|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
75.233.62.192
In Reply to: RE: Cable article link posted by sbrians on September 30, 2013 at 11:38:55
Very good artocle, however it helps to be in the cable business since 1996. Having doen a LOT of simulations and measurements, I can tell you that interconnects do matter a lot. The revelant parameters are not losses or inductance though. These are capacitance, skin-effect and crystal lattice symmetry. All of these are related to electric field, which is mostly what you have in interconnects. In speaker cables you have inductance being the most important because of strong magnetic fields.
Looking at this loss and that loss is not very interesting IMO. It it the non-linear effects across amplitude and frequency that are more important.
I do agree that passive crossovers are up there with the main things that need improvement, however the best way to go about that is up for debate. If one gets passive crossover
components closer to the ideal texbook in performance, it can be evry bit as good as bi or tri-amping.
Steve N.
Follow Ups:
"crystal lattice symmetry" ????????????????
Well, metal is a crystal, no?
Like they say “everything maters” but often the issue is to what degree?
The cool part for the experimenter, one can build a test rig that actually lets you hear what cables do to the signal while passing it.
Obtain two good microphone or wide band signal isolation transformers (like Jensen etc) same model etc.
The high Z primary (for a mic trans) or input sides + and – for each are connected one transformer for each cable end, thus the signal across each end of the cable is the input for each transformer.
Then, the transformer outputs are connected in series BUT are out of polarity so that two exactly equal input signals cancel out.
Fwiw, a further refinement is a trimmer pot so that when both inputs are connected to the same source, the null or cancellation is as deep as possible.
It is easier to use the inputs from a good sound card etc but with that set up it is risky to audition loudspeaker cables this way.
Either way, when passing a signal, ANY difference between the input and output ends shows up as a signal that didn’t cancel out and so there is a sum which you can amplify and listen to with headphones etc.
You can also disconnect one end temporarily and audibly compare the magnitude of the sum or difference relative to the actual signal.
Best of all you can do this “sum difference auditioning” with any music you choose, any level you choose and with speaker cables (using signal transformers and using resistive attenuators if running at more than a watt or two).
Best,
Tom
This null test will tell you what the cable does to the signal that passes through it, but it won't show the effect of the cable on the source. (Extreme example: the capacitance of the cable makes a poorly designed amplifier oscillate. Most fun when this is a tube amplifier and one sees the plates glow red.)
Ultimately, only end-to-end performance matters, namely what makes it through the air and into your ears. Technical examination of the pieces of a playback chain are useful for diagnostic and engineering purposes, but have no functional value to an audiophile.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Hi Tony
Your right, the end to end difference test only shows the difference one end to the other or like applying it to an amplifier only lets you hear the difference between the input and output..
For an amplifier, capacitive loading was most problematic on the early SS amplifiers. Back in the 70’s when I built a lot of Electrostatic speakers and worked at an amplifier company (Grommes), it was the tube amps which were most comfortable driving that kind of load and an examination of the Safe operating area protection circuits would usually reveal an output stage which didn’t tolerate much capacitive reactance before “protecting”.
Now days, it would be unusual to find an amplifier that was parasitic due to a cable’s capacitance or normal drivers although once in a while they find one at work (where amps sometimes have a large amount, many hundreds of feet of cable) . The twin low loss coax cable that leads out to the measurement tower outside only has about 2K Ohms capacitive reactance at 20KHz and it’s 125 feet long.
For speaker cables, it is the series inductance and Rdc which are larger factors, series L especially in generic extension cord..
Your right about measurements, once the design is done, the use of engineering as a refinement tool is also over. Sadly, measurements are used to sell, to imply quality. Worse, measurements are intended to compare input to output in various ways or to a standard, to define how accurate the device is by quantifying the differences.
For the audiophile, this is a problem because subjective enjoyment is not strongly tied to signal accuracy. For example, a large level of reverberant sound can make a choir sound heavenly but at the same time, it can prevent you from understanding the spoken word or prevent you from creating a strong mono phantom image and so on.
If one wants to “hear” how faithful a device or loudspeaker is, one can make a generation loss recording of the DUT as well as the recording medium as a reality check.
It is ironic that most electronics will pass many generations before sounding significantly degraded but it is rare for a loudspeaker to go more than one or two generations before sounding very funky, they are the weakest link in the chain in several ways.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: