John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
AX tech editor
Joined 2002
Paid Member
Right you are, Jan, no doubt. But this model example could be considered as a very big exaggeration of possible real situations, fast transients in audio signals, etc. And this example gives a hint (I am guessing), how deep GNFB amp-speaker interactions could affect low-level signal constituents.

"Very big exaggeration of possible real situations" means, of course, that it is not a possible real situation.

We had another member here years ago who invented the term 'first cycle distortion', something similar as what you showed.
Several people suggested he limit the signal to a reasonable bandwidth, say 30kHz. Of course he refused, because it would destroy his carefully set up argument. I can prove anything when allowed very big exaggeration.

As to the second part of your post, unfortunately I have no idea what you are talking about, or what you are guessing, apologies.

jan
 
Disabled Account
Joined 2004
Do you think you can hear a difference between a 10' and an 80' speaker cable?

Now if you make the LC&R the same but just the length different will you hear a difference. (I am sticking to circuit theory on that question as few here know much about field theory.)


Very different cable geometries would be needed. As for field theory when someone here likened the limiting case of skin effect in coax to waveguides only DF96 spoke up (beat me to it). TEM waves are rather their own subject.
 
And how much cable length is required to introduce a delay of 1 uS?
That will depend on the characteristic impedance of the line, the impedance of the load at the specific frequency, the length of the line, and the impedance of the source at the specific frequency.

At very high frequencies, the line to load mismatch will cause delays in the 5 to 10 uSec range as long as the amp output is sufficiently stiff. If the amp is incapable of correcting the reflections as they occur, the settling time will get worse.

At lower frequencies, the R of the cable will make the situation even worse.

If you want to see the delay and the reflection, just use an HP8721A.

Roughly 1ns/ft for a signal. Insulation can move that up or down a skosh.

Within a coax with a dielectric relative permittivity of 4, the prop velocity will be half lightspeed, roughly 2 nS per foot. For zip, this is a useful rough value, but thicker insulations can increase this to almost 3 because it spaces the wires farther apart.

Hey! Look! A squirrel!

WHERE??

jn
 
Disabled Account
Joined 2004
And how much cable length is required to introduce a delay of 1 uS?

Ed, I would suggest bundling 6 pair RG400 in the two lengths. I don't think the resistive difference matters much but you could pad the short one with the tiny R left over.
 

Attachments

  • coax.JPG
    coax.JPG
    40.4 KB · Views: 212
Last edited:
PMA, what subject do YOU want to talk about on this thread?
Apparently, everything that I have brought up recently, even vetted material, and not written by me, is considered 'useless' by you, to the point of covert insult. What is it that we NEED to debate or learn here? Contact me in private, if you wish, but please give me a clue.
For the record, I spent a fair amount of time locating that early AES paper from B&K for YOU, specifically. I first thought that it would reinforce your position on two tone testing, and maybe give even you some further insight. On reading the paper, myself, I found it more 'self-serving' to B&K equipment than I would have liked, but I did not know that in advance. I was trying to give some 'interesting and relevant' input to a colleague, which happened to be you. The paper was too long to put up, but if you would send me your e-mail address, I can send it to you personally.
I am only trying to help.
 
That will depend on the characteristic impedance of the line, the impedance of the load at the specific frequency, the length of the line, and the impedance of the source at the specific frequency.

At very high frequencies, the line to load mismatch will cause delays in the 5 to 10 uSec range as long as the amp output is sufficiently stiff. If the amp is incapable of correcting the reflections as they occur, the settling time will get worse.

At lower frequencies, the R of the cable will make the situation even worse.

If you want to see the delay and the reflection, just use an HP8721A.



Within a coax with a dielectric relative permittivity of 4, the prop velocity will be half lightspeed, roughly 2 nS per foot. For zip, this is a useful rough value, but thicker insulations can increase this to almost 3 because it spaces the wires farther apart.



WHERE??

jn

So now we can proceed to how many factors come into play on just a 10'-100' speaker cable?

Then we could go on to how would we test to see what cable length does create a difference that someone could hear?

That would raise the issue of how many variables were actually under control in the listening test. Also did we really account for all of the variables?

Which brings us back to, does the length of speaker cable affect how the audio power amplifiers global feedback behaves? And is this an issue that one should be concerned with.

The rational answer would be I don't know, but my opinion is...

Jan,

If the issue of TIM is well understood, then one could calculate for any given design, how much would be present and use that in design trade-offs. Just saying that you know the ground rules to avoid it is the first step. Based on the thread, it seems that is not as well known as you seem to assume.

ES
 
Status
Not open for further replies.