Warszawa

The sound of speaker cables

Recommended Posts

5 minutes ago, rabski said:

I can't be bothered to find the relevant bit, but the vertical scale doesn't seem to give a baseline. A 20dB difference would be massively apparent if it was a difference between 0dB and 20dB, but a difference between say -140dB and -160dB would not be.

From the article I'd thought the idea was for it to show the difference in the waveform at either end of one of the cable conductors, with the values representing the ratio on a dB scales such that 0 dB was no difference. Although looking again the varying offsets from the 0 dB line even at lower frequencies doesn't make any sense.

Whatever has been done that graph is definitely misleading!

Share this post


Link to post
Share on other sites
17 hours ago, TheFlash said:

Sorry but what is a "reflection", how does it manifest itself, and what has this PCB/digital circuitry thing to do with speaker cables?

Be gentle and clear. We don't all work with this stuff. Thanks.

Lets say a cable or pcb track carrying a signal has a Characteristic impedance of say 75ohm at a given frequency. If it is terminated by a 300ohm resistance there will be a  reflection of a portion of the signal back to the amplitude of which is related to the difference in the impedances. The same thing could happen at the source end if say the source end is say 5ohms so there could be multiple reflections back and forth until all the signal is finally transmitted. If the source impedance is 75ohms then there will be no return reflection. This is what I meant by "back matching". 

I mentioned that it can happen with digital pcb traces just to show that length per se is not a factor in whether reflections occur or not. Length does matter if standing waves are considered but lets not make it anymore complicated than it already is.

Your last comment is the most important one ie what has it got to do with speaker cables. The short answer is I dont know.  Ive known of this work that Max has been doing for some time and discussed it with him. As I said though these reflections occur at such a high speed that even multiple reflections should only affect very high frequencies. LTSpice simulation shows this also. I suppose this could affect some amplifiers.

Max's measurements show something is going on but quite what and how that relates to reflections is not clear at least to me. If someone can prove or disprove this link then Im interested.

I hope this is of help.

  • Like 2

Share this post


Link to post
Share on other sites
14 hours ago, hifinutt said:

I think it's called rigor mortis  :D

Will it reach the parts, Heineken refresh?  Of so I'm out there tonight.  

Share this post


Link to post
Share on other sites
2 hours ago, zeta4 said:

I mentioned that it can happen with digital pcb traces just to show that length per se is not a factor in whether reflections occur or not. Length does matter if standing waves are considered but lets not make it anymore complicated than it already is.

But length with relationship to wavelength at that frequency is what it's all about! It's only a factor when the trace or cable is a meaningful fraction of a wavelength.

It has absolutely nothing to do with speaker cables, analog interconnects or AC mains cords. In fact it only starts to apply to longer digital S/PDIF cables.

Edited by Speedskater

Share this post


Link to post
Share on other sites
2 hours ago, Speedskater said:

But length with relationship to wavelength at that frequency is what it's all about! It's only a factor when the trace or cable is a meaningful fraction of a wavelength.

It has absolutely nothing to do with speaker cables, analog interconnects or AC mains cords. In fact it only starts to apply to longer digital S/PDIF cables.

We have measured a phenomenon and have postulated an explanation that has been peer-reviewed by two highly experienced electronic engineers who agree that the conclusions are correct.
You obviously disagree, so could you please explain why there is a difference in the error voltage, shown in fig 3, that correlates directly with the characteristic impedance of the cables?

7 hours ago, MartinC said:

From the article I'd thought the idea was for it to show the difference in the waveform at either end of one of the cable conductors, with the values representing the ratio on a dB scales such that 0 dB was no difference. Although looking again the varying offsets from the 0 dB line even at lower frequencies doesn't make any sense.

Whatever has been done that graph is definitely misleading!

7 hours ago, MartinC said:

From the article I'd thought the idea was for it to show the difference in the waveform at either end of one of the cable conductors, with the values representing the ratio on a dB scales such that 0 dB was no difference. Although looking again the varying offsets from the 0 dB line even at lower frequencies doesn't make any sense.

Whatever has been done that graph is definitely misleading!

The graph shows the voltage between the black terminal at the amplifier and the black terminal at the speaker. The abscissa, or x axis, is the frequency in Hz. The ordinate, or y axis represents the voltage dropped across the cable at the relevant frequency. The dB scale is relative and is a feature of the Room Equalizer Wizard (REW) when used as a spectrum analyser, to measure voltage with respect to frequency in real time. 0dB was chosen for the short circuit, as it was easily repeatable. White noise is chosen as it represents every musical note and is constant in amplitude across the audio band.

As the voltage drop across the cable gets higher, the trace rises up graph and represents loss, which is a measure of the error in the cable. The greater error, the further up the trace is and the lower the trace is,  the error is less. The lower the trace is, the lower the error. The short circuit, trace 1,  is the ideal. The DC resistance of each cable causes the rise in the traces at lower frequencies and all the cables have a similar loss up to about 500Hz. The rise in the traces at higher frequencies represents a greater voltage error. With a perfect cable, there would be the low frequency rise at the left due to resistance only and the trace would be essentially horizontal. 

The trace 2 of the two flat strips Zo 18ohms is very close to the ideal with just a slight rise above 10kHz. All the other cables 3,4,5,6 and 7 all show an increase in loss as the characteristic impedance rises. Note that cable 7 is identical to cable 2, except that the spacing between the inductors is increased significantly in case 7, which lowers the capacitance and rises the inductance and hence the characteristic impedance from 18ohms to about 1600 ohms. The DC resistance is not changed. The video clip shows this with sound. The brighter sound from cable 7 is due to the higher loss at high frequencies. The most important conclusion is that the performance of a cable changes with the geometry and it is clearly shown. 

Share this post


Link to post
Share on other sites
Moderator
15 minutes ago, Townshend Audio said:

The graph shows the voltage between the black terminal at the amplifier and the black terminal at the speaker. The abscissa, or x axis, is the frequency in Hz. The ordinate, or y axis represents the voltage dropped across the cable at the relevant frequency. The dB scale is relative and is a feature of the Room Equalizer Wizard (REW) when used as a spectrum analyser, to measure voltage with respect to frequency in real time. 0dB was chosen for the short circuit, as it was easily repeatable. White noise is chosen as it represents every musical note and is constant in amplitude across the audio band.

I can't quite understand why if you measured a voltage drop, you do not display this on the graph. Displaying it as a dB measurement seems a little disengenuous, as obviously a 10dB or more difference is very large, but according to your statement here, that merely reflects some voltage drop.

I must admit your explanation seems a little baffling. Your equipment is designed to measure the voltage drop on the cable. REW does not measure a voltage. Your graph shows a substantial dB difference, but you are now saying this actually represents a voltage measurement. Please can you therefore point to where the actual thing you measured is noted?

Edited by rabski

Share this post


Link to post
Share on other sites
43 minutes ago, Townshend Audio said:

The graph shows the voltage between the black terminal at the amplifier and the black terminal at the speaker. The abscissa, or x axis, is the frequency in Hz. The ordinate, or y axis represents the voltage dropped across the cable at the relevant frequency. The dB scale is relative and is a feature of the Room Equalizer Wizard (REW) when used as a spectrum analyser, to measure voltage with respect to frequency in real time. 0dB was chosen for the short circuit, as it was easily repeatable. White noise is chosen as it represents every musical note and is constant in amplitude across the audio band.

Could you write an equation for how Ga(f) and Gs(f) relate to give the dB values plotted?

Share this post


Link to post
Share on other sites
42 minutes ago, rabski said:

REW does not measure a voltage.

REW does actually have an Impedance measuring mode that I believe makes use of electrical measurements. It's not intended for what has been done here though.

Share this post


Link to post
Share on other sites

To help your people out. This is a chart of a typical coax cable with a 75 Ohm @ 1 megahertz  Radio Frequency Characteristic Impedance.

While it's rather approximate, after scaling, it should give a good guide to what audio frequency band, characteristic impedance's might be.

Frequency Impedance
1 megahertz 75 Ohms
100,000 Hz 110 Ohms
20,000 Hz 160 Ohms
1000 Hz 800 Ohms
100 Hz 2000 Ohms
20 Hz 4000 Ohms
Edited by Speedskater

Share this post


Link to post
Share on other sites
Moderator
12 minutes ago, MartinC said:

Could you write an equation for how Ga(f) and Gs(f) relate to give the dB values plotted?

My issue Martin is that the graph is the only place I can see where any actual measurements are shown, but the graph is stated to show a voltage difference, whereas the scale is in decibels.

In other words, unless I have missed it, there are actually no figures given anywhere for what was measured. What is shown as at least 10dB difference, may in fact reflect an extremely small difference. Either way, the graph is simply incorrect, as it suggests something completely different to what it actually portrays.

  • Upvote 1

Share this post


Link to post
Share on other sites
2 minutes ago, rabski said:

My issue Martin is that the graph is the only place I can see where any actual measurements are shown, but the graph is stated to show a voltage difference, whereas the scale is in decibels.

In other words, unless I have missed it, there are actually no figures given anywhere for what was measured. What is shown as at least 10dB difference, may in fact reflect an extremely small difference. Either way, the graph is simply incorrect, as it suggests something completely different to what it actually portrays.

Hence me asking for the equation to try to clarify this point...

Edited by MartinC

Share this post


Link to post
Share on other sites
Moderator
Just now, MartinC said:

Hence me asking for the equation to try to clarify this point...

I know. It's just I'm not sure there is any equation.

Share this post


Link to post
Share on other sites
1 minute ago, rabski said:

I know. It's just I'm not sure there is any equation.

The dB values plotted must represent something.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Recently Browsing   0 members

    No registered users viewing this page.