I'm lost , which thread ??
this one ?
http://www.diyaudio.com/forums/showthread.php?s=&threadid=102083&perpage=10&pagenumber=3
wait.. probably this ?
http://www.diyaudio.com/forums/showthread.php?s=&threadid=28448&perpage=10&pagenumber=8
this one ?
http://www.diyaudio.com/forums/showthread.php?s=&threadid=102083&perpage=10&pagenumber=3
wait.. probably this ?
http://www.diyaudio.com/forums/showthread.php?s=&threadid=28448&perpage=10&pagenumber=8
john curl said:
I have page after page of notes.
Tips from you.
I appreciate your time/knowledge and your sharing.
Hi all
this thread seems to be going full -steam since I last logged on.
I'd like to take up the point Bob raised some posts ago. The time delay of speaker cable is about 3 nS per foot (from sqrt(LxC)). So if you have 20 feet of speaker cable that is 60 nS or 120 nS round trip, as Bob said as much.
There will be reflections on this line to a pulse, but whether audio triggers such pulses is one question, the answer to this is probably no.
However, there ought to be reflections back and forth because the amplifier output impedance is pretty low (let's say <0.1 ohm and effectively a short circuit to incoming) while the cable impedance (from sqrt (L/C)) is around 50 ohms, so it will never match an 8 ohm speaker.
Normally the reflections in a short line (the line needs to be a lot longer to generate standing waves) should damp exponentially, but maybe they react with the output coil to give some interesting transients.
We can easily use a simulator to generate a transmission line model of the speaker lead, and use an 8 ohm load, to investigate, provided the model does not cut off the frequency response (because a lumped transmission line model cuts off more abruptly than a real Tx line)./
I'll try this unless someone else does, but I suspect this may be getting to the root of the issue?
BTW all of you may know L is in henries and C in farads, but another way of specifying these is ohm.s for L and s/ohm for C. Isn't it neat that multiplying the two cancels the ohms and dividing cancels the seconds?
cheers
John
this thread seems to be going full -steam since I last logged on.
I'd like to take up the point Bob raised some posts ago. The time delay of speaker cable is about 3 nS per foot (from sqrt(LxC)). So if you have 20 feet of speaker cable that is 60 nS or 120 nS round trip, as Bob said as much.
There will be reflections on this line to a pulse, but whether audio triggers such pulses is one question, the answer to this is probably no.
However, there ought to be reflections back and forth because the amplifier output impedance is pretty low (let's say <0.1 ohm and effectively a short circuit to incoming) while the cable impedance (from sqrt (L/C)) is around 50 ohms, so it will never match an 8 ohm speaker.
Normally the reflections in a short line (the line needs to be a lot longer to generate standing waves) should damp exponentially, but maybe they react with the output coil to give some interesting transients.
We can easily use a simulator to generate a transmission line model of the speaker lead, and use an 8 ohm load, to investigate, provided the model does not cut off the frequency response (because a lumped transmission line model cuts off more abruptly than a real Tx line)./
I'll try this unless someone else does, but I suspect this may be getting to the root of the issue?
BTW all of you may know L is in henries and C in farads, but another way of specifying these is ohm.s for L and s/ohm for C. Isn't it neat that multiplying the two cancels the ohms and dividing cancels the seconds?
cheers
John
I get the impression that many of you have not seen my
ancient article which mostly deals with this subject.
www.passlabs.com/np/spkrcabl.pdf
😎
ancient article which mostly deals with this subject.
www.passlabs.com/np/spkrcabl.pdf
😎
Nelson, I could not pull up the page, but thanks for checking in. We must get these young 'whipper-snappers' in line. 
myhrrhleine said:
Sorry to go a little off-thread, but since hearing occasionally appeared to be offered as proof that the effect of output coils is audible:
I would agree with your first sentence - but then you go right on and in the second imply that in a way humans (their hearing) are superior to measuring??
In exhaustive research over the last decades human hearing has been quantified to an extent that really leaves little room to suggest that there is still "something out there" that is not known (at least in this context). These tests involved batches of "calibrated" listeners, statistic analyses etc, done at universities and medical auditory research centres, and the results were published inter alia in J.A.E.S., auditory-medical journals etc. Much of this was in the more serious interest of acousto-medical assistance for hearing disability rather than hi-fi.
On the way there these tests also showed that hearing acts like all our other senses in that subjectivity plays a large role and it can be "bluffed" - medically a totally natural thing, but apparently very hard to grasp by a minority of technologists who simply will not accept the results of research! (But that is quite off-thread - not that my humble repetition will change minds that could not even be budged by the research itself).
Hi Estuart,
---I have also bootstrapped my I/P stage based on the output signal, which (the latter) is flat to about 250kHz. It out performs a cascode topology.---
I am interested by this. Is there somewhere where the schematics can be seen ?
---I have also bootstrapped my I/P stage based on the output signal, which (the latter) is flat to about 250kHz. It out performs a cascode topology.---
I am interested by this. Is there somewhere where the schematics can be seen ?
John Ellis,
Your post #705. This has been done, and as you expected with no remotely audible result. But please do not ask me to quote now; it was more than 15 years ago. Yes, it is easily done on Spice. The line is simply too short for this to play a role, especially with the totally overriding effects of any real loudspeaker impedance as a load.
Your post #705. This has been done, and as you expected with no remotely audible result. But please do not ask me to quote now; it was more than 15 years ago. Yes, it is easily done on Spice. The line is simply too short for this to play a role, especially with the totally overriding effects of any real loudspeaker impedance as a load.
John C
Nelson's paper appeared to address the relative merits of cables from different manufacturers. It does not really matter if any one uses a gold-plated cable with "zero oxygen copper" or suchlike or another, they will still have inductance and capacitance. The point I was trying to get the debate to open up on was the contribution of speaker reflections in the presence of an output coil in the amp. I'm not sure this has been specifically adressed up to now.
You certainly lit the blue touch paper with this topic, and you might have also missed the point I made in a previous post that actually you may be right, but the reasons have had to be dragged out.
cheers
John
Nelson's paper appeared to address the relative merits of cables from different manufacturers. It does not really matter if any one uses a gold-plated cable with "zero oxygen copper" or suchlike or another, they will still have inductance and capacitance. The point I was trying to get the debate to open up on was the contribution of speaker reflections in the presence of an output coil in the amp. I'm not sure this has been specifically adressed up to now.
You certainly lit the blue touch paper with this topic, and you might have also missed the point I made in a previous post that actually you may be right, but the reasons have had to be dragged out.
cheers
John
john_ellis said:
Hi John,
If you want to play around with transmission lines in SPICE, it's not necessary to use lumped approximations. There's a model called LTRA that is a true distributed model. I believe most SPICE simulators support it. This post has an example LTSpice project and links to two articles about the model by the UC Berkeley guys who created it.
john curl said:
This young whipper-snapper is still waiting for a hint of an answer from you Mr. Curl.
By the way, my elder daughter of 27 is working her doctorate in developement biology, my son of 25 is an IT consultant now in Mexico DF, and my younger daughter at 23 is midway her Civil Engineer career, go figure my youth.
Moderators may be called to the floor to attest my posts have unnerringly being respectfull of everyone, no matter how far I may find myself from their positions. I consider myself deserving the same level of respect which sadly I have not got from you Mr. Curl, and note I am not even challenging the intrinsic badness of an output coil or not, just trying to learn.
You may keep ignoring my requests, feel confindent I will not ask any more from you from now on, in case you feel this is harassment.
Rodolfo
Some people are convinced that the plastic insulation of signal leads significantly affect the sound even if the send and return leads are not physically connected i.e. air gap between the the plastic...
...can anyone explain this ?
...can anyone explain this ?
Johan Potgieter said:
While branching off to a new thread might be best, I don't know how.
For many years, temporal alignment of loudspeakers was thought a waste of time because nobody could hear the phase error.
The 'golden ears' heard something, but the 'lead ears' did not, and so the 'golden ears' were dismissed as not knowing what they were talking about.
It's not that temporal response could not be measured, it was assumed inaudible and therefore not measured.
It was many years before temporal alignment was accepted.
It may take a long time for the mind to detect a difference. Therefore standard A, B, X testing may be inadequete for subtle differences.
What I'm suggesting is that we should not automaticly dismiss something because we do not notice it when we listen, or it doesn't show up when we do our measurements.
rdf said:
I used to own a nice stacked pair, and I still have 4 of those
fabulous 100:1 step up transformers. They are the basis of my
comment on the minimum series impedance being 0.5 ohms
in an electrostat.
But back to the subject of unterminated speaker cables, there is
no question in my mind that the high frequency resonance can
grab the amplifier's loop by the nose and create all manner of
IM products in the audio region.
😎
myhrrhleine said:
Good point, furthermore, it has been only recently that the active nonlinear nature of the hair cells in the basilar membrane has been characterized.
Linear frequency related distortions not considered significant (meaningless with sinusoidal tests) do alter waveforms which in turn becomes important under this light.
Rodolfo
- Status
- Not open for further replies.