December 2011 Issue - page 7

Article Index
December 2011 Issue
page 2
page 3
page 4
page 5
page 6
page 7
All Pages



In the reply to a reader’s letter Noel stated that audio cables do not have a characteristic impedance. This is incorrect. Even a piece of string has a characteristic impedance. Whether it matters sonically is another question but I think its important to get the facts right.

The usual equation used for the calculation of CI is Zo = ( L/C)1/2 where L and C are the values per unit length. However this only applies at high frequency as we shall see.

The full equation for the calculation of CI is Zo = (( R + jx2xpixFxL)/(G + jx2xpixFxC))1/2 where R, L,C,G  are per unit length and F is any frequency. G is the conductance of the dielectric and is usually very, very small so can be ignored. So a cable such as an audio cable has a CI depending upon these parameters. Cable length does not matter.

At low frequency the terms jx2xpixFL and jx2xpixFC are very small compared to R so R dominates. As frequency increases the above terms become much larger so the equation simplifies to (L/C)1/2.

A cable specified as 75ohm, is that calculated for high frequency not audio frequency. A typical coax cable will have a CI very much higher than 75ohms at audio frequencies. So the idea of loading 75ohm coax with 75ohm will not work from the impedance matching point of view as well as it’s a bad way to treat a line stage as Noel says.

So unless you deliberately  manipulate the cable’s parameters most audio cables will be a mismatch to their source and load impedances. This will result in some of the signal being reflected back down the cable and reflected again. The number of reflections being related to the amount of mismatch. The speed of propagation of the reflected signal is related to the dielectric material of the cable but is usually between 3 to 5ns per meter. This effect can be seen experimentally on the leading edge of a square wave transmitted down a mismatched line. The leading edge has a series of small steps at twice the expected delay time.

If you adjust the load (or source) impedance towards the nominal CI the height of the steps decrease. to a null. If the cable did not act like a transmission line this would not happen. Delatraz described the experimental set up in detail with the same results. Similarly Richard Black. I have verified these results myself.

Geoff Mead

Technically that is so and I suspected someone might pick this up! But as you say the reactive components fall out of the issue at low frequencies and, as they are the determining factors of Characteristic Impedance, in practice a cable does not effectively have a characteristic impedance at low audio frequencies.

Time delays of nanoseconds (10 exp-9 seconds) where cycle times are 0.05 milliseconds or more are not especially significant I suspect and unlikely to explain cable sound differences, but it is interesting all the same. NK


I have an old Townshend Rock turntable which I am modding and trying to bring up to scratch. The platter is very early and consists of a formed dish approx 1.5ins thick and filled with plaster of paris. Due to the method of manufacture and poor design it requires an alloy centre to locate the platter correctly. However it is impossible to centre the platter correctly as one cannot establish a datum to work from. I therefore am considering turning up a new platter using acrylic as the material.

My question is can you advise which grade of acrylic I should use and a possible supplier? The platter is quite heavy in its original form and I would like maintain or slightly add to the original weight. Manufacturing the platter from scratch enables me to make sure everything is correctly in line and to the best standard I can achieve.

I hope you can assist me to keep this old warrior alive and fighting; at its best as it can give quite a respectable account of itself. During my attempts to ‘up the ante’ I have pulled together various redundant parts and brought them into use, it comprises a Rega RB300 arm rewired and counterweight modded, and a power supply put through a smoothing device which enables me to adjust speed whilst running, and maintains the use the original AC motor. This is all finished off by the addition of a rebuilt Decca Supergold (London) cartridge.

The addition of the Decca cartridge has had the most effect: the music has a vitality that is quite astounding and challenges my Orbed GyroDec SE, which makes me even more determined to carry out what will be the final mod to this deck.

John Lancaster

PS The original platter weighs 5.5lbs

Hi John. Grades of acrylic are a bit beyond our knowledge I am afraid to say. I suggest you contact Townshend Audio for this info. NK


I am so pleased to see your old Buying Guide back up and running on the internet. I was using it on a very regular basis up until it’s disappearance from Net and quite frankly, (as daft as it may sound) when it went I was lost without it! Thing is, all the stuff that I wanted in the mid 1990’s but couldn’t afford is now available 2nd hand at good prices (e.g. eBay and cash converters) and this guide is absolutely ideal for helping out with research on this older (but golder) stuff (for example turntable cartridges and mid end CD players as upgrades). Apart from product reviews,I find it very useful for checking this 1990s stuff’s price when new to give a good feel for where the item sits in the pecking order and welcome help with likely component matches.Today’s Buying Guides are pretty useless to me. I’m a lot happier now I’ve got the use of yours again.

Richard Franks

Thanks for writing in Richard. We resurrected this guide in response to your request and this of so many others who wrote in from around the world. Now we understand the value I hope we can improve this listing slowly. NK


Add your comment

Your name:
  The word for verification. Lowercase letters only with no spaces.
Word verification:


Hi-Fi World, Powered by Joomla!; Hosted by Joomla Wired.