wire: stranded vs. solid core

[mook]

strickly talking in the POWER section of an amp where higher voltages are of concern:

are there any advantages in using stranded vs. solid core? I'd prefer to use solid, because I feel stranded breaks easier.

I know you need a "thicker" gauge solid core vs. a thinner stranded due to voltage handing concerns, but other than that, why use stranded?


Any opinions?


Mook

[defaced]

Voltage handling is dictated by the isolation over the wire, not the wire gauge or wire type.

Current carrying capacity is where wire gauge comes into play, but even in a power section, you're max looking at like 500 ma, which is not alot of current.

Use whatever kind of wire you like, but make sure the gauge and isolation voltage rating are sufficient for what you're doing.

I use 22 gauge stranded teflon wire for my amp builds (minus the heaters, I like 14g solid for that). I like it because it's flexible and easy to work with. Other people like solid wire because it stays where you put it. In my experience though, I've had more failures from solid wire either breaking loose from the solder or straight up breaking than stranded. Plus solid wire is a pain to work with, IMO.

American Wire Gauge table and AWG Electrical Current Load Limits with skin depth frequencies

[Enzo]

I think you are confusing some issues.

Wire thickness has nothing to do with voltage. Wire thickness is a concern with current. And tube amps don't use a lot of current. 500v only needs 100ma to make 50 watts.

When you see "600v" on wire, it refers to the insulation, not the copper wire inside. That means the wire can lay right on the grounded chassis, with 600v running through the copper, and it won't arc to the chassis.

Stranded or solid doesn't matter much. 18ga is 18ga as far as we are concerned.

I personally find solid more likely to break. I find stranded more flexible, and it is bending that breaks wire. Your experience may be different.

In my view, solid is easier to form into shapes that will stay. Make a nice crisp right angle for example. But stranded I find easier to keep smooth. Solid can get crinkled if you bend it some, and then harder to make look pristine later.

I'd say the choice is more about what you prefer to work with than anything electrical about it.

[Chuck H]

Let's not forget "skin effect" ??? I don't know much about it really. It has something to do with the surface area of lead wire being the most active area. So it's been argued that stranded wire is better because of it's greater surface area. Probably has more to do with signal leads than power leads, but I'd be interested in more clarification.

Chuck

[defaced]

Yea, I don't know all the ins and outs of it either, but according to the link in my last post, you've got to be using 18 gauge or larger wire to be in the audible range and in the 15-16 gauge wire to be in the top end of the guitar range.

As for sonic implications, who knows. We're not dealing with high current devices, and according to the almighty Wiki, all skin effect does is raise the effective resistance of the wire above the skin frequency. The more I think about it, as long as there's sufficient cross section of wire to carry the current after you take into account skin effect, I don't think it's really going to effect much. Add that to the fact that we're usually working with input impedance in the 100k of ohms anyway (at least in the preamp), and I can't see (with my limited knowledge) this being a huge importance.

[Steve Conner]

Skin effect is a non-issue at guitar amp frequencies and impedances.

Even if it was an issue, stranded wire doesn't help, because the strands are in electrical contact, so the whole thing behaves like a solid conductor. To combat skin effect you need Litz wire, which is made from strands of varnished magnet wire. It's commonly used for winding transformers in switchmode power supplies, and available off the shelf in various sizes.

Another popular trick in high power RF circuits is to use copper pipe for conductors. The current all flows in the outside layer, so they just leave out the copper in the middle as it's not doing anything.

For really high power and frequency, you silver-plate the pipe, which makes a big difference because most of the current actually flows in the thin layer of plating.

[daz]

Tonal considerations aside, his question actually revolved around the use of solid because he felt stranded breaks easier. This caught my eye because i don't understand why the OP would think this. stranded will out do solid 1000 times better when it comes to reliability as breaking goes ! Solid can break with years of use just due to vibration and such, especially if it's even slightly nicked. Stranded....solder joints aside, i cannot even imagine a piece of stranded in the sizes we use in an amp breaking EVER unless it's stressed over and over again many times such as moving it while working on it. And then i can't imagine it breaking unless you move in on the order of 100's if not 1000's of times. I use 18 gauge stranded and this suff is so sturdy i think it would literally never break no matter how long the amp lasts. So to the OP i say go ahead and use stranded. If thats the only reason you prefer to use solid don't worry about it. Stranded if far more reliable.

[R.G.]

To recap:
- solid/stranded has little or no effect on sound at audio frequencies
- Litz wire (actually a shortening of something like "Litzendraught" or some such) is good if you have really high frequencies and high current. But it's a PITA to work with - I have.
- at audio, 18Ga is 18Ga, solid or stranded.
- the wire insulation, at audio frequencies has no effect, given only that it really does insulate and not break down.
- prettiness is best done with solid if you like the rigid-right-angle style or insist that wires stay exactly where they were put; otherwise, stranded is more easily bundled if you like the bundled-and-tied look. Not many people know how to hand a spool of waxed linen cable tying cord these days, though.

Daz hit a big point that most people have no clue about - nicking the wire while stripping it. There is essentially no way to strip wire mechanically and guarantee not to nick it. To do that, you must use thermal or chemical stripping. Chemical stripping is best used on magnet wire - or our friend Litz wire, which is often made from strands of magnet wire. The chemicals are very, very nasty. Thermal stripping is so much better that it's like sex - you never go back to what you used to do before.

Thermal stripping used to be expensive and moderately dangerous, in that you could burn yourself just like with a soldering iron. But there exists at least one thermal stripper that's quick, easy, non-dangerous and just so slick that I keep a spare. It's the PTS-10thermal stripper. Mine cost me about $35. They heat up in seconds, take 3W at idle, and are so much better and faster than diagonal strippers, knives, claw strippers, teeth, whatever that, like I said, you'll never go back. There's even an adjustable stop so that every wire is stripped the same length. Stick in a wire, twirl it with thumb and forefinger, pull it out. It's a one-handed operation. Each strip is about 2 seconds.

...oops, I'm waxing poetic... sorry.

The US military used to specify thermal stripping on all wires specifically to prevent nicking the wire.

In severe-vibration environments, stranded is much better because it's more flexible, and does not work-harden with vibration as much. Except that you can and do convert stranded to solid at each solder joint as the solder wicks into the wire. Completely unrestrained stranded will break first at a nick, then at the place where the solder stops. In high vibration environments.

[Chuck H]

Ok, I'm real interested in this thermal stripper.

R.G. Does it also have the advantage of helping minimize insulation re-shrink due to soldering heat? Being that the insulation has already been cut at high temp and all. That would be a great side benefit.

Chuck

[R.G.]

Ok, I'm real interested in this thermal stripper.

R.G. Does it also have the advantage of helping minimize insulation re-shrink due to soldering heat? Being that the insulation has already been cut at high temp and all. That would be a great side benefit.

Hmmm... I don't know about re-shrink. Probably/possibly not. I think most re-shrink comes from the heat conducted up the wire while soldering, maybe. The PTS-10 intensely heats a tiny area so fast that it does not conduct to the insulation even slightly away from the blade.

But here's a tried-and-true way to not reshrink/melt insulation. It's a two-step.
1. Work with a hot iron, and work FAST. Thermal dynamics being what they are, there is a transmission speed of heat up a wire, affected by the wire mass and external heat removal. If you solder much, you know this intuitively. Some wires you can solder while holding the other end of the bare wire, and not get burned if you work fast. The whole trick is to use a hot enough iron with enough heat capacity to get the joint under the iron to soldering temp immediately, make the joint, then get the iron off it before the heat can travel up the wire. A low temp iron is the surest way to melt things far away, paradoxically.
2. Use a heat sink. Those little aluminum sprung tweezers that used to come with cheezy soldering kits are perfect for stopping heat from heading up a wire. They're useless for soldering transistors, but they're great for stopping heat up a wire. So are a pair of needle nose pliers, but I lost my third arm back in the 60's and it's never regrown. 8-)

[mook]

Let's not forget "skin effect" ??? I don't know much about it really. It has something to do with the surface area of lead wire being the most active area. So it's been argued that stranded wire is better because of it's greater surface area. Probably has more to do with signal leads than power leads, but I'd be interested in more clarification.

Chuck

I thought "skin-effect" was only with AC.....not DC. Right?

[R.G.]

I thought "skin-effect" was only with AC.....not DC. Right?

Yes, it is. DC uses the full cross section of the conductor. AC causes electromagnetic fields inside the wire that forces the conductors toward the outside. Frequency is the big determiner. At frequencies below 20kHz, the change in resistance of a conductor is minimal, especially if there is any other resistance connected to the wire at either end... 8-)

Skin effect is not really a "skin". It's a graded distribution which increases the current density from the center out, while suppressing it at the center. The frequency determines the distribution. The "skin depth" is an arbitrary definition of how deep you have to go in the distribution to get X amount reduction in current density. There is still current below the "skin depth", just less. At high radio frequencies, sometimes massively less. But ignore skin effect at audio.

[mooreamps]

I thought "skin-effect" was only with AC.....not DC. Right?

It is more pronounced at radio frequencies ; above 100 mhz. It also is part of the reason why waveguide works the way it does.

-g

[R.G.]

Calculating skin effect

The first entry google returns when you put in "skin effect" contains this:

The skin depth is defined as the depth below the surface of the conductor at which the current density decays to 1/e (about 0.37) of JS. It can be calculated as follows:

\delta=\sqrt{{2\rho }\over{\omega\mu}}

where

ρ = resistivity of conductor
ω = angular frequency of current = 2π × frequency
μ = absolute magnetic permeability of conductor = \mu_0 \cdot \mu_r , where μ0 is the permeability of free space (4π×10−7 N/A2) and μr is the relative permeability of the conductor.
The resistance of a flat slab (much thicker than δ) to alternating current is exactly equal to the resistance of a plate of thickness δ to direct current. For long, cylindrical conductors such as wires, with diameter D large compared to δ, the resistance is approximately that of a hollow tube with wall thickness δ carrying direct current.

Putting in some numbers for copper at 20kHz,
ρ = 1.68E-8 ohm-M
ω = 2*pi*20kHz = 1.256E5
μr = 0.999994 for copper; call it 1!
μ0 = 4*pi*1E-7

And if I do the math right, we get skin depth for copper at 20kHz is
... drum roll ... 1.459mm.

For the approximation of a tube of wall thickness equal to the skin depth being the same resistance as a solid wire of the same diameter, the diameter of the wire has to be larger than 10x the skin depth, so the approximation holds for wires of diameter over 29mm, over 1 1/8 inches diameter. That's heavy for welding cable! Anything of smaller diameter never gets to the current distribution where you could call it a "skin". The whole interior conducts. Wire of diameter close to twice the skin depth is a bit over 2.9mm, which is AWG 9 size.

Net result? If you're not using frequencies over 20kHz, and wires less than 29mm diameter, skin effect is not going to be a big consideration. The copper wire's DC resistance is more fruitful to play with.

Pre-apologies if I did the math wrong. This is a favorite question on EE physics and fields tests, and it's been almost that long since I did one of these.

[soundmasterg]

R.G., how did you get your thermal stripper for $35? That link shows a price of $69. Maybe you got yours years ago or on sale?

It looks like a pretty useful device. If I got one of those and a solder pot, I'd be in business.....

Greg