4

I'm using two shift registers to control 16 digital outputs using only 3 pins on my microcontroller. The parallel outputs carry <= 20 mA DC, their sole purpose is to power LEDs. My microcontroller pins do the following:

  • One is a serial binary data stream (5 or 0 Volts DC). In my case it's a 16 bit periodic pattern which mutates every second.
  • One is the input clock pulse for the shift registers. It sends a HIGH pulse for each bit in the data stream. Depending on how I tune the program, the frequency could be anything from 50 Hz to about 16 MHz.
  • One is the output latch clock pulse for the shift registers. It sends a HIGH pulse for every 16 bits in the data stream.

Does it matter whether I use solid, multi-stranded or Litz wire in this project?

I've read that the main point of Litz wire is to reduce AC resistance at high frequencies. I suppose nothing here really qualifies as AC, and I'll admit the question is mostly academical. But I'm a bit curious since there are some (possibly) high frequencies involved. This formula and these values from Wikipedia give a skin depth of 65 µm for copper at 1 MHz, so the skin effect certainly is significant, at least for true AC.

Emil Lundberg
  • 143
  • 4
  • 3
    The main point of multistrand wire at logic level voltages and currents is typically flexibility of the wire, and longevity against repeated flexing - a single-strand wire will crack at a flex-point far easier than a bundle of much thinner strands. – Anindo Ghosh Jul 15 '13 at 20:14

2 Answers2

7

For all practical purposes, there is no difference between stranded and solid wire. The only exception is for very high speed signals (>1 GHz), where solid is arguably better.

The main purpose of stranded is to make the wire more flexible, and has nothing to do with AC resistance at high frequencies.

There is something called the Skin Effect, where high frequency signals tend to travel on the surface of the wire. But stranded wire behaves as a solid conductor in this regard, because the electrons can travel freely from one strand to another. Litz Wire, on the other hand, has insulation between the individual strands and so does behave better with high frequency signals (how much better is open for argument). I have never seen a commercial product use Litz Wire (ignoring audiophile people). I'm sure there is a practical use for Litz wire, but suffice it to say that it is rare.

  • Actually, Litz wire is very common in cheap, thin headphone cables from my experience. The reason definitely isn't audiophoolery, since the headphones were mostly in the $5 range. – AndrejaKo Jul 15 '13 at 20:22
  • @AndrejaKo You're right, but it is also not used for it's "Litz-ness" since you don't have much skin effect on the 40+ gauge strands with 20 KHz max frequencies. I suspect that it is the low cost of the enamel insulation, high flexibility, and the labor required to assemble it is very cheap. –  Jul 15 '13 at 20:42
  • I completely agree. Do note that I did not claim that it was used because of the skin effect. Perhaps it's simply cheaper/more available than tinsel wire. – AndrejaKo Jul 15 '13 at 20:48
  • @AndrejaKo No problem. It's all good. –  Jul 15 '13 at 21:01
  • More on Litz wire: http://www.w8ji.com/skindepth.htm (near the end) – Phil Frost Jul 15 '13 at 21:17
  • Ah, of course. I did read about Litz wire, but then went on to mix up the terms. What I really meant was whether Litz makes a difference in this case. I realize now that the main point of logic circuits is voltage, not current, so I suppose they don't suffer too much from the skin effect. – Emil Lundberg Jul 15 '13 at 21:51
  • 1
    @EmilLundberg Some signaling standards are current sensitive, while others are voltage sensitive. Also, current and voltage are only independent when talking academically. In any practical circuit voltage and current are intertwined. As for Litz wire, in 20+ years of being an EE I have never had a reason to use it-- odds are that you have a couple of years before needing to even consider it. –  Jul 15 '13 at 22:01
  • @DavidKessner Hahaha, you make it sound like it's purely based on time in the industry. "Welp, I've been an engineer for 25 years, time to start using Litz wire!" – NickHalden Jul 16 '13 at 00:33
  • 3
    @NickHalden That's not what I said. Everyone knows that 30 years is the Litz Wire anniversary, geez! 25 years is the unnecessarily complicated vacuum tube amp anniversary. 20 years is Gold Plated Through Hole Resistors. And for the newbies, the 1 month anniv. is "CPU out of discrete transistors". –  Jul 16 '13 at 01:34
  • I've seen litz wire used in induction hobs. I suspect it's common in that sort of high-frequency, high power applications. – Connor Wolf Jul 16 '13 at 01:56
  • Some SMPS transformers use Litz, I suspect it's the ease of shaping it that makes it easier to get higher density of copper in a limited space. – Spoon Jul 17 '13 at 19:28
5

If by single-strand, you mean "solid", solid tends to break through repeated bends, but stranded does not. Stranded is a bit easier for me to work with-- just remember to tin the wire before soldering.

Scott Seidman
  • 29,274
  • 4
  • 44
  • 109