Regarding loudspeaker frequency response linearity, I believe that this is one of the reasons why my speakers sound so linear. They were made to work 40-18.000 Hz +/- 1.5 dB and are matched to within 0.5 dB one from another.
Their impedance modulus, nominally at 8 Ohms, ranged from a minimum of 6.5 Ohms to just over 12 Ohms at 20 kHz.
How does this tie to the uncommonly flat response?
I should mention that all of these were measured with an input of 2.83 V, at which time it produced 92.5 dB at 1 kHz, taken as a reference point/value.
Their impedance modulus, nominally at 8 Ohms, ranged from a minimum of 6.5 Ohms to just over 12 Ohms at 20 kHz.
How does this tie to the uncommonly flat response?
I should mention that all of these were measured with an input of 2.83 V, at which time it produced 92.5 dB at 1 kHz, taken as a reference point/value.
We simply have to use our critical thinking skills.
If the conclusions of the reviewers come from non scientific methods, then we must put less weight upon them. It could be that other factors are influencing the reviewers, for instance price, looks, past reviews etc...
This is technically true- if you're doing a side by side comparison with rapid switching, 0.1dB of level is a just-noticeable-difference for sharp ears. It is JND for some frequency response differences. But in a non-instantaneous comparison or playing a dynamically varying piece of music, I suspect that this would not be an audible issue- as well, John is absolutely right when he points out that most speaker wire runs will swamp the amp's output impedance if it's a low one (as in most engineered amplifiers).
If there's data showing audibility of the effect of varying speaker impedance with level due to amplifier source impedance, I'd be interested to know about it.
Pavel, behave. That’s misotuby(and Zout can be adjusted via NFB, tubes and SS alike).
What’s wrong with you?
George
George, would you mind posting a single link as a proof. Full set of measurements of a tube power amplifier, preferably from Stereophile, or a similar complete set. Thank you
All the best,
Pavel
George, would you mind posting a single link as a proof. Full set of measurements of a tube power amplifier, preferably from Stereophile, or a similar complete set. Thank you
All the best,
Pavel
Pavel
If Stereophile had measured a million+n tube amplifiers and their valid and respected measurements had shown that all these amplifiers were of considerable distortion (of any kind) figures and of highish Zout, would this make for a positive rational conclusion that tubes by their nature distort and that Zout of a tube amp circuit is not a function of the amount of NFB?
George
In principle you are right but the available loop gain in the overwhelming majority of commercial tube amps is too low to have any appreciable loop feedback. This coupled with an inherently larger Zout (open loop) than most SS, makes a clear difference.
Add the average tube designers' aversion to feedback and there's your recipe.
As SY noted, some designers are able to create tube amps that are really 'high fidelity' in the original sense, like Morgan Jones' creations and Bob Cordell's tube amp from Linear Audio, and these have low Zout as any respectable amp should have.
But those are the exceptions that confirm the rule.
Jan
That's why feedback exists. Just like with those highly nonlinear p-n junctions in solid state amp output stages which are linearized by various feedback schemes.
The problem with commercial designs is that the customers for those amps WANT them to sound different, to alter the signal, to be an effects box. Like the guys who thought my designs sounded like high feedback solid state amps.
That is irrelevant to the technology, it's a market question.
Pavel
I decided to commend only because I see you insist on blaming the tubes and not the many or all Hi-End amplifiers built with tubes.
You know that tubes by their nature are high R devices, that they can be microphonic but I trust you know also that their principle of operation allows them to function as more linear amplifying elements than SS amplifying elements ( before applying any feedback around them, provided proper DC bias and load line)
George
I decided to commend only because I see you insist on blaming the tubes and not the many or all Hi-End amplifiers built with tubes.
You know that tubes by their nature are high R devices, that they can be microphonic but I trust you know also that their principle of operation allows them to function as more linear amplifying elements than SS amplifying elements ( before applying any feedback around them, provided proper DC bias and load line)
George
Yes, it is very hard to achieve. One of the reasons we did it was because we had around 40 tweeters to choose from, ditto for mid range drivers and bass units. Not much of a population, but the Son Audax had rather stringent manufacturing standards, so that did help.
AS for levels, I said that was at 2.84Vrms input, coresponding to 1W/8 Ohms, at which time the speaker produced 92.5 dB SPL at 1 m and at 1 kHz, which we took to be our zero point of reference. Therefore, our overall level 40-18.000 Hz varied from 91 to 94 dB.
Using the traditional +/- 3 dB variation, the response would have been 36-22.200 Hz, rapidly falling beyond those points. Phase shift is even better, that's the spec I am really proud of, worst case is -25 degrees, just above the bass driver resosnance point.
We tried thee bass reflex port tuning points: 33, 36 and 39 Hz. 33 Hz would havegiven me a Hz lower down repsonse, but at the cost of some unruly bass driver behavior. In the end, we judged that 36 Hz was the best solution overall. BTW, the bass driver is a 10" unit, at the time the best Son Audax made, and the enclosed volume is just above 70 litres. Obviously, the mid and high range drivers are in separate enclosures.
We really went to town on that speaker, but it took us 6 months to get it just right.
I don't remember the exact prices of the drivers, that was 12 years ago, and the prices I got were distributor prices (the friend I did it with was the offical Son Audax distributor for Serbia and Montenegro), but eve so, the RRP for the woofer was about €330 oer unit, midranges were around €120 per unit and the tweeter I remember as costing €97 per unit. So, just under €1.100 for the drivers alone. Corrected for 3% annual inflation, that's like €1.600 today.
But believe me, it was well worth it.
Last edited:
While I agree with you completely that tubes are elements which can be connected in many ways, thus determinig their sound, I must also point out that in the local tube community at least here in Serbia, tubes are thoughtof as magical ONLY when operating in zero global NFB mode only.
Once they have a global NFB loop added, no matter how small or large, they are said to lose their acoustic charm and to sound more like solid state.
My personal experience with this is exactly zero, so I have no personal view on that matter.
Last edited:
Using your critical thinking skills, how do you really explain the sound description all being nearly the same on any particular speaker system? That those descriptions come from different times and places and spaces. Always a coincidence or always because reviewers and listeners are impressed by price and looks etc? [Note this occures with middle price and low price speakers as well.] Well, I dont think so! The consumer would know if it didnt sound the way it was reviewed. I hear the speakers pretty much as the better writer/reviewers describe them.... in my own system and space.
I think your arguments are a stretch to avoid some unpleasant truths.
THx-RNMarsh
- Status
- Not open for further replies.