Measured (!) speaker cable difference!?

[preload]

What Amir posted is system level and the large variances were clearly due to amp roll off. In the results wrt this thread a reference no cable run shows no such amplifier roll off. The large variances are not believable nor is the shape of the curves.

Nope. If the FR variatiom were solely due to amp rolloff, then all cables would have measured the same. The fact that the cable FR's differed demonstrates that the cables were responsible (or more precisely, the cables interacting with the reactive load). The degree of variation is definitely believable because it can be explained by analyzing the circuit using basic LCR parameters. I don't have an engineering degree and even I can follow the math.

 

[preload]

This thread is not about the cable influence on frequency response. This thread is about the magnitude of the FR variation from cable to cable, and can be with a real speaker or not.

I need to make sure it's clear that the load matters. If the load is a non-inductive resistor, the cable properties won't make much of a difference. The op demonstrated this as well. Real life loudspeaker loads have varying impedance vs frequency. It's this property that interacts with cable dcr and inductance that causes the FR deviation. It's that simple.

 

[pma]

This is my typical measurement of speaker cable transfer function when loaded with a real speaker (more in the link already posted). It is a voltage transfer of

A(f) = V(speaker terminals)/V(amplifier terminals)

As I have already stated, the isolated cable frequency response when it is loaded with the real speaker is quite easy to measure and quite easy to simulate, if we have the speaker impedance already measured. I have prepared a comparison of REW measurement with VituixCad simulation of the cable frequency response when loaded with my CNO-T25 speaker and 47ohm parallel resistor. As VituixCad allows maximum zoom to 10dB vertical span, I have also accommodated the REW plot to the same.

This is the measured frequency response of the cable

Simulated cable frequency response

CNO-T25 impedance with 47ohm resistor in parallel

 

[preload]

As I have already stated, the isolated cable frequency response when it is loaded with the real speaker is quite easy to measure and quite easy to simulate, if we have the speaker impedance already measured. I have prepared a comparison of REW measurement with VituixCad simulation of the cable frequency response when loaded with my CNO-T25 speaker and 47ohm parallel resistor. As VituixCad allows maximum zoom to 10dB vertical span, I have also accommodated the REW plot to the same.

View attachment 148573
This is the measured frequency response of the cable

View attachment 148574
Simulated cable frequency response

View attachment 148575
CNO-T25 impedance with 47ohm resistor in parallel

What are the L and R parameters of your cable?

 

[pma]

What are the L and R parameters of your cable?

0.07 ohm + 5 uH approx.

 

[audio2design]

Nope. If the FR variatiom were solely due to amp rolloff, then all cables would have measured the same. The fact that the cable FR's differed demonstrates that the cables were responsible (or more precisely, the cables interacting with the reactive load). The degree of variation is definitely believable because it can be explained by analyzing the circuit using basic LCR parameters. I don't have an engineering degree and even I can follow the math.

You can follow the math but didn't do the math .... See curves below that had largest variance and guess what, all the cables are similar as the biggest variable is the amplifier (Amp-A).

Since you can follow the math, tell me why in the test associated with this thread there is almost no discernable difference with the 8.4R resistor between the 1.5mm and 6mm cable. Since you can follow the math, tell me how big a db difference there should be then explain why it is not there.

Now, since you understand the math, tell me the comparative likely inductance profile of the 1.5 and 6mm cable and how that is likely to impact frequency response.

 

[solderdude]

We can doubt OP's study but we cannot say its wrong/flawed till we have determined that it is indeed wrong (would need to verify this experimentally).

1: The measurement can't be repeated because the exact measurement setup is not known.
2: It is clear from the results the measurement is flawed as has already been explained.
The plots don't make sense unless speaker wire current crept into the measurement path.

Regarding the article by Amirm, I have read it and I am not doubting the contents. However, we cannot simply use it as an "authority" to determine that OP is wrong.

This is just one (old) study. There are more. Then there is common electronics sense as well. They all tell the measurement is flawed.

Cables all do measure differently. There is nothing shocking about that. You can minimize this by using large diameter cable and keeping distances short. One can play with geometry and materials to create a different L and C as well.
When the R is low the sound impact is low. When it is low enough it is below audible levels.

 

[KSTR]

As for the amplifier relevance, both in general as well as in those measurements from that AES paper, it all depends on the measurerment strategy and the desired entity. We either want to see the effect of the amp or the cable alone, or both in combination, it all depends....

In my proposed setup, only the cable influence is measured, the amp's contribution is normalized out. The amount cable influence of course depends on the amp output impedance, amp impedance should be significantly lower than the cable impedance to see maximum effect from the cable. As explained, with a current drive amp, no change of speaker voltage (and SPL FR output) will form regardless of any series impedance... which is technically correct as per goal we want to measure the effect of the cable alone, and with current drive there is no effect by definition... obviously, the FR will be all over the place, completely spoiled (but could be EQ'd to flat), yet the cable itself makes no difference.

@pma's setup is slightly different. He also uses a ratio, though a different one: speaker-side voltage normalized to amp-side voltage. For very low amp output impedance it yields the same results, but for higher impedances the result start shifting away from representing the true speaker terminal voltage change. Again, using current drive as the exteme case for illustration, the voltage on the speaker will not change no matter what cable but the measured ratio is changing as it is measuring (V_speaker)/(V_speaker + V_cable)=1/(1+V_cable/V_Speaker). Of couse Pavel is aware of that and pointed out the influence of the amp impedance.

In the Davis Experiment, they used the same strategy, normalizing to the amp output, albeit they did it point-wise by manually forcing the amp output to be 1V at any of the test frequencies. Therefore, the effect characteristics seen are the same. Change of the (unknown) amp output impedances is reflected in the results even with one and the same cable, as it should be. And vice-versa, change only cables with same amp and get similar but topologically different results: the leg in the voltage divider where the variable is introduced now is the other leg, the bottom leg. Finally you can alter both legs at once and get yet another set of plots, less meaningful ones (two variables at at time -- no no).

 

[pma]

and with current drive there is no effect by definition... obviously,

For this very reason, I made a current transfer module at link level about 20 years ago. The purpose was to eliminate audiophile feelings of different sound of different link cables. It did work, however it of course did not convince the audiophile crowd .

The principle can be seen below, the current level was similar to industrial 4-20mA.

 

[pma]

The amount cable influence of course depends on the amp output impedance, amp impedance should be significantly lower than the cable impedance to see maximum effect from the cable.

Klaus, I have some troubles with your explanation. Please see below the simulated effect of change in output impedance of the voltage source (=amplifier) (R15) from 1ohm - 10ohm to isolated cable voltage transfer. Nada, nothing. Voltage transfer of the linear 4-pole is of course independent on source impedance.

V(A) would be amplifier output voltage, L9+R14 would be the cable and the rest is crossover loaded with simulated drivers. V(in) would be voltage at speaker terminals. V(in)/V(A) would be cable voltage transfer. I hope that R15 = 10 ohm is high enough to simulate any bad amplifier.

 

[preload]

You can follow the math but didn't do the math .... See curves below that had largest variance and guess what, all the cables are similar as the biggest variable is the amplifier (Amp-A).

No, all the cables are NOT similar. There is still a 0.3dB variation across over 3 octaves across cables with this particular combination (amp A / speaker A).
Yes, I agree with you that common tendency for the cables 7-12 to roll off when combined with Amp A and Speaker A is largely influenced by Amp A (since this amp had rising output impedance with freq). However, it doesn't change the fact that it also confirmed variation in FR introduced by the speaker cables.

Since you can follow the math, tell me why in the test associated with this thread there is almost no discernable difference with the 8.4R resistor between the 1.5mm and 6mm cable. Since you can follow the math, tell me how big a db difference there should be then explain why it is not there.

A scientific approach would be to first ask what what the total DCR was between the 1.5mm and 6mm cable lengths. If they are similar, I would not expect a large dB difference. Secondly, I would want to know if the scales were normalized for the purposes of presentation (that's how I would present it, btw = normalized). And thirdly, I would want to know if the cables were swapped out in one sitting vs. measured on separate occasions and/or whether the gains were adjusted in between measurements.
A non-scientific approach would be to start with your conclusion (i.e. "that cables make 0 measure difference") and assume that any data that contradicts your world view is inherently flawed. A non-scientific approach would also be to completely dismiss pertinent findings (i.e. FR difference across cables) because you found ONE component that didn't make sense to you right away (i.e. the absolute dB level with smaller cables).

Now, since you understand the math, tell me the comparative likely inductance profile of the 1.5 and 6mm cable and how that is likely to impact frequency response.

Now, IF you understood the math, you would already know that the geometry of the cable (particularly spacing of the conductors) would influence inductance, and the cross sectional area/length/resistivity would influence DCR - so why ask.

 

[preload]

0.07 ohm + 5 uH approx.

Your simulation looks like it makes sense. A relatively large gauge cable with a short run and average inductance appears to be introducing only 0.1dB fluctuations (spanning several octaves) in FR across the audible range.

 

[pma]

Your simulation looks like it makes sense. A relatively large gauge cable with a short run and average inductance appears to be introducing only 0.1dB fluctuations (spanning several octaves) in FR across the audible range.

I would like to mention that I posted both simulation and real world meaurement and the results are almost identical.

 

[preload]

I think we still do not understand each other. We do not argue that there is variation with frequency. The conversation is that between various cables, the difference is not significant (thus audible):
View attachment 148336

I would argue that +/- 0.3db spanning 3+ octaves is "potentially audible."
For a while I played with similar shelf filters in Roon to tune my Genelecs, and I could notice a 0.3dB shelf from 2kHz-20khz when I was looking for it.

 

[preload]

I would like to mention that I posted both simulation and real world meaurement and the results are almost identical.

As they should be! They're just LCR simulations after all. Ohm's Law is a law, not an estimate (like preference scores, haha).

I think we all know that a standard, paired,12GA copper/stranded speaker cable isn't going to introduce significant FR deviations in most cases. But when we start using longer runs, or using smaller gauge, or non-copper materials, or the impedance curve of the loudspeaker is exotic, THEN we can start to see FR deviations that have more and more potential for audibility (i.e. greater dB deviation over multiple-octaves).

 

[Raindog123]

and I could notice a 0.3dB shelf from 2kHz-20khz when I was looking for it.

"level matched"?

 

[KSTR]

Voltage transfer of the linear 4-pole is of course independent on source impedance.

This is entirely correct, full ack. You actually get different transfer functions for different cables even with current drive... but, and that is the point here, under current drive the actual speaker voltage has not changed at all, thus the cables have no impact. So, a highish source impedance tends to reduce cable net effect as seen at the speaker voltage whereas the cable transfer function remains the same and thus is not a good representation anymore.

 

[pma]

under current drive the actual speaker voltage has not changed at all,

That's correct, however the cable voltage transfer function remains unchanged, as there is no other option. We were looking at the subject from different views. As the current drive of speaker box with crossover is meaningless ......, and high amplifier output impedance results in modulated frequency response of the "normal" speaker, I still see the voltage transfer of the cable investigated in isolation as important and addition of the amplifier output impedance as misleading, especially for the layman audience. The influence of speaker output impedance was shown as well in post #150. Together with cable contribution.

 

[DimitryZ]

It's not controversial to state that an amplifier with a high output impedance, paired with a speaker with a lowish and FR variable impedance will result in substantial FR variations.

 

[audio2design]

Now, IF you understood the math, you would already know that the geometry of the cable (particularly spacing of the conductors) would influence inductance, and the cross sectional area/length/resistivity would influence DCR - so why ask.

Bud, grow up and stop trying to impress people. No one cares.

The person who did the experiment claimed same length. Maybe some reading and less judging? The 1.5 and 6mm cross sections must measure different with an 8.4 ohm load. Basic engineering which you lack. That is just one of the obvious measurement mistakes and suggests other methodology errors. Typically due to spacing the 1.5mm will have lower inductance than 6mm round. The graphs don't make sense.


The Amp A shows 1.5db of drop but cable to cable is as you noted 0.3db so thank you for validating EXACTLY what I said which is the large drop is almost exclusively from the amp and not the cable (my statement to refute an equivalence claim wrt magnitude of changes, as the experiment wrt this thread had large differences cable to cable but the reference no cable is flat indicating the amp is relatively flat).

Perhaps try to look erudite to someone who doesn't have multiple engineering degrees, worked in academia and industry and who can do this stuff in their sleep.