Cable burn-in: Is it real or imagined?

[dmnc02]

Hello Dnmc02



It doesn't however to be a wise guy from a skeptical point of view are the cables changing or is it your perception that is changing. That's why I would like to see some kind of before and after measurement's.

I am a bit confused why any changes in the dielectric from burn in would have an effect in interconnect cables. It's a real potential issue with capacitors where the signal must pass through the dielectric material simply because of the basic construction of a capacitor. With interconnects the dielectric is used as a barrier to insulate the conductor runs. The signal does not pass through the dielectric like it does in a capacitor. Why would changes in the dielectric effect the sound in interconnects??

Rob

Rob, I am not the best person to answer your question in the last paragraph, as I do not have the required training.

As to your first paragraph, I just wanted to be absolutely clear that I fully agree with you that a set of measurements showing exactly what happens with burn-in would be ideal. However, I simply do not believe that having been unable (so far) to identify the correct set of measurements is sufficient ground to dismiss the evidence provided by listening tests.

 

[thedudeabides]

Not happy with the sound of your new amplifier, just wait until it has burned in!
Actually of course the only time when you might genuinely hear any differences in components is within the first 20 minutes until your ears have become accustomed to the new sound.
Keith.

Let's substitute amp for preamp and consider the Conrad Johnson ET5 or other preamps, electronics that have large banks of teflon capacitors.

The need to "break in" these type units is quite well documented but I suspect you will consider this BS also.

Happy Thanksgiving.

 

[DonH50]

I am a bit confused why any changes in the dielectric from burn in would have an effect in interconnect cables. It's a real potential issue with capacitors where the signal must pass through the dielectric material simply because of the basic construction of a capacitor. With interconnects the dielectric is used as a barrier to insulate the conductor runs. The signal does not pass through the dielectric like it does in a capacitor. Why would changes in the dielectric effect the sound in interconnects??

The dielectric acts like a distributed parallel capacitance. Using transmission line equations, which are not really all that valid in this situation, the impedance (Z) of a cable is roughly

Z = sqrt[ (R + jwL) / (G + jwC) ]

where
R = resistance (in series with the signal)
L = inductance (also in series with the signal)
G = conductance (in parallel with the signal, i.e. shunt leakage)
C = capacitance (in parallel with the signal)
j = sqrt(-1)
w = frequency in radians/second; w = 2*pi*f where f is the frequency in Hz

The kinds of things I have measured related to dielectric absorption, charge traps, and so forth have been well below audibility IMO. Didn't some company include a bias box with their cables a few years ago? No idea who or if they still do. The purpose would be to fill those charge storage traps in the dielectric, but that ignores some basic physics about the way the traps works (at least in my hairy-knuckled engineering understanding) and their magnitude. Bias helps but does not completely eliminate the problem, and the level is just in the mud relative to the signal level, again IMO.

Component burn-in is a completely different animal. And speaker break-in yet another.

IME/IMO! - Don

edit: Found a Wikipedia article, should have looked first but the t-line equations are burned into my little pea brain.

http://en.wikipedia.org/wiki/Characteristic_impedance

HTH - Don

 

[dmnc02]

The kinds of things I have measured related to dielectric absorption, charge traps, and so forth have been well below audibility IMO. Didn't some company include a bias box with their cables a few years ago? No idea who or if they still do. The purpose would be to fill those charge storage traps in the dielectric, but that ignores some basic physics about the way the traps works (at least in my hairy-knuckled engineering understanding) and their magnitude. Bias helps but does not completely eliminate the problem, and the level is just in the mud relative to the signal level, again IMO.

Don, Look back at post #86. Audioquest cables still include a bias box and Vandersteen bias some of the components in their crossovers. Obviously, they think this makes an audible difference.

 

[thedudeabides]

Component burn-in is a completely different animal. And speaker break-in yet another.

HTH - Don

Don, I agree but was responding to a previous post.

However, this type of dismissive, absolutist attitude appears to go well beyond the cable burn in issue.

Sigh!

 

[microstrip]

(...)
The kinds of things I have measured related to dielectric absorption, charge traps, and so forth have been well below audibility IMO. Didn't some company include a bias box with their cables a few years ago? (...)

Component burn-in is a completely different animal. And speaker break-in yet another.
(...)
IME/IMO! - Don

Don,

IMHO probably you measured the wrong thinks ... And what values are you considering for the threshold of audibilty?

BTW Harman uses batteries to bias the film capacitors of the top JBL speakers. As far as I know they did not present any measurements to justify this technique.

 

[Robh3606]

BTW Harman uses batteries to bias the film capacitors of the top JBL speakers. As far as I know they did not present any measurements to justify this technique.

Hello Micro

That's true they call it Charge Coupled Networks. They have measured them and they actually slightly increase the measured IM distortion levels. I use the technique in all of my DIY speakers. I am a believer but it's subtle and where you can hear it is in low level passages added smoothness and detail. There is the info from Greg Timbers who designs the JBL high end systems at the end of the thread that he sent us to post on the site. The justification is it sounds better.

http://www.audioheritage.org/vbulletin/showthread.php?3555-Bypassed-and-Biased-Capacitors

Rob

 

[DonH50]

I measured the usual things when trying to fully characterize cables. Not sure I remember everything, but:

Z and S parameters (includes frequency response)
DC shield and center conductor resistance
Leakage
Noise (long and short term)
Dispersion
Impulse and step response, hysteresis
DA, hysteresis
Loss tangent
TDR/TDT characteristics
Vibration/flex tests
Power (voltage) sweeps, looking for nonlinearity
Temperature sweeps, ditto
etc.

The noise and hysteresis we saw from charge traps and such we saw were generally <-100 dBc and were a royal pain in some cables because we needed much better than that. However I felt that was below audibility but as I said that was IMO. we did see higher spikes now and then but they were hard to capture. It was years back when I did anything really complete like that; most measurements were much simpler, typically DC plus S-parameters. Note that burn-in was not a consideration unless we saw measurements change during testing (usually found to be some other cause). And finally this was focused on RF/mW/mmW applications, but because I designed baseband data converters for IQ radar systems (among other things), DC to mW frequencies was all important.

This was not targeting audio testing and I am quite sure you and others could dispute any claims of mine by challenging the test methodology, manner and number of tests, etc.

I did mention earlier that we did in fact bias cables in some applications with some improvement. That was in a system targeting >140 dB dynamic range. The benefits did not seem evident until below 100 dBc, and did not completely eliminate noise related to charge traps in the dielectric. We eventually went a different direction using nitrogen-filled cables.

 

[Robh3606]

Thanks Don for the response and links

Rob

 

[FrantzM]

Is this technical documentation with measurements or is it anecdotal documentation?

Happy thanksgiving to you. I have already burned the bottoms of my pumpkin pies I did last night, I put them in the lower rack of the oven and so the crusts in the bottom are burned....so added a note to cookbook not to make that mistake again, and now, the dilemma I am in is I am not sure if I put the turkey in the oven at 10:30 or 11:30 AM, man its not going good this year for food and me.....my reputation is about to crash down with my grandkids....

Was is done with full knowledge?

Happy Thanksgiving People!!!

 

[andromedaaudio]

Capacitators and cables are not the same thing , althougfh i think transparent uses caps in his cables, they place for example in some cables a cap in parallel to create a filter so " only lower" freqs are arriving at the speakerconnectors .
New caps can sound a ittle harsh when first on the juice even a skeptic like me agrees on that

 

[DonH50]

I did not say capacitors and cables were the same thing, I presented the standard telegraphers transmission line equation that includes the distributed capacitance of a cable. An related my experience with the effects of dielectric charge storage, a capacitive-like thang.

I don't think I'm adding anything to this except more circles to go around, back to practicing... - Don

 

[microstrip]

I did not say capacitors and cables were the same thing, I presented the standard telegraphers transmission line equation that includes the distributed capacitance of a cable. An related my experience with the effects of dielectric charge storage, a capacitive-like thang.

I don't think I'm adding anything to this except more circles to go around, back to practicing... - Don

No one is saying such thing. However since both use dielectrics and conductors, are said to influence sound in similar unexplainable ways, I think everything we can learn form capacitors will probably be relevant to explain cable sound. Just MHO.

BTW, what was the bandwidth (low and high limits) of your measurements? And what was the value you were accepting as audibility threshold?

 

[thedudeabides]

Is this technical documentation with measurements or is it anecdotal documentation?

Hi tomelex,

I hope the turkey and stuffing had a better outcome but I'm sure the pies were just as good.

Basis for my statement is my own personal experience with the ET5 (started a thread where I documented my listening results) as well as other CJ owners. If I had to decide whether to keep the unit, after the first hour or even the first 250 hours of listening, I never would have done so. My previous pre was the CJ Premier 18 LS.

I think you can probably find technical reasons if you research. I did some initially and relied on owner experiences, which ended up being very similar to mine. Maybe we all had expectation bias or something else to explain all the similar experiences but I seriously doubt it. I'm not technically qualified (nor am I interested in this type of information) to interpret the "objective" data that may explain what happens when the teflon caps "break in".

As I've stated on many occasions, I totally trust my ears.

Best.

 

[DonH50]

@microstrip: I was responding to andromedaaudio's comments about capacitors. I am not sure if he was addressing me or another poster.

Frequency range was DC to whatever upper end was deemed appropriate for the system. That would range from a low of maybe 100 MHz to a high of 100+ GHz, but most of my work targeted X-band systems (8 - 12 GHz) with an IF of maybe 1 GHz. I'd guess most measurements were taken to 20 GHz or a little above, many to 44 GHz (upper end of one of our standard HP/Agilent/Keysight VNAs), only a few above that. Highest I ever dealt with was a few things up around 110 GHz, and a brief sojourn around 300 GHz but that was all waveguides. As I said, these were primarily baseband systems using in-phase/quadrature converters so response to DC was required. To the front-end RF guys, X-band was often baseband and 1 GHz they considered "DC". I am not really an RF guy, just worked around it a lot since I had to design or at least deal with the RF front end in spec'ing the data converters and associated circuits I designed.

As I said, these were not audio measurements, or at least not targeting audio systems. Measurements I have taken specifically on audio cables were much more mundane. Many of the cables were similar, with the caveat that virtually none were 75 ohm cables that are commonly used in the audio world (rader/ELINT/EW/com systems use 50 ohms almost exclusively, although CATV and video satellite links use 75 ohms). And as mentioned in my previous post I was assuming below -100 dBc was inaudible, or at least thoroughly masked by room noise, the music or movie signal, etc. I am aware some (many?) feel that too low a threshold.

microstrip, I am not trying to debate or prove myself credible to anyone else, just reporting what I found in the past and have acknowledged most won't understand nor care, and that it was not for audio systems. The main thing I wanted to highlight was that there are subtle things that happen in cables, but IMO they are too subtle to be audible. That said, I have not done a real controlled listening test, nor done measurements anytime recently that would bolster my position vis a vie audibility.

However, since you do not find my experience credible for audio, what exactly do you think should be measured? The catch for me now is that my current job is focused on different signaling (computer stuff, SAS/SATA/PCIe/DDR) and we have no good audio-frequency test equipment so I don't have an easy way to do LF measurements. The VNAs, SA's, and such have LF limits well above the audio band, and I do not have fA and nV meters as I had back then.

 

[andromedaaudio]

Hello don, no i was responding to a post by the dudababbies , as he entered a cap filled pre in to the equasion

greetings hj

 

[DonH50]

Ah, my apologies, thank you.