Cable differences are real

[mansr]

Cable differences are real and measurable. Here is the frequency response of a few random interconnects I had lying around.

The variations seen with the RG-59 (red) are due to the mismatch between its 75 Ω characteristic impedance and the 50 Ω inputs and outputs used for the measurement. The same thing happens with the Maplin cable (green) which also seems to have a few defects causing the irregularities in the curve. The AudioQuest (purple) is totally bonkers and the real reason I posted this.

 

[Blumlein 88]

That AQ cable sucks. It is already down 1 db by 25 mhz.

 

[pkane]

Cable differences are real and measurable. Here is the frequency response of a few random interconnects I had lying around.

View attachment 103777

The variations seen with the RG-59 (red) are due to the mismatch between its 75 Ω characteristic impedance and the 50 Ω inputs and outputs used for the measurement. The same thing happens with the Maplin cable (green) which also seems to have a few defects causing the irregularities in the curve. The AudioQuest (purple) is totally bonkers and the real reason I posted this.

AQ cable works to reject some RFI, then, but only if you have an FM radio transmitter at exactly 47MHz near-by. Who knows, may actually have a positive effect on some circuit somewhere in the world

 

[AnalogSteph]

The AudioQuest (purple) is totally bonkers and the real reason I posted this.

That must be a severe impedance mismatch in there, makes me wonder whether there isn't actually a break at one end. That would show up as substantial loss across the board in a 50 ohm system though... must be some very peculiar cable stock (some sort of twisted pair or even twin lead perhaps?).

 

[Ron Texas]

 

[Killingbeans]

20KHz is what, roughly 2/3 of a pixel?

 

[preload]

Cable differences are real and measurable. Here is the frequency response of a few random interconnects I had lying around.

The variations seen with the RG-59 (red) are due to the mismatch between its 75 Ω characteristic impedance and the 50 Ω inputs and outputs used for the measurement. The same thing happens with the Maplin cable (green) which also seems to have a few defects causing the irregularities in the curve. The AudioQuest (purple) is totally bonkers and the real reason I posted this.

What on EARTH? Why do we care about the frequency response in the MHz-range? Here, let me fix your graph so it only shows the audibly-relevant range:

 

[Helicopter]

MHz. Who cares? But yeah, don't buy audioquest. I think we settled that.

 

[Sgt. Ear Ache]

MHz. Who cares? But yeah, don't buy audioquest. I think we settled that.

The people who have them real revealing systems care! That's who!

...oh, and them sweet golden ears!

 

[Rock Rabbit]

You are measuring exactly lambda/4 with 0.64c velocity and 47-48 Mhz, you have 1 m of good cable and almost 0 volts at lambda/4 a complicated way to measure cable length. Nothing wrong with that cable

 

[mansr]

That must be a severe impedance mismatch in there, makes me wonder whether there isn't actually a break at one end. That would show up as substantial loss across the board in a 50 ohm system though... must be some very peculiar cable stock (some sort of twisted pair or even twin lead perhaps?).

I think it's a twisted pair with a shield connected at one end only.

 

[Sal1950]

What on EARTH? Why do we care about the frequency response in the MHz-range?

 

[Wes]

cables suk - use a waveguide

 

[Helicopter]

Are we talking about RCAs for S/PDIF?

For audio, you will never need more than 90kHz and then, only for training puppies and rats. 20k is good for humans, and 16k is probably fine for most people old enough to buy fancy gear.

I use @Wayne A. Pflughaupt 's recommendations for best noise rejection in my cables, tying the shield in on both ends.

 

[mansr]

Are we talking about RCAs for S/PDIF?

The RG-58/59 are general-purpose coax cables. The rest are intended for analogue audio.

 

[anmpr1]

There are a lot of excellent reasons to never buy anything from an outfit like Audioquest. Cable response at 50 megahertz is not one of them.

 

[mansr]

Same selection of cables into a 75 Ω input:

 

[Robin L]

 

[mhardy6647]

View attachment 103880

 

[mansr]

Someone mentioned S/PDIF. Here's what the S/PDIF output from an iFi Nano looks like using two different cables into 75 Ω.

The AudioQuest obviously isn't intended for this. Nonetheless, and despite the mangled waveform, the zero-crossings remain clean. It would probably work just fine.