[nick_charles]

Quote:

So after reading a significant portion of the thread that I'd like to make something very clear. This is only to test the wire. Not the source or the destination. The great thing about dB is you can just add them and subtract them. Seriously its like the coolest thing ever, waaaay better than trying to compare things in volts haha. 
 
Also I just realized that it was Nick Charles himself who gave me the link, thanks! I think its a bummer how that thread turned into a flame war. Hopefully that doesn't happen in this one. You did a great job with analyzing the data, I think.
 
So again, the test I'm going to perform will only going to show the difference that the wire itself makes.  

From my own experience I hope you are fairly thick-skinned and I wish you the very best of luck, I would reccomend you read some of PIO2001's posts on testing.
 
I could have done better on my tests but my resources were limited so my tests were largely by proxy (using DA and AD intermediaries which inherently add some error) and from what I have learned since I would do some things differently. I am fairly happy with the data analysis though.
 
I have two pieces of advice (take them or leave them , I will not be offended);
 
1. Do lots and lots of trials , I found substantial random measuring variation between trials, I suggest at least 10, 20 is better and average the results. You want to establish if the between cables variance is greater than the within cables variance
 
2. Once you have established the actual db differences between cables ask yourself how will you determine if the differences might be significant or not ?
 
 

 

 

[Pio2001]

Quote:

So again, the test I'm going to perform will only going to show the difference that the wire itself makes.  

 
But... that totally depends on the impedance of the load. The frequency response, for example, is just a voltage divider game between the cable's impedance and the load's impedance.
 
You can easily get losses of 0.000001 dB with an oscillo (1 MOhm) , and 0.3 dB on a speaker (3 Ohms) with the same cable.
 

 

[Uncle Erik]

Quote:

I think its a bummer how that thread turned into a flame war. Hopefully that doesn't happen in this one.

 
Prepare for it.  This is not an academic debate.
 
There are tens or hundreds of thousands of dollars at stake for those who make money from cables.  You can be sure that their, er, "representatives" will show up to mire results in fear, uncertainty and doubt.
 
Poke a few holes in someone's income and they'll make a stink.

 

[brconner]

Quote:

 
But... that totally depends on the impedance of the load. The frequency response, for example, is just a voltage divider game between the cable's impedance and the load's impedance.
 
You can easily get losses of 0.000001 dB with an oscillo (1 MOhm) , and 0.3 dB on a speaker (3 Ohms) with the same cable.
 

 
 Thats true. Which is why I'll be measuring the impedance of the wire. If you have the impedance you can calculate what would happen based on the load. I'll also have a load on the wire as well. Probably 32 or 16 ohms since that seems to be a fairly standard value used in headphones.
 

 

[Anaxilus]

 

[AVU]

These $5000 interconnects are on sale for ONLY $2500!  You should get them to test!
 
look at this thread under "cable DEALS" 

 
http://www.head-fi.org/forum/thread/503194/cable-deals#post_6794362

 

[Currawong]

The reason I suggested plugs as well is, I have, with a very cheap and nasty plugs on some cables, clearly had harsh distortion. However, what would be interesting to test is wire that people believe makes a clear tonal difference.  It would be interesting if it were possible to measure why.  It's a pity though those Nordost Heimdalls linked are sold, they are one very good example of tonal distortion caused by cables.

 

[grokit]

I want to know the SQ "difference" between gold-plated, rhodium-plated, cryo'd, and plain old metal connectors.

 

[Anaxilus]

Gold and rhodium is strictly for corrosion resistance.  There is no other value than that.  Cryo is for strengthening a metal by altering the molecular alignment.  Common practice in race engineering.  So I guess your interconnects become faster.  

 

[brconner]

Quote:

Gold and rhodium is strictly for corrosion resistance.  There is no other value than that.  Cryo is for strengthening a metal by altering the molecular alignment.  Common practice in race engineering.  So I guess your interconnects become faster.  

I actually wasn't sure what those were either, thats really interesting. From an electrical standpoint, so long as there is a solid connection between two materials the effect is extremely small and not noticeable until the signal frequencies are in the Megaherzt range or higher.

 

[grokit]

Agreed, with the caveat that rhodium is much harder and more therefore more durable than gold; it also has lower resistance and therefore higher conductivity. Also agreed that the actual effects of this higher conductivity are negligible regarding connector plating, so rhodium is more for durability when you are constantly plugging and unplugging. As far as the cables, silver has the highest conductivity, but is also the most prone to corrosion. Copper is of course the next highest in conductivity but also corrodes easily.  There are many cables made of silver-coated copper, but there are a few manufacturers that swear by gold plating. It might be cool to try and measure these differences.

 

[brconner]

Here are what I'm planning on testing so far assuming I can find some good prices and get things in reasonable lengths.
 
I want to test the following gauges: 20, 22, 24 and 26 AWG.
 
For each gauge I'd want to test pure copper, 0.999 silver, and 14k gold.
 
As I said earlier, I'd like to test each configuration at 3, 2 and 1 meters. I'll probably test something short like 10 cm as well.
 
Depending on prices, I may not test everything I've listed. I'm thinking in particular the gold wire could get rather expensive at 3m.

 

[grokit]

Quote:

Here are what I'm planning on testing so far assuming I can find some good prices and get things in reasonable lengths.
 
I want to test the following gauges: 20, 22, 24 and 26 AWG.
 
For each gauge I'd want to test pure copper, 0.999 silver, and 14k gold.
 
As I said earlier, I'd like to test each configuration at 3, 2 and 1 meters. I'll probably test something short like 10 cm as well.
 
Depending on prices, I may not test everything I've listed. I'm thinking in particular the gold wire could get rather expensive at 3m.

Silver and gold plated wire wouldn't be so bad, price-wise. But all the best wire, even the best copper is still quite expensive. And that is what we all want to know, is if these super pure, exotic and yes, expensive wires make a difference, right? Perhaps a retailer with a liberal return policy would be willing to play along.

 

[brconner]

What I'll be getting is just the wire. Not a cable, so I will be getting the highest quality wire that I can. I'll also get some cheap stuff (RadioShack) to compare with.

 

[nick_charles]

Quote:

What I'll be getting is just the wire. Not a cable, so I will be getting the highest quality wire that I can. I'll also get some cheap stuff (RadioShack) to compare with.

Okay, but several exotic cables use wire arranged in rather specific ways, braided, with different dielectrics and so on, so that testing the gauge is just one variable, if you look at Pear, Audioquest and others they have the strands or cores for solid core configured in very particular arrangements which they claim to alter skin effect , RFI and so on. By testing only two variables (length and gauge) you are missing all the exotic arrangements you could have done such as having two different gauges of wire in the same cable , individuallly sheathed strands, and so on.
 
Thus your findings however useful only scratch the surface of the cable construction argument, the exotic cable proponents will simply say "but you did not test cable X which has a,b,c".