Pooh,
So am I reading this correctly that you are using a 6" cone driver on a horn for the midrange? This has been my preferred method for over 30 years now and I am surprised by those who insist on trying to make a compression driver, even a 2" exit work down into the lower 150 to 200 hz range.
So am I reading this correctly that you are using a 6" cone driver on a horn for the midrange? This has been my preferred method for over 30 years now and I am surprised by those who insist on trying to make a compression driver, even a 2" exit work down into the lower 150 to 200 hz range.
Some of the treble horns are polyester DP. If designed right these over kill materials can often be match or even exceeded. Be may be hah. high modulus but many direct driver tweeters using it have good impulse but lousy step test. Too much retained energy. This may be for Be foil or PVD andthey will not be the same. Same with Ceramic magnets. It is not simply neo is better. Most of these things are swings and roundabouts. I Like the use Kinderhorn of using a direct radiator for the lower to mid idea as the impedance match at the lower frequencies is good. Not like a 25 mm tweeter. The ribbon thing is a joke because there are aluminium domes and ribbons so why is RAAL so revered. It is also how they are used. Lampizator tried the horn on P24 but he stopped. Now he is selling a $$$$ full enclosed horn speaker. so its not P24 for life.
Pano and Kinderhornman- we all have to make our own way and it is good to see that we rely on mostly on our own resources and we share some of that
Well, this is your opinion.
Anyhow, do you have any explanation for how changing the metal surface of the plug influences the sonic signature (be it by changing the distortion, or in any other way)?
I don't suppose you could back up the hundreds of posts like this with something tangible???
There's a low pressure metal to metal contact there, imperfect by a very large measure. Depending upon precisely what the metals are, the pressures, how smooth the surfaces are, how long the prongs have been undisturbed there will be a contact resistance which will be not a simple resistive quantity, it will be non-linear in nature. Which will impact upon the harmonic content of noise riding on the mains line which passes through to the component power supply, and the precise shape of current pulses drawn.
I've done the exercise of eliminating that "weakness", as an exercise a long time ago, by hard wiring the mains directly to the power cord - naughty, naughty! - but it proved the point, the sound significantly improved. The best approach is to consider the power being fed down the power cord to your gear as being severely contaminated by everything it's seen up to that point, so clean it up by whatever means seem reasonable, and effective.
"Tangible" is hard, because it's mostly about the hearing at the moment - measurements currently struggle to get a handle on the matter. People like John Curl who have access to plenty of expensive, precision equipment haven't got easy answers, and the effort isn't being put into finding smarter testing techniques.
Probably the best approach is to record at the highest quality level what emerges from the speaker drivers - because this is a total system behaviour - and then do sophisticated analysis to differentiate that from the source data. Things like DiffMaker are essentially useless in dealing with the realities of playback - something much, much, more capable is necessary ...
I suspect he may be more interested in your solutions to these problems. What have you done that has eliminated or mitigated them?
So I take it we are never going to hear your secret recipe for audio nirvana. Glad you've reached it.
Frank,
I don't think I would recommend that people start hard wiring the mains connections, you are starting down a path that most would never attempt. If you are going to go to that extreme then why now twist the mains connection wires from the box to the outlet? It sure wouldn't be easy with solid core wire and I am not sure that you can use stranded wire on a mains connector and meet most building codes? It would be just about impossible in a new home in the USA as they have pretty much stopped using conduit for the wires and they are now all using Romex wiring now with parallel wires running the entire length of the conductors.
I don't think I would recommend that people start hard wiring the mains connections, you are starting down a path that most would never attempt. If you are going to go to that extreme then why now twist the mains connection wires from the box to the outlet? It sure wouldn't be easy with solid core wire and I am not sure that you can use stranded wire on a mains connector and meet most building codes? It would be just about impossible in a new home in the USA as they have pretty much stopped using conduit for the wires and they are now all using Romex wiring now with parallel wires running the entire length of the conductors.
My experience is that 2nd harmonic has very low audibility ... my perceptual threshold is around 1~3%, which is consistent with the old literature from the Forties. 0.5% 3rd harmonic, by itself, adds a noticeable roughness to the sound, tilting the subjective balance towards treble. The combination of 1% 2nd harmonic and 0.25% 3rd harmonic restores the tonal balance, and thickens up the sound a little.
The higher harmonics are another story. The 5th, 6th, 7th, and 9th are extremely dissonant. Back in the Fifties, both D.E.L. Shorter of the BBC and the popular audio writer Crowhurst suggested weighting harmonics by either the square or cube of the order, suggestions that were not taken up due to the computational limitations of the day. Instead we got the simple THD number, which in most equipment is dominated by 2nd and 3rd harmonics, which are also the least audible.
The same problem applies to electronics and loudspeaker drivers; the harmonics that dominate the measurement are not the most audible. In the case of 2nd harmonic, audibility is quite low, surprisingly so.
So the real question with vacuum-tube electronics is the spectra of the distortion, not the magnitudes, unless the numbers are so high that a defect is indicated. This is where tubes differ; some have a very fast decline in upper harmonics, while others create a comb spectra, with only a very slow decline. In practice, the comb-spectra tubes may have intentionally uneven windings on the grids, so they can be used as limiters in IF strip in a tuner, or as an AGC detector.
Low-level nonlinearities always have to be carefully examined and removed if possible. They generally fall outside the scope of modeling and simulation programs because real-world electronic devices always have more complex behavior, due to manufacturing defects, than the simple nonlinear models used in sims.
In loudspeakers, the field geometry in and out of the gap has an effect on the linearity of the driver. A loudspeaker is basically a linear motor, and can never be more linear than the drive field applied to the voice coil. If the field lines are bunched or more widely spaced in the region traversed by the voice coil, that affects the drive.
Overhung and underhung voice coils are very different. Unity-hung VCs are used for guitar speakers, where nonlinearity is desirable, and the tone sharply changes depending on drive level; it makes the loudspeaker more responsive to the player's touch. In hifi, we don't want this, so overhung (most common) and underhung are preferred.
Overhung has been used for 95~99% of hifi drivers since the Sixties. It delivers the longest excursion (deepest bass) of all topologies, and has the additional advantage of most efficient use of a given magnet size.
Underhung goes back to the Thirties and Forties; the VC is shorter than the gap, not longer. Why was it ever used? Well, the field lines in the gap are straight, unlike the field lines outside the gap, so operation is linear ... as long as the VC does not leave the gap.
Overhung makes a different compromise. Since a substantial portion of the VC is always outside the gap, the "outside" portion is exposed to the stray field outside the gap. The better grade of drivers use field-shaping techniques to linearize this out-of-gap field, but it has to be kept in mind the external field lines are never as straight as the ones in the gap itself.
It's similar to the tradeoff between Class A and Class AB operation for opamps and power amplifiers. Class A operation is the most linear at low levels, but for the same-sized device, maximum power is reduced by 4 to 10 times compared to the more efficient Class AB. That's an unattractive and uneconomical tradeoff for most manufacturers and buyers.
Similarly, underhung drivers have shorter usable VC travel compared to a same-efficiency overhung driver, but within that short travel, the VC is exposed to straighter field lines than the overhung alternative.
The "secret recipe" is easy to say - it's being fussy, considering the impact of everything, taking nothing for granted, assuming nothing. Actually following through on that, and doing the "testing" that establishes what matters, and what doesn't, takes time, focus and continual playing with things. In essence, it's treating getting optimum playback as a troubleshooting exercise, rather than a mix 'n' match process, going on forever.
What needs to be done for a particular system will vary from the next one. There's no "secret recipe" for fixing a car that's not quite running properly, every car's issues will differ from the other. But it's persistence, and a history of experience by the mechanic that gets the situation under control.
Or how about something that does not contradict all available data and theory.
Why do you keep posting this mantra?? We've seen it repeated hundreds of times, and it gets us absolutely no where. I've said all I'm going to say about this to let the moderators rest.
Not recommending it at all. This was 30 years ago, many, many houses ago, when I just wanted to know what mattered, what made a difference - haven't done that since. Of course, as I pointed out, the real solution is to treat power out of the wall as "dirty", so the job is to clean it up - or, make the audio equipment impervious to any sort of interference, from any cause. Over the years I've tried many ideas for doing this, all of which had some effect at the time - my 'ways' keep evolving.
The whole process of getting the good sound is like being a detective: maybe it's this, maybe that, that's stopping the sound "working" well - trial and error is the way I've used to get my solutions ...
Last edited:
Because it happens to work ... I got a very good result, 30 years ago, purely by accident, and have evolved the process over the years. On the other hand, I hear plenty of audio systems, built the "correct way" and that results in "dodgy" replay - the recent hifi show in Sydney stank of this lesser quality sound, so the 'pure' way is not working, from where I happen to be looking at it ...
It was the square, which is a big step in the right direction. But, it still does not account for the difference in nonlinearities that grow with level (clipping) and ones that grow as the level is reduced (crossover). For any single metric to work this level effect has to also be accounted for. Looking at the square of the harmonic levels versus signal level will pretty much tell the whole story - but who does that! (I have.)