macgyver10 said:
Hmmm..I looked at the link you posted. I did not see any specifications. So I cannot comment on their voltage divider design.
Is this system a simple sine sweep? Does it have the ability to run a DC along with one or two different sines superimposed?
The ultimate test setup will be able to push one or two hundred watts from DC to say, 50 Khz, it will need to be able to run three simultaneous signals, and be able to discern the relative timing of the three, at the loads after a crossover network.
I do not think your LMS setup is capable of what is needed, but am willing to help if I can..
Cheers, John
jneutron said:
And the amp you will be using to do that is...?
Not a lot of audio amps are DC coupled. Do you have one in mind, or would it be custom/modified?
Hi All,
At the rate this thread is proceeding, if the worthy Mr Hawksford were to have written his article now, he might have needed to factor the speed of this thread in as well compared to the "slow" speed of electricity in his findings.
John,
I respect your right to your opinion, but no alternative numbers, info, etc.?
Pardon for not re-posting a whole lot of entries (clog things up), but I would refer you to my post #178 (p. 18), where I gave my measurements. Those lead to the conclusion that with normal cable under normal domestic conditions and in audio, none of L, C and R come close to being able to affect response. (I also referred to the situation that qualitive statements seem to abound in this field, whith very little quantative data around.)
You later question the ability to be able to measure the inductance of a single wire. Why? I laid similar lengths of tested cable in the same geometry, which would indicate the presence of extra inductance - which could be easily calculated and subtracted. But the different values obtained with different cables seemed to show that this was not serious. But I should perhaps have indicated this shortcoming at the time, and I apologise.
The salient point in an engineering approach was, however, the finding that none of L, C and R values would make any difference under practical circumstances and in audio. (How the suspected absence of this condition keeps popping up in many of the excellent mails posted here.) Why go to further refinement if first approximations indicate that contributions would lie way up somewhere in the r.f. region, in addition to the fact that one's approximations were worst case? That would have been a worthy physics procedure, but of what value in audio?
I am not one to soap-box my long experience with audio, but while I also have analysed these parameters with a view to at least find some position in the wilderniss of qualitive statements, I would wholeheartedly agree with the several posts that put cable parameters in the bigger picture of the orders of magnitude larger contribution of loudspeaker and room characteristics.
I would also like to comment on a later post by you, but do not seem to get quotations from several pages in my replies.
Thus regards for now!
At the rate this thread is proceeding, if the worthy Mr Hawksford were to have written his article now, he might have needed to factor the speed of this thread in as well compared to the "slow" speed of electricity in his findings.
jneutron said:
John,
I respect your right to your opinion, but no alternative numbers, info, etc.?
Pardon for not re-posting a whole lot of entries (clog things up), but I would refer you to my post #178 (p. 18), where I gave my measurements. Those lead to the conclusion that with normal cable under normal domestic conditions and in audio, none of L, C and R come close to being able to affect response. (I also referred to the situation that qualitive statements seem to abound in this field, whith very little quantative data around.)
You later question the ability to be able to measure the inductance of a single wire. Why? I laid similar lengths of tested cable in the same geometry, which would indicate the presence of extra inductance - which could be easily calculated and subtracted. But the different values obtained with different cables seemed to show that this was not serious. But I should perhaps have indicated this shortcoming at the time, and I apologise.
The salient point in an engineering approach was, however, the finding that none of L, C and R values would make any difference under practical circumstances and in audio. (How the suspected absence of this condition keeps popping up in many of the excellent mails posted here.) Why go to further refinement if first approximations indicate that contributions would lie way up somewhere in the r.f. region, in addition to the fact that one's approximations were worst case? That would have been a worthy physics procedure, but of what value in audio?
I am not one to soap-box my long experience with audio, but while I also have analysed these parameters with a view to at least find some position in the wilderniss of qualitive statements, I would wholeheartedly agree with the several posts that put cable parameters in the bigger picture of the orders of magnitude larger contribution of loudspeaker and room characteristics.
I would also like to comment on a later post by you, but do not seem to get quotations from several pages in my replies.
Thus regards for now!
macgyver10 said:
By ringing I presume you mean a damped overshoot (repetative) at the transition of a square wave half-cycle. In that case the frequency would obviously indicate whether it was r.f., which it almost never is due to the (limited) bandwidth of audio amplifiers. However, the fact that there is an overshoot indicates that the bandwidth is not limited - let me put that better.
The bandwidth of all audio amplifiers are limited somewhere, but ringing typically indicates an increase in gain somewhere prior to roll-off, i.e. a peak. This is readily evident from a sine wave amplitude/frequency plot. It is this peak rather than the point of roll-off which causes ringing (again supposing we both mean the same thing). With reference to cables, you will not find a practical effect from them as in reflections, again because of cable parameters already discussed.
The above includes the effect of the load and amplifier stability especially with NFB. (And please, not "feedback is bad" again! With greatest respect to the Good Book: "Forgive us our platitudes, as we forgive those that cliche against us.")
Regards!
Hi panomaniac,
Hi John,
-Chris
Krell. Disconnect the protection if it has.
Hi John,
I can only imagine the fun to be had in your lab! Now, just to gauge the effect of the tester on the process, we'll run a series without beer, and another with beer. 😀
-Chris
jneutron said:
Aha! I think this might be the thingamajig that makes this happen:
http://www.linearx.com/products/accessories/Vibox/VIBOX_1.htm
Which I don't have, and at this point, can't afford.
Johan Potgieter said:
Hi Johan, how do you mean this? Does it allow for RF effects reflected back into the audio domain?
Hi John,
I had to hunt for this just now.
This is puzzling at first, but are you talking of a delay at one ear with respect to the other, of an audible tone? A hasty calculation seems to indicate that if you have a source of say 5 KHz at a distance of 20m, then an increased delay of 2 uS at one ear would be caused by moving that source about 70mm in a cross direction. (The equivalent phase shift is about 3,6 degrees.)
Some two decades ago I was involved in just such a research situation, but from memory I have a feeling that such a small shift of the source was not audible (We worked on 100m, where the equivalent shift would be 35 cm.) Still, and though this is a little off-thread, I would appreciate further info from you.
You also pose a most interesting point of whether/how one can equate this to a frequency ......
Kind regards.
I had to hunt for this just now.
jneutron said:
This is puzzling at first, but are you talking of a delay at one ear with respect to the other, of an audible tone? A hasty calculation seems to indicate that if you have a source of say 5 KHz at a distance of 20m, then an increased delay of 2 uS at one ear would be caused by moving that source about 70mm in a cross direction. (The equivalent phase shift is about 3,6 degrees.)
Some two decades ago I was involved in just such a research situation, but from memory I have a feeling that such a small shift of the source was not audible (We worked on 100m, where the equivalent shift would be 35 cm.) Still, and though this is a little off-thread, I would appreciate further info from you.
You also pose a most interesting point of whether/how one can equate this to a frequency ......
Kind regards.
Hi macgyver10
Yep, $120 US, on sale. Not bad considering the workmanship. But then again, all you need is a schematic.
Many times these things are a simple circuit and a collection of jacks. They could have asked a lot more.
I hear you on the $$$.
-Chris
Yep, $120 US, on sale. Not bad considering the workmanship. But then again, all you need is a schematic.
Many times these things are a simple circuit and a collection of jacks. They could have asked a lot more.
I hear you on the $$$.
-Chris
rdf said:
Hi Rdf,
No (must you be difficult? 🙂 🙂 )
I was simply talking of the possible audible effect of cable parameters (L, C, R), i.e. the effect on the signal supplied by the amplifier - since one often encounters quite wild claims in this regard (not referring to any contribution in this thread).
To understand you correctly, I take it you mean induced r.f. from an outside source. This could of course be a problem depending on the situation, e.g. when the system is close to a broadcasting station. To me this would be a secondary effect, but which certainly might demand e.g. a shielded speaker cable. We are also now on the terrain of how susceptable an amplifier will be to r.f. injected in the loudspeaker cable - large possible variations there!
I was fortunately not yet involved in such a situation (I do not find cell phones to interfere, say from the next room). Mercifully I am also not aware that any of my customers had such a situation - still, very possible.
Regards.
macgyver10 said:
Perhaps we can talk them into lending us (you) one? If we carefully lay out the test and publish the results of the "Group DIY" cable test, they might be willing to go along. Good publicity for them.
Johan Potgieter said:
Then you must not have NEXTEL in South Africa!
US NEXTEL users will know what I mean. Horrible things. Seem to drive everyting crazy, speakers , mics, monitors. We had to ban them on shows, they killed the wireless mics so badly. Click, click, click. Yes, even amps/speakers. I once asked a Motorola engineer how they got these things past the FCC. He just grinned...
Johan Potgieter said:
If it's any consolation my boss says the same thing. 😉
No, I was more thinking about highly reactive RF cable loads causing device instability well above the audio domain and the potential effects of low-level oscillations. To side track, a couple of nights ago I was experimenting with single tube grounded-cathode/cathode follower cascades. Changing the screen bypass cap of the pentode CF from a 3.3 uf Solen polyprop to a physically much larger 33 uf Solen caused the THD spectral analysis noise floor to go wild. It was traced to a low level oscillation in the 60 MHz range. A silver mica to ground on both the CF plate and screen cleaned it up, but more relevant to the discussion was the extent to which an RF oscillation thousands of times higher in frequency caused obvious measurable effects in the audio domain, even with a 22 kHz low pass measurement system.
Somewhere - I thought perhaps Self but can't find it on his site - someone posted a series of measurements showing cable effects on outputs in the megahertz range. Many showed low level peaking. Obviously there are too many variables in a world of gear and cable to claim a single cause but I'll put forward for consideration the possibility some reported differences are the result RF reactances causing low-level oscillations. Smoke is proof this has happened in the wild. It would seem to me high feedback, wide bandwidth designs are most at danger.
Is there a prize for the best designed test procedure?
My own experience with different speaker cables is that with the Bose 301, Normal parallen ran Monster Cable type cable, the sound was clean and slightly dead. Once I braided the same cables the way mentioned in Speaker Builder, the sound was more live and sounded like it had more depth.
With better speakers, you can hear image depth change and more awareness of where each musical instrument is located. Of course this also depends on the recorded source.
Recently I've been thinking about getting the LinkWitz Lab Demo CD to see whether it can be used as a listening reference.
My own experience with different speaker cables is that with the Bose 301, Normal parallen ran Monster Cable type cable, the sound was clean and slightly dead. Once I braided the same cables the way mentioned in Speaker Builder, the sound was more live and sounded like it had more depth.
With better speakers, you can hear image depth change and more awareness of where each musical instrument is located. Of course this also depends on the recorded source.
Recently I've been thinking about getting the LinkWitz Lab Demo CD to see whether it can be used as a listening reference.
Johan Potgieter said:
The ringing I was refering to is more symmetrical than you describe, and appears typical of what you would expect if there were not enough higher order odd harmonics to sharply define the transitions of the square wave.
More a case of "ripple" on the top and some increased slope on the rise and fall.
I realize this is not, perhaps, what was being referred to as "ringing".
However, I recall audio equipment tests in some stereo magazines publishing results of square wave reproduction on CD players. This always made me laugh, considering the brick wall filter at 1/2 nyquist. How could a CD reproduce a square wave (or any more complex repetative waveform) with any accuracy above 11Khz, for example?
Actually, CD players can't reproduce much of anything but sine waves as continuous tones above about 7 khz. I once took a scope snapshot of an Ampex 351 I modified the playback electronics of and with Saki heads at 7.5 inches per second that demonstrated a correct phase relationship between a 7.5 khz square wave fundamental and its third harmonic when recorded and played back upon that tape deck.
Not super-sure of my facts, but I believe the "Fight or Flight" mechanism was extremely primitive. It was very sensitive to the smallest sharp noise, triggering the impulse to flight before a direction of the threat was discovered. For example, the ability to discern whether a noise occurence is directly in front or behind you generally only works when you know how loud the noise already is. For hunting, L/R discrimination was very important in letting the hunter know where to look, but with few exceptions (movie samurai warriors is one) human hunter did not use their hearing to locate their prey.anatech said:
But bats...hmmm. They must have one of the most perfect phase discrimination systems in existence. I wonder what makes their hearing system different from ours?
macgyver10 said:
I think I was the one talking about "ringing." Talking about how easy it is to see in video signals. What video techs usually call "ringing" is the alternating light & dark bands or ripples that are seen to the right of an abrupt brightness transition.
My monitor here are work rings badly beacuse it goes thu a switch (3 computers, 1 monitor.) But I used to see ringing all the time when I was a video projectionist. Hard to hide it on a 20 or 30 foot screen. It looks like a sine wave that trails off. Oscillation, you know. Almost always a bad cable or connector.
Maybe this just isn't a problem at AF. Wavelengths are just too long, if it is wavelength related.
anatech said:
That was all I could afford that did not sound irritating at that time while I was young and single and in Texas.
Actually, I have an old Tiger 250 pair in the basement. I do not know what the upper freq limit is yet, but it is dc coupled. Otherwise, it will indeed be custom.panomaniac said:
Johan Potgieter said:
Numbers, sure..
For now, a simple sine signal into a single, resistive load, worst case:
4 ohms, 20 Khz sine, I = 10 amperes peak, 10 feet zip, load with zero inductance. Vpeak = 40 volts
I = 10 sin wt... w = 2 pi f, w = 2 pi 2 * 104
I = 10 sin 1.2 * 106
dI/dt = 10 w cos wt = 10 * 1.2 * 105cos wt
dI/dt = 1.2 * 106cos wt
maximum dI/dt = 1.6 amperes/microsecond.
10 feet zip = 2 uH inductance
Cable reaction:2.4 volts peak during energy storage or shedding, this is 90 degrees lagging.
Summation of this inductive voltage with the stimulus voltage shows a zero crossing shift of 500nSec, or 1/2 microsecond, with a peak shift also...(I will not detail this aspect.)
This simple analysis shows 10 feet of zip DOES NOT shift a full power 20Khz sine sufficiently for humans to detect, even if it were only one channel being shifted. (hope I did the math correctly, this html code is painful..)
So, sine excitation will not produce results which are consistent with the premise of audibility.
It is not the correct tool.
The correct tool must first be able to see this shift, this requires 100 nSec (give or take)resolution...but seeing this shift is still not proving audibility..
The correct tool will see this shift easily, and then must be able to discern this with two or three orthogonal signals superimposed on the wire, and feeding a two or three branch frequency splitting node. That is where the effort lies. What complicates the analysis is the fact that the error signal which shows up across the wire is a zero power signal, a new "class" of signal which will not show up on an FFT.
Johan Potgieter said:
The analysis I posted above arrives to the same conclusion, that is why I posted it. I also point out the reason this analysis is incomplete.
Johan Potgieter said:
We concur. I, however, point out a weakness in the standard engineering approach.
You read about the martini's, huh?anatech said:
- Status
- Not open for further replies.