The lumped sum parameter equivalent is only an approximation. It pays no heed to the line length, prop velocity, or impedance mismatch at the load. While the equation is well understood, it's application to audio loads and frequencies is terribly mis-applied and mis-understood.
I concur, Belden is the best. There is a lot however, that they do not know about wire. Try buying 75 thousand feet of low smoke, zero halogen, radiation resistant cable tray rated 2/0 single conductor...
And their papers on t-line theory use approximations..Still, they are the best out there.
Or, possibly, problems that have not to date been discernable using standardized blind testing. I remain open to that possibility.
Actually, they do for first order numbers. Specifically, the capacitance and cable impedance are. But second order effects such as absorbtion, no.
I do wish however, that they'd include inductance per foot for all their cable. They've been tending to include cable Z and cap per foot, and we have to re-calc to get the inductance.
Actually, "poorly designed amplifier" requires some slight re-statement. If the amplifier is hot enough that tweeter unloading causes sufficient phase shift before unity gain, is it the designer's fault? The designer may have had speed/bw requirements that did not include stability in all regimes.. I would blame the one responsible for the initial design requirements.
Note I did not say excessive capacitance. It is because the tweeter impedance rises sufficiently well above cable z before unity gain. Cable capacitance is not the real culprit, but the load mismatch is.
Cheers, John
Last edited:
I am back after a short break, I see most of you seem to never take a break
Pavel,
If I measure the Cepstrum of a directly heated cathode vacuum tube amplifier without global feedback and a modern well designed solid state unit, do you think I would find time domain information that might show a difference?
ES
I take this as the high-end has given up trying to trickle down any of their ideas into more affordable mainstream products and has decided to serve their tiny market of followers with ever more ridiculously priced, as Kantor says, fetish objects.
Actually, the statement is only partially correct. The distinction is meaningless to the general public, as they normally have no control over the load characteristics at high frequencies. To people who do e/m for a living, the distinction is huge.
So..
Linn amplifiers will oscillate when driving a capacitive cable that is terminated in a load which decouples at higher frequencies.
A linn amplifier will NOT oscillate when it drives the exact same cable when that cable is terminated in a load which matches the cable impedance up to the unity gain point of the system.
Cheers, John
ps.. holy mackeral, that orange was horrible..red not much better...blue it is..
Last edited:
The same might be said for speakers with regions of pathologically low impedance, the requirements for .1dB error from cable resistance might come into play.
Agreed. Iff .1dB is of significance. Otherwise, it's just a number. When the termination impedance falls below line z, the system starts to look inductive. While I do have cases where inductance can cause instability, it's pretty much limited to load inductances outside consumer interest, anywhere from 5 to 70 henries.
What I am speaking of with hf instability however, is based in T-line theory. I do not have sufficient understanding of electronics to even think about a resistive divider...perhaps next semester??
ps..happy belated bday, you old guy...
Cheers, John
Yeah? Was rather more Jack Bennyish I thought.
I was just trying to figure out what John's point was as I wasn't really able to make one out from his previous posts.
He said something about getting better "fidelity" and then went on about attending audio shows. Particularly THIS year's CES/THE Show.
I still don't see what his point was, if any.
se
What exactly is "true sonic satisfaction"? As opposed to what? Untrue sonic satisfaction? Are you saying that of those who are satisfied sonically, not everyone is truly satisfied sonically? Truly by whose measure exactly?
To some, yes. To others, no.
Who's right?
So it is a numbers game, just that it's about numbers we can't yet measure?
But not without the benefit of knowing what you're listening to, i.e. you have to peek.
So you don't REALLY know. You just believe.
se
When it comes to IC op amps, I prefer video designs with HIGH open loop bandwidth. I also prefer IC's with class A output stages. These are aspects of what I always design into my discrete op amp designs. I also can use jfets for both the first and second stages without much trouble. This always gives a slightly smoother distortion distribution, usually without 5th or 7th easily seen or measured in any significant proportion to the rest of the distortion residual.
How much time information do you expect to see from a solid state amplifier? What would cause it?
Some measurements require skill to understand properly, others slap you in the face with it.
What kind of time information might you see from the direct heated cathode no global feedback vacuum tube amplifier? What would cause this that is not in a solid state amplifier?
ES
Scott, I see you're you're all prepped for the super bowl.
Or are you just communicating the deeper meaning here?
I figured I'd spend the next few days reminicing about 1967.
That saying really was over the locker room door at Lambeau Field. When they asked Fuzzy Thurston how he prepared for the 13 below game he said he drank 10 Vodka's, times have changed. The avatar shows some lack of mouthguards back then, who is it? SY be quiet.
Not sure ED, but once you determine the time issues how do you filter out the 5-15% distortion of the SET amp?
QUOTE=jneutron;2456137
"The lumped sum parameter equivalent is only an approximation."
All mathematical models are only approximations. What else is new?
"It pays no heed to the line length"
Actually it does. The value of each parameter depends on length. This is why the specifications are expressed as resistance, conductance, capacitance, inductance, susceptance, admittance per foot or per meter. When you solve the equation for a specific case, you fill in the blanks.
"prop velocity"
That is a physical characteristic, not a circuit parameter unless it can be expressed as a qualifier to a varaiable that is a circuit parameter such as capacitance.
"or impedance mismatch at the load."
Not its function, it describes only the wire itself. It's up to the user to put it in context of the network.
"While the equation is well understood, it's application to audio loads and frequencies is terribly mis-applied and mis-understood."
That's not a shortcoming of the equation. It's like any other tool. If you don't know which end of the soldering iron is the handle, don't blame the manufacturer for turning it on and getting burned when you grab the wrong end.
"I concur, Belden is the best. There is a lot however, that they do not know about wire. Try buying 75 thousand feet of low smoke, zero halogen, radiation resistant cable tray rated 2/0 single conductor..."
Actually I tried buying 561,000 feet of plenum rated RG62 and 100,000 feet of plenum rated 27 twisted pairs in 1700 foot put-ups from them in 1984 They couldn't meet my delivery schedule and they wouldn't guarantee put-ups over 800 feet. The order went to the Habia Cable Company in Ronkonkoma LI. It was about $350,000. The quality was outstanding and delivery was on time.
The largest cable I've bought was about 750 MCM, the largest underground was around 500 MCM rated for 15 KVA use in an underground duct bank. I think it had to be meggared at 90KV. Also 3300 twisted pairs 3.3 inches in diameter.
"And their papers on t-line theory use approximations..Still, they are the best out there."
All papers use approximations whether they are right, wrong or otherwise. There are no exact models.
"Or, possibly, problems that have not to date been discernable using standardized blind testing. I remain open to that possibility."
Interesting program on Nova Science last night on PBS about how the brain works and can be tricked. Interesting demo about perception of sound. Also a lot of magicians tricks showing how the brain is distracted from noticing what a magician does to create illusions. Nothing on false advertising though.
"Actually, they do for first order numbers. Specifically, the capacitance and cable impedance are. But second order effects such as absorbtion, no."
Absorption is not an element in circuit analysis. It should not appear in the telegrapher's equation. It may be used as a physical parameter in another equation to calculate parameters used in that equation though.
"I do wish however, that they'd include inductance per foot for all their cable. They've been tending to include cable Z and cap per foot, and we have to re-calc to get the inductance."
You can always measure it in your lab using impedence using lissajous pattern theory among other techniques.
"Actually, "poorly designed amplifier" requires some slight re-statement. If the amplifier is hot enough that tweeter unloading causes sufficient phase shift before unity gain, is it the designer's fault?"
Yes. Stability under varying loads including step changes in magnitude and phase of load are a criteria for electrical design of a good audio amplifier. A lot of what has appeared on the market seems to have been designed by tinkerers. Not all audio amplifiers were designed by bona fide electrical engineers.
"The designer may have had speed/bw requirements that did not include stability in all regimes.. I would blame the one responsible for the initial design requirements."
If the electrical engineer didn't point that out as a requirement to whomever asked for the design, he still didn't do his job. Often managers will ask for things that are incompletely specified, incoherent, impossible to achieve, or otherwise flawed. It's up to the engineer to explain in laymen's terms to him what they are and what they mean. If someone asks for a car design that has no wheels, the engineer has no valid excuse building oone when whoever asked for it finds out it won't go anywhere.
"Note I did not say excessive capacitance. It is because the tweeter impedance rises sufficiently well above cable z before unity gain. Cable capacitance is not the real culprit, but the load mismatch is."
When the cable design is as badly defective as some so called audiophile cables are, the cable itself becomes an element of the load. The entire idea of setting up standards for different classes of wire for different wire applications by the mainstream wire industry was to eliminate wire as a significant factor in electrial performance of systems as much as possible. The selection of wire as a control element is diametrically opposed to that goal. It flies in the face of creating products that are inexpensive, effective, and appropriate so that the engineer doesn't have to think about that as a variable critical element too. Of course this doesn't generate profits for people who make and sell exoitic wire. For them it's best to ignore that and stick with magic.
"The lumped sum parameter equivalent is only an approximation."
All mathematical models are only approximations. What else is new?
"It pays no heed to the line length"
Actually it does. The value of each parameter depends on length. This is why the specifications are expressed as resistance, conductance, capacitance, inductance, susceptance, admittance per foot or per meter. When you solve the equation for a specific case, you fill in the blanks.
"prop velocity"
That is a physical characteristic, not a circuit parameter unless it can be expressed as a qualifier to a varaiable that is a circuit parameter such as capacitance.
"or impedance mismatch at the load."
Not its function, it describes only the wire itself. It's up to the user to put it in context of the network.
"While the equation is well understood, it's application to audio loads and frequencies is terribly mis-applied and mis-understood."
That's not a shortcoming of the equation. It's like any other tool. If you don't know which end of the soldering iron is the handle, don't blame the manufacturer for turning it on and getting burned when you grab the wrong end.
"I concur, Belden is the best. There is a lot however, that they do not know about wire. Try buying 75 thousand feet of low smoke, zero halogen, radiation resistant cable tray rated 2/0 single conductor..."
Actually I tried buying 561,000 feet of plenum rated RG62 and 100,000 feet of plenum rated 27 twisted pairs in 1700 foot put-ups from them in 1984 They couldn't meet my delivery schedule and they wouldn't guarantee put-ups over 800 feet. The order went to the Habia Cable Company in Ronkonkoma LI. It was about $350,000. The quality was outstanding and delivery was on time.
The largest cable I've bought was about 750 MCM, the largest underground was around 500 MCM rated for 15 KVA use in an underground duct bank. I think it had to be meggared at 90KV. Also 3300 twisted pairs 3.3 inches in diameter.
"And their papers on t-line theory use approximations..Still, they are the best out there."
All papers use approximations whether they are right, wrong or otherwise. There are no exact models.
"Or, possibly, problems that have not to date been discernable using standardized blind testing. I remain open to that possibility."
Interesting program on Nova Science last night on PBS about how the brain works and can be tricked. Interesting demo about perception of sound. Also a lot of magicians tricks showing how the brain is distracted from noticing what a magician does to create illusions. Nothing on false advertising though.
"Actually, they do for first order numbers. Specifically, the capacitance and cable impedance are. But second order effects such as absorbtion, no."
Absorption is not an element in circuit analysis. It should not appear in the telegrapher's equation. It may be used as a physical parameter in another equation to calculate parameters used in that equation though.
"I do wish however, that they'd include inductance per foot for all their cable. They've been tending to include cable Z and cap per foot, and we have to re-calc to get the inductance."
You can always measure it in your lab using impedence using lissajous pattern theory among other techniques.
"Actually, "poorly designed amplifier" requires some slight re-statement. If the amplifier is hot enough that tweeter unloading causes sufficient phase shift before unity gain, is it the designer's fault?"
Yes. Stability under varying loads including step changes in magnitude and phase of load are a criteria for electrical design of a good audio amplifier. A lot of what has appeared on the market seems to have been designed by tinkerers. Not all audio amplifiers were designed by bona fide electrical engineers.
"The designer may have had speed/bw requirements that did not include stability in all regimes.. I would blame the one responsible for the initial design requirements."
If the electrical engineer didn't point that out as a requirement to whomever asked for the design, he still didn't do his job. Often managers will ask for things that are incompletely specified, incoherent, impossible to achieve, or otherwise flawed. It's up to the engineer to explain in laymen's terms to him what they are and what they mean. If someone asks for a car design that has no wheels, the engineer has no valid excuse building oone when whoever asked for it finds out it won't go anywhere.
"Note I did not say excessive capacitance. It is because the tweeter impedance rises sufficiently well above cable z before unity gain. Cable capacitance is not the real culprit, but the load mismatch is."
When the cable design is as badly defective as some so called audiophile cables are, the cable itself becomes an element of the load. The entire idea of setting up standards for different classes of wire for different wire applications by the mainstream wire industry was to eliminate wire as a significant factor in electrial performance of systems as much as possible. The selection of wire as a control element is diametrically opposed to that goal. It flies in the face of creating products that are inexpensive, effective, and appropriate so that the engineer doesn't have to think about that as a variable critical element too. Of course this doesn't generate profits for people who make and sell exoitic wire. For them it's best to ignore that and stick with magic.
New? Interesting...
You stated: quote"The mathematical model for cables of all types is one of the most well understood and highest correlation formula when compared to testing data we have. I'm referrig to the "Telegrapher's Equation." It gives the lump sum parameter equivalent of the distributed parameter network of a cable. "
You are incorrect. From your response, you do not understand it nor it's implications with respect to speaker loads, line length, and amplifier stability.
Perhaps I should update the wiki page to reflect reality?
Again...
Lumped element models do not consider the relationship between the load impedance and the cable impedance. Nor, do they correctly consider the physical length of the speaker wire connected to a load which does not match the wire impedance.
If you wish to learn this stuff, give me a PM..I don't think this thread is the correct place to teach you.
Your mis-understanding is btw, that of the majority. Not your fault..
I said:
"Or, possibly, problems that have not to date been discernable using standardized blind testing. I remain open to that possibility."
you responded:
Interesting to know, but has nothing to do with what I said.
I said:
"Actually, they do for first order numbers. Specifically, the capacitance and cable impedance are. But second order effects such as absorbtion, no."
you responded:
Never made a high accuracy S/H eh?
Or hi-potted a large capacitive structure only to find that simply shorting the terminals momentarily doesn't prevent recovery?
What??
Lissajous pattern theory???? Like, in ....time tunnel?
I have a novel idea..how about an inductance meter..
Or use the equation LC = 1034 DC, L in nH per foot, C in pf per foot.
No. and ...No.
A low Z cable has to be considered independently from the load. That is where guys like you need to learn reality. Googling this stuff ain't gonna find what I speak of...I haven't put it on wiki..
PM me if you wish.. We can even catch up on the years since AR..
Cheers, John
ps..it would help if you address the entity without adding lots of off-topic stuff, your posts remind me of JR's diatribes. If you can do it short and concise, other's won't have to piece together a coherent post by wading through lots of off topic stuff.
Last edited:
Pavel,
Will get there, it is interesting. But there is no single answer as vacuum tubes even with the same type numbers behave differently in the time domain.
Scott,
You know perfectly well the normally considered nonlinear distortions don't hang around after the excitation signal.
Which sounds worse 5% 2nd order thd or 4,000hz hanging around for 10ms. (to -60db)?
Which sounds better .001% 3rd harmonic or 350hz reverbing for 125ms?
Just for grins anyone can try a bit of frequency selective reverb of very short duration. It can go from increased warmth to nasty sibilance. Nothing new here, acousticians have know this for decades even if they are still refining their criteria now that better instruments exist.
Next trick is to supply the filament to the tube under test with an audio power amplifier set to provide the proper DC voltage with a small sweepable AC component. Guess which tubes sound better the ones with the least filament sing, the ones with resonances in the lows or mids, or the ones that hid upper mids to highs?
Oh yeah what is soundstage?
ES
- Status
- Not open for further replies.