeus "You found what some problems with going too tight,you also squeeze the wire too much." seems to address that point. – FreeMan Aug 15 '22 at 14:37
  • 2
    @FreeMan I argue that it addresses nothing. It merely re-stated OP's action rather than explaining a hazardous result of the action. "Too much" has no qualifier. If there was an explanation of what bad result does "too much" achieve then I'd be inclined to upvote. – MonkeyZeus Aug 15 '22 at 14:52
  • 3
    @MonkeyZeus Think you might be right, the problems not fully explained. Will try an edit to try to make it better. – crip659 Aug 15 '22 at 14:56
  • 1
    I like the edit. Worth noting that the cause of wire weakening is microfractures in the metal. – MonkeyZeus Aug 15 '22 at 15:10
  • I upvoted but the answer is a little hard to read, maybe that's why the downvotes. – jay613 Aug 15 '22 at 15:30
  • 1
    @jay613 Per your comment, I just made some edits to improve readability. – End Anti-Semitic Hate Aug 16 '22 at 00:14
22

This was investigated more formally by researchers, and it was determined that your folk "knowledge" is only half the story. Too tight is also a problem.

Also, they discovered via testing that pro electricians could not "guess" screw torques any more accurately than their spouses or business managers.

The result of this was NEC 110.14, which requires torque tools be used to set screw torques to the specification - anywhere a torque is specified.

This is actually nothing new: NEC 110.3(B) already required installers to follow instructions and labeling, including torque specifications. It simply deletes the (proven wrong) excuse of "I know what 18 inch-pounds feels like!"

UL approves the instructions and labeling as part of listing equipment. So this allows UL to require that manufacturers either specify a torque, or show in testing why it doesn't matter.

times in the past where I cammed out the screw head

If you look closely at receptacles you'll see the combo standard/Philips head is "funny". It's a little extra hollow in the middle. This is actually a Robertson socket. You won't have cam-out problems if you use Robertson.

Harper - Reinstate Monica
  • 300,628
  • 26
  • 286
  • 734
  • 2
    Yeah -- the kicker here is the OP is asking about wire binding screws, whereas the tests you mention were on mechanical set-screw type lugs (breaker/panel lugs IOW)...so it might be the case that overtightening is less harmful here, but I can't tell for sure – ThreePhaseEel Aug 14 '22 at 05:39
  • 3
    Robertson is Canadian for square not Torx. – crip659 Aug 14 '22 at 10:31
  • 1
    Yeah, to be clear, Robertson and Torx are very different-looking screw drives (Robertson is a square, Torx looks like a six-petaled flower). But maybe Robertson is kind of the Canadian equivalent of Torx. – Tanner Swett Aug 14 '22 at 10:45
  • 1
    @crip659 thanks for that info, edited. – Harper - Reinstate Monica Aug 14 '22 at 17:18
  • 1
    The screw heads for electrical receptacles are ECX drives. These are essentially a combination of flat head, Philips, and square drive. They can be tightened with any of these drives individually, but will work best with a specialty ECX bit. – Glen Yates Aug 15 '22 at 14:28
  • 1
    Now the trick is to get contractors, builders, and unlicensed "handymen" to actually follow the code, which was created for good reasons. – End Anti-Semitic Hate Aug 16 '22 at 00:16
  • A pro will at least have some experience on what will work, what will not in practice. The trap to avoid is suggesting to anyone else - especially apprentices and amateurs - that they should wing it in order to be more "pro". – rackandboneman Aug 17 '22 at 00:24
  • Torx drives also have that problem of being "too good" - you can apply insane forces on a screw without something feeling off :) – rackandboneman Aug 17 '22 at 00:26
4

Conceivably, if you grossly over-torqued a connection, you could deform the wire and squeeze much of the conductor metal out of the connection. That could lead to a bad connection (excessive resistance) and/or mechanical failure (if there's less metal hanging on, it could break off more easily than it should).

I think it would probably be hard to achieve this by hand on typical receptacle/switch screws, but if a device specifies a torque range, you should just use a torque screwdriver and follow the specified range.

nobody
  • 6,670
  • 3
  • 31
  • 41
  • It's not that hard to cut a #14 wire by accident by over torquing. It's happened to me a couple of times. Not enough to detect a "why" pattern, but possibly a) wires had hardened and/or b) device design, particularly the ones with a screw-down wire clamp accessed through a small hole in the back ... I think. It's definitely more to do with the design than the actual torque. – jay613 Aug 16 '22 at 00:38