The Smith Chart and why power cords matter

From your first post: "Here's an example what happens when the impedance of the "transmission line" does not match the "load impedance". "

The transmission line in this case (since we're talking about electricity for your system) starts at the power station. It's the power grid. It's not the area between a wall receptacle and an amp.

 

Have a nice day, gents.

 

This Smith Chart thingy is beautiful! I want one for my wall.

I will never be separated from my pile of ex-computer cables gathered from the years at work.

I'll never read any justification for their products from companies like Shunyata, Audioquest, etc. The only company I trust to give me probably-genuine information is Blue Jeans. And even then, I'm aware that I could be falling for another kind of marketing.

But hey, I once thought I heard a difference made by a mains lead. It is filed under highly-doubted, because I'm a human, and our senses are not infallible. But it's also filed under just-maybe... And rat-hole-I-won't-go-down.

I'm not a scientist, much less an Electrical Engineer. But I'm aware that some of our members are. And I like to listen to them, because they sound different.

/snarky-sceptic
/grumpy-old-guy

 

I read the Shunyata justification. From what I can parse......

It *seems* like one of the things they are trying to say is that noise within the amp, specifically bridge rectifier switching noise, can be 'dissipated' by the power cable. Let's just say I'm skeptical. Without wanting to be a measurement nazi, that is absolutely the sort of thing you could directly measure - both on the AC and DC side. It is also the sort of thing that you can mitigate by by using tube rectifiers (no switching noise), by properly-sized snubber caps on traditional Si rectifiers, or by using modern zero-recovery SiC diodes.

But I still don't see how characteristic impedance of the power cable is what makes the difference here. Especially in the absence of any impedance measurements on the products in question. A lot of the Shunyata products have 'data' that shows less noise with their cables. But many of these products also have integrated filters that despite the mumbo jumbo seem like they wouldn't be much more different than traditional C-L-C power filtering. Separating out the mumbo jumbo filtering versus the cable is far more effort than I'm willing to put into this.

 

Don't be a hoarder It is OK to let go of those DB25, token ring, and parallel SCSI cables. They had a good life.

 

I bought an expensive Audioquest USB cable once. My computer started sporadically not recognizing the DAC. Went back to USB printer cable. That ended my "quest" to hear things that only instruments can measure.

 

The theory of the case I can tease out of that article (which, interestingly, is largely claims about what a power cable doesn't do and examples of power cord designs that may have more downside than upside) is that the power cord is effectively part of the transformer's primary winding and therefore has an effect on related electromagnetic properties.

I don't have the knowledge to evaluate that claim myself, and to the extent that it makes sense I don't know what exactly you would do to a power cable to make it better at this and not worse, or how someone would know which one to buy. I might ask an equipment designer how they feel about that, but I figure the fact that the places I've bought from have never recommended or attempted to upsell an expensive power cable is sort of an implied answer.

Something else it says that I take issue with is the idea that the power cable is not transmitting a signal, because it most certainly can, even if we discount the main power signal itself. Powerline networking at home, ripple control signals, etc... and given some of the troubleshooting I've had to do, these can definitely cause audible effects to connected audio equipment (although I believe powerline networking frequencies are MUCH higher than should ever be humanly audible, so it was probably a weird interaction with some other connected device). So it's pretty weird for them to say that.

 

Of course, I meant power cables but forgot to type the word.

Not that I don't have some outdated stuff in dusty corners. I never worked with token ring, but DB25 (Oh hey, come to think of it, I have a tin full of bits that includes pins for those) and parallel printer cables are very possible. I used to have a small set of thin-ethernet Ts and Terminators: a tiny complete network at the back of a drawer.

Honest, USB and Ethernet usually just works. And when it doesn't, more-expensive is not the answer.

I learnt more about cat-5 cables from Blue Jeans than I ever knew when actually managing computer systems. I had no idea that most of them are actually nowhere near spec, and yes, of course that can be measured. Thing is, they mostly still work. Ethernet is wonderfully resilient. (But not for music I hear some people cry! Music is different! Delicate! Bollocks, I reply: ask a network engineer, not an audiophile.)

Then why should that stop at the plug/socket? Actually EE*s, is there any possibility of that being true? Because if it isn't, the whole idea of that final 1.5m making a difference, unless it is more than just wire, goes down the pan

Real problems; stuff that it doesn't take an "audiophile" to hear. I used to follow Archimago's blog, eg, if you have a genuine hum which is passing down a USB cable, you need an isolator. or to go completely optical.

Now I'd better stop all this. Because I seriously gave up these arguments some years ago. But it's like nicotine: it can get you when you off guard, even decades later. LOL.




* ee is Tamil for housefly.

 

Counterfeit ethernet cables are unfortunately a thing. When I was working one of our new buildings was built with it and it caused all sorts of seemingly random issues till we finally figured it out. Cost the contractor a lot of money to remediate.

 

Iconoclast?

I had a prepro that used DB25-to-RCA for its multichannel out. Those seem expensive for what they are...

 

DB25s are the best… in the studio. Not sure why you’d need them in your house, other than maybe some kind of multichannel thing. But even there, I never see them in consumer gear.
Regardless, if you find a *quality* db25 cable you don’t need, put it on eBay, they’re not cheap to make or buy so there’s a market for them.

 

I used to buy cheap flyleads. If I'd known what I know now, I probably wouldn't have done, but as I said, problems related to a cable were almost insignificant in number. Building cabling is another matter: no skimping there! Had a very trusted contractor. Two-man company: in suits for network consultancy, in jeans for installation work.

Counterfeit? I can believe it. But even when not, there is a spec, but no testing, and anyone can print "CAT whatever" on the cable/package. Shameless Plug: buy from Blue Jeans, Each flylead is tested and certified. Don't remember the pricing, but it won't be ridiculous like Audioquest et al. Plain ordinary ethernet flyleads, but certified. I do not promise they'll sound any better. Ethernet is just so resilient you have to try hard to stop it working!

Not something I buy much of these days. But when I do: Amazon Basics!

 

Interesting. Did not know they were still used, so my bad. My DB25 cables went away with my last LPT printer. Maybe I should hoard more lol. I do still use RS-232 to this day, so at least some of these things have aged well.

I have several interconnects from BJC, agree they are excellent value and without all the marketing hype.

 

Yeah, DB25 is just a connector, like 1/4" or XLR or whatever. You can run whatever signal you want through them, within reason. For studio use for example, you often see them used to connect 8 balanced channels (using 24 of the 25 available pins, 3 per channel) to an audio interface. That's a lot easier and smaller than 8 individual XLRs or 1/4" TRS. They also come in other sizes, like DB9 or DB36 etc. I hear they're also used in aviation to connect electronics, since they can be screwed in to be secure.

 

To take this back on topic, I think this is a really telling fact about power cables. Even if all the physicists and EEs* were wrong and power cables did make a difference, wouldn't component makers jump at the chance to include them with their products to ensure they sounded their best? Instead, the people spreading the message that they matter are those with a vested interest in that message - the people making the cables, selling the cables, or writing reviews where power cable makers advertise.

Not to get all ASR or anything, but this is one area where they're not wrong.

*I'm neither, but I do believe them because they're usually very smart and spent a long long time learning really hard shit to be able to do what they do.

 

In the first 45 seconds of the video, information crucial to understanding the concept of standing waves in electrical conductors is presented. It is quite important to keep in mind the relevance of

ƛ = C/f where C = speed of light, 3 x 10^8 meters/S.

If f=50 Hz then ƛ = 6,000,000 meters or 3728.23 miles.

In the first 45 seconds of the video they give an example of how a rope moves when the wavelength is very large compared to the rope.

So when it comes to power cables of length approximately 1 meter in length, they represent only 0.000000167 of the wavelength. Not even on the radar. Application of SWR, Smith Charts begin at Radio Frequencies (RF) which are commonly considered at least 540 KHz. Even in the example in the video they quickly moved to 100 MHz for demonstrations. There are exceptions such a Very Low Frequency radio such as used in communication with submarines (3 ~ 30 KHz) but that is a very special case.

Power Cords may or may not have an influence on the sound quality but in this case the science originally cited doesn't apply.

 

See? Real scientists just sound better.

 

I do not understand the science related to cables or electronics. I know enough to blow stuff up, stupid shit. I have had many many different cables and power cables in many systems and they all sounded different, why I don't know and its what I have heard many times over the last 35 years. I always think of tube amps that test horribly and sound fantastic or SS amps that are the very best with all of the tests and sound like dried dog shit. There you go

 

To me, with one possible differing experience, they have all sounded the same.

There are some things that I just plain don't believe in. But hey, I just posted elsewhere that ego is not a certificate of infallibility so I'd better swallow my own medicine: Maybe that's why they all sound the same to me.

 

Power cables make a noticeable difference when the chain is high enough quality.
To me this is difficult to swallow, as @Beefy pointed out - reasoning before the audio cable (random installation cable and outlet) and after (transformer, rectifier diodes + caps, questionable wire quality used inside the electronics and whatnot) gives little justification to even think about a 1 m long cord in between. Even so I trust my perceptions before my reasoning.
Keep in mind I am much less of a scientist (although trained as one) as @atomicbob and prone to disregarding science when it does not agree with what I perceive.
So far in my testing transmission line effects do not appear to matter, at all - nowhere in analog audio chain.

What does appear to matter to sound as far as cable construction :

• conductor cross section (more appropriately conductivity) appropriate to the load - and this is not what you will see connected to the printer as adequate
• magnetic field effects - why twisting the cables into braid have an effect in the first place
• electric fields interacting with insulation and connectors (more or less parasitic capacitance effects)
• metal to metal transitions and non-ideal contact between conductors
• metallurgy (cryo treatment is real thing, monocrystal is real thing, impurities are a real thing)

All of the above is hard to measure relevant to our hearing.

What does not appear to matter for analog audio cables:

• transmission line effects (characteristic impedance and the Smith plot)
• skin effect and Foucault's currents, Litz and flat wire as remedy - in extremes the effect can be noticeable but for practical audio cable cross sections the effect is over hyped
• cable capacitance and cable inductance - again, it could be issue (especially with high feedback amp connected), but for practical cables it is really not