What specs exceed 20kHz?
Spec is short for specification, which comes from the word specific.
So far you haven't even quantified what it is you're on about so how can you have any sort of spec for it?
All you've been doing is throwing a bunch of stuff up against the wall and hoping something eventually sticks.
se
Well, presumably, you'd be using a source Z that wasn't 600R, and you'd be using a Cinemag MC stepup rather than the line transformer that Steve cited in a totally different context; the MC transformers are more like -1dB at 100kHz. As a side note, I've used that particular Cinemag line transformer and had no increase in HF noise (see spectrum below).
By "flat to 40kHz," you mean that you're not implementing the so-called Neumann 3.18us time constant? I personally can't hear the difference, but some (notably the late Allen Wright) claim that the 3.18 zero makes a major audible improvement.
Attachments
JC, it seems that standing by these specs above for your new MC preamplifier, you are about to attenuate the increased HF output of the MC cartridges by other means, after the input transformer.
The –7db at 100KHz of the mentioned Cinemag , might be a bonus in this case (and act as an RF-pollution filter as well). I mean that the Cinemag people, may have LP filtered their product for a reason.
http://www.diyaudio.com/forums/analog-line-level/146693-john-curls-blowtorch-preamplifier-part-ii-732.html#post2396527:
This isn't engineering, it's insanity. This is the same advertising hype that came straight out of the 1960s. University Sphericon tweeter, response to 40 khz. HK A1000 amplifier, flat from DC to light. Audio Fidelity's Frey Stereophonic Curtain of Sound with response extended to 25 KHZ. This kind of talk is a complete rejection of the mathematical principles of Fourier which leaves no room for doubt that response beyond 20 khz serves no useful purpose. But it does have a down side. It not only makes the system susceptable to noise from stray RF and magnetic fields which can interact with the signal at lower frequencies, it wastes power (signal handling capability) at best and at worst can result in spontaneous oscillation that can damage amplifiers and tweeters. A bad choice.
No, that is the whole point I am trying to make. The mechanism exists anyway. The random thermal fluctuations in the core sum do not sum to zero, but sum to a thermal energy - just like a resistor. They generate a field which induces a noise voltage in the windings.
Transformers with identical windings, but different degrees of coupling to eddy currents in the core, will give different thermal voltages. You can calculate the thermal voltage by first measuring the resistance at a particular frequency - this will be greater than the DC resistance. The increase may come from skin effect, proximity effect, core coupling (the mechanism for eddy currents), radiation resistance etc. etc. Which ones dominate will depend on frequency. The appropriate temperature to use for the calculation may also depend on frequency, but is likely to be transformer temperature for all except radiation resistance. Sorry to drag us back to this point, but I think this is still not understood by some in this debate.
To try to settle another point, current in an ordinary wire does not generate shot noise. How do we know this? Well, for a start how could we measure shot noise in a valve (which is much smaller than sqrt(2ei)) if the wires attached to the valve contributed even more shot noise too? No, in a wire the electrons can 'see' each other so their flow is correlated - they behave in a similar way to a classical gas. If an electron is 'here', then other electrons are repelled so less likely to be 'here'. As a result they tend to arrange themselves in a fairly uniform way so there is no shot noise. If wires gave shot noise, then any junction would also generate partition noise - all our electronics would get a lot more complicated!
I now want to toss in a new thought. Is it possible that the insulation between laminations may permit some electron tunneling - it is very thin. If so, this may be a source of shot noise which would be signal dependent. I have no idea what the magnitude might be, so this could be a red herring. It would get smaller with thinner laminations, provided that the insulation did not get much thinner too. This is because there would be more insulation layers in series between any two points, so presumably less current.
When it comes to designing, I follow my own design rules, not someone else's.
The basic design, that will be used the VAST MAJORITY of the time, will be flat, to better than +/- .1 dB from 20Hz to 20KHz. I have ALWAYS maintained that standard, for the last 30 years. To add the CineMag transformer, that was shown by SE, would be down almost .3dB at 20KHz. This would spoil all the work that I did to get the RIAA flat in the first place.
I just injected that comment to remind people that some transformers are more 'equal' than others. It was a pleasure to poke fun at SE's transformer.
A curve like this, would certainly put me off, and I would chose something else.
The basic design, that will be used the VAST MAJORITY of the time, will be flat, to better than +/- .1 dB from 20Hz to 20KHz. I have ALWAYS maintained that standard, for the last 30 years. To add the CineMag transformer, that was shown by SE, would be down almost .3dB at 20KHz. This would spoil all the work that I did to get the RIAA flat in the first place.
I just injected that comment to remind people that some transformers are more 'equal' than others. It was a pleasure to poke fun at SE's transformer.
A curve like this, would certainly put me off, and I would chose something else.
What SY is referring to is exactly opposite what Soundminded is 'ranting' about. I think that SY has a better case, although I will probably not go in that direction.
We are talking about adding a 'shelf' at some high frequency, usually at 40kHz-50kHz to the RIAA to compensate for the 'shelf' in the pre-emphasis built into the disc cutter, which is also between 40KHz and 50KHz. Makes sense, or does it?
From Soundmind's opinion, ANYTHING over 20KHz doesn't make any sense.
But what of more open minded audiophiles, who want TRUE reproduction?
Well, add this 'shelf' and shoot yourself in the foot! (more later)
We are talking about adding a 'shelf' at some high frequency, usually at 40kHz-50kHz to the RIAA to compensate for the 'shelf' in the pre-emphasis built into the disc cutter, which is also between 40KHz and 50KHz. Makes sense, or does it?
From Soundmind's opinion, ANYTHING over 20KHz doesn't make any sense.
But what of more open minded audiophiles, who want TRUE reproduction?
Well, add this 'shelf' and shoot yourself in the foot! (more later)
I was thinking when I listened to NPR last night that you wouldn't want to lose 0.2 db of distortion at 20khz with an SPL of 130 db because that wouldn't be right at all.
Soundcheck: - WNYC
http://www.wnyc.org/shows/soundcheck/2011/mar/11/
50 Years of Making Fuzz, the Sound That Defines Rock 'n' Roll - William Weir - Technology - The Atlantic
The program discusses how an accidental wiring error in a sound mixing board resulted in the invention of distortion that defines rock and roll. The guys who invented the fuzz box and other great discoveries like the big muff explain how what seemed like a horrible mistake turned into a boon for them. Now who would want to distort distortion? I mean really, where would the Grateful dead or Jimi Hendrix have gotten without it? Why they might just have become ordinary office clerks wasting away their lives without purpose and have been totally forgotten.
BTW, the golden element they all love.....germanium transistors. Ever think of using them in your designs JC? Who knows, it might start a whole new trend.
Soundcheck: - WNYC
http://www.wnyc.org/shows/soundcheck/2011/mar/11/
50 Years of Making Fuzz, the Sound That Defines Rock 'n' Roll - William Weir - Technology - The Atlantic
The program discusses how an accidental wiring error in a sound mixing board resulted in the invention of distortion that defines rock and roll. The guys who invented the fuzz box and other great discoveries like the big muff explain how what seemed like a horrible mistake turned into a boon for them. Now who would want to distort distortion? I mean really, where would the Grateful dead or Jimi Hendrix have gotten without it? Why they might just have become ordinary office clerks wasting away their lives without purpose and have been totally forgotten.
BTW, the golden element they all love.....germanium transistors. Ever think of using them in your designs JC? Who knows, it might start a whole new trend.
Last edited:
Soundminded, Germanium, YES I would if I could, but Kirkwood Rough, or Roadbagger has cornered the market. I visited him the other day and he had LARGE plastic bags full of Germanium transistors that he uses for musical instrument amplifiers. Ask him yourself.
He and I are both REAL AUDIO design engineers, AND we love Germanium! Heck, OUR QuanTech's are full of them! Praise to the early engineers who could make them work.
He and I are both REAL AUDIO design engineers, AND we love Germanium! Heck, OUR QuanTech's are full of them! Praise to the early engineers who could make them work.
Now, SY, getting back to your point. Why NOT put in a 50KHz shelf to compensate for the lack of boost (3dB) at 50KHz. Why roll off, forever?
Well, unfortunately, the disc cutter designers, like Ortofon and most probably Neumann knew they had a problem, SO THEY FIXED IT, by resonating the cutting head to boost 1 dB or so at 20KHz TO MEET THEIR SPEC. SHEET. Therefore, IF you try to compensate and add a 50K shelf, you create a BOOST at 20K approaching 1 dB. Not so good, unless you use lossy high frequency transformers to compensate. Didn't know that, did you? Of course, the 50KHz resonance of a typical MC phono cartridge will easily fill in any losses of output in that region created by the LACK of a shelf, and after 50KHz, even I don't even care much, because that is the LIMIT of the best practical microphones (low noise is important too), master tape recorders, my UltraMaster 30 ips tape recorder is flat to approximately 45KHz, and 1/2 speed cutting systems.
Then, letting the RIAA just keep on rolling off, just makes sense.
Well, unfortunately, the disc cutter designers, like Ortofon and most probably Neumann knew they had a problem, SO THEY FIXED IT, by resonating the cutting head to boost 1 dB or so at 20KHz TO MEET THEIR SPEC. SHEET. Therefore, IF you try to compensate and add a 50K shelf, you create a BOOST at 20K approaching 1 dB. Not so good, unless you use lossy high frequency transformers to compensate. Didn't know that, did you? Of course, the 50KHz resonance of a typical MC phono cartridge will easily fill in any losses of output in that region created by the LACK of a shelf, and after 50KHz, even I don't even care much, because that is the LIMIT of the best practical microphones (low noise is important too), master tape recorders, my UltraMaster 30 ips tape recorder is flat to approximately 45KHz, and 1/2 speed cutting systems.
Then, letting the RIAA just keep on rolling off, just makes sense.
Some people listen to classical music and other music played on acoustical, un-amplified instruments only. Some of them dislike any form of distortion. Undistorted reproduction gear will reproduce all types of music fairly faithfully to the way it was recorded, mixed and mastered.
JG, you have your definition of music and I have mine. It's not a matter or which language it's spoken in. An adolescent shrieking his lungs out 2 inches from a microphone, that signal put through fuzz boxes and big muffs, and then amplified to 130 db does not fall into my definition of music. Perhaps yours, perhaps other people's but not mine. If that is all there was to music, I would be perfectly content with a $79 boom box. IMO that is the stuff audio nightmares are made out of. The sound from hell the world would have been far better off without.
Steve,
One of the demos we do around here is to play a 30,000 cycle per second sine wave through an old Spica speaker. Everyone says they don't hear any thing. Then I turn it off and just about everyone realizes they can detect it going off and then back on. It does not sound like a clean tone but the energy is certainly perceived.
There are a number of researchers who believe it is possible to perceive small amounts of phase shift at 20,000 cycles per second.
If you tell me you can't hear it I will believe you, If John tells me he hears things who am I to argue?
ES
Of course classical music was not considered music at first as there was no singing. That Stan Kenton must not have had a clue either when he produce something clearly not music where people just sang notes.
And of course that Picasso guy ruined painting.
To me music is an expression of emotion. You are welcome to your world, you may need a passport to get into others.
Hi John,
I am also under the impression that some do and some don't choose to compensate for the Neumann time constant of 3.18 us. It is certainly easy to do if one chooses to do it; just add a zero at about 50 kHz by putting a resistor in series with the capacitor that gives you the 75us time constant.
However, since the 3.18 us zero is of first order, and it lies little more than an octave above 20 kHz, it will certainly make a difference in RIAA equalization at 20 kHz - on the order of 0.7 to 1 dB.
So, I think we can all agree that it makes a difference, but it is less clear whether it is right or wrong to include it. I am also under the impression that not all recording studios adhered uniformly to this.
Cheers,
Bob
Then where is the engineering justification for extending FR in a sound reproducing system beyond 20 KHZ?
The transformer I provided the graph for was a 10k:10k input transformer with a 600 ohm source Z. Not a transformer you'd be using for a moving coil step-up.
More comparable would be the CineMag CMMI-8PCA, which is -1dB at about 85kHz.
By the way, did you note that the spec for the Lundahl is with the secondary open circuit?
Will you be using it to drive the input FET's gate directly? Or will it be driving the 47k ohm of the MM input?
se
Pay attention, John. The graph was for a 10k:10k input transformer with a 600 ohm source Z. Not something you would be using for a moving coil step-up.
SY already pointed this out. Why do yo continue to ignore it?
se
- Status
- Not open for further replies.