All that RF may not manifets itself as audio on the output but it can both add distorion as the junctions processing your audio try to process the rf and if its successful the following amp then needs to deal with it. Further, if the amp can then the speaker nees to. The first time you see a tweeter smoke with no sound you face the reality of dealing with RF.
Switching amps are a special case, the "RF" (100-500 KHz) carries the information. If you look at the current flowing through the speaker you will see a whole lot less of the RF that you might have expected to. (The inductive nature of all speakers helps with this.)
Switching amps are a special case, the "RF" (100-500 KHz) carries the information. If you look at the current flowing through the speaker you will see a whole lot less of the RF that you might have expected to. (The inductive nature of all speakers helps with this.)
PMA said:
Would this be by using speakers, ears and rooms with typical background noise floor of 40dB or higher ? And using typical home listening levels of 80-90dB ?
I've seen some rather fancy methods of extracting the signal well below the noise floor (radio astronomy is current state of the art in this field (*)), but 70-80dB below is totally unheard of. I'd dearly like to understand the mechanisms involved.
(*) there could be some more advanced stuff that is still classified for all I know
john curl said:
Whatever they do, there could be far more effective ways of doing it, without any woodoo at all. Shield RF ? No problem, very effective methods are well known, tried and tested. Block cosmic rays ? Absorb neutrinos ? Modify the time-space curve around the output stage ? Just spell it out and then we can all go and try to fix it anyway we could.
But these devices are touted to do something mystical, something we don't yet know about; nevertheless, we still claim that it somehow magically "improves sound". Using sighted tests.
Well, that is the approach that makes me cringe, and that is where any scientific discussion ends.
Hi Demian,
For the RF plots: Advantest 9kHz-3GHz spectrum analyzer with Tracking generator, all set up is 50 ohm impedance (Rout, Rin, BNC-SMB RG316 cables and microstrip boards)
For the low freq. plots: DDS sweep generator and EMU 1616m soundcard with Spectralab SW. Also using the same cables and microstrip boards as before.
Hi Hannes,
I think the archive is downloadable w/out login. It's big, I don't know how to attach it here
1audio said:
I recently heard two people talk about class D: Martin Colloms (Munich Hi-end show) from the pov of audio reviewer, and Bruno Putzeys (AES Amsterdam) from the pov of Hypex designer. Both came to the same conclusion: the class D that sounds best are the ones that control RF and EMI issues best.
Making a class D linear in the audio band is not that big a deal, the other stuff is. There were tales of class D amplifiers that distorted the sound of an audio chain when they were switched on, even when they (class D amp) weren't even connected to the chain. Or class D amps that spouted so much RF back out their input connection that it totally upset the driving CD player.
Jan Didden
Hi Roberto, thanks for pointing that out! It's indeed downloadable without registration, it's here:
http://www.audiofaidate.it/forum/uploaded/caps_tests.zip
Thank you very much for that marvelous work! Very interesting!!
Have fun, Hannes
The problem with audio is that it takes a human to evaluate it, ultimately. No matter how you test, conjecture, or rationalize, you may find that many attempts to make quality audio will not be very successful, and you should not have wasted your time in the first place, and just bought something. The reason for this is beyond simple harmonic distortion measurements, especially THD which can be almost useless, unless you know the harmonic content of the distortion. This has been known for about 70 years, and has been in print in the 'Radiotron Designers Handbook' since 1941 at least.
Each audio designer has faced the same questions from the beginning. Many of us have just learned from experience what is necessary, above and beyond what would appear on a spec sheet. In fact, almost any op amp made for the last 40 years or more should be able to satisfy many audio needs. Why then are we always looking for a better sounding one, or making discrete designs still?
Yes, 40 years ago, I actually thought that a first class preamp could be made with a 4 channel IC chip. Maybe it will be yet, but I haven't found the device as of yet.
Even when I buy a premium car radio or tv, I must live with the sound compromises that I know do not have to be there. This is often because the people in charge of these designs are too indifferent to what would improve sound quality, backed by engineer's rationalizations.
I
Each audio designer has faced the same questions from the beginning. Many of us have just learned from experience what is necessary, above and beyond what would appear on a spec sheet. In fact, almost any op amp made for the last 40 years or more should be able to satisfy many audio needs. Why then are we always looking for a better sounding one, or making discrete designs still?
Yes, 40 years ago, I actually thought that a first class preamp could be made with a 4 channel IC chip. Maybe it will be yet, but I haven't found the device as of yet.
Even when I buy a premium car radio or tv, I must live with the sound compromises that I know do not have to be there. This is often because the people in charge of these designs are too indifferent to what would improve sound quality, backed by engineer's rationalizations.
I
john curl said:
Here we go again. It is impossible to discuss if you keep shifting the subject, building your own straw arguments and answering your own questions.
I asked a simple question - what is it that those magic devices do to make an audio device 'better' ?
I've met Shun Mook for example, and if there is a definition of a con man, he fits it perfectly. There is not one iota of engineering in him. "Successful audio designer". Sure.
PMA said:
And what would that be ?
You see, we know that we can make a toothbrush much better than ones we have in our bathrooms. Aircraft aluminium, 316 stainless steel, CNC machined handle, bristles of equal thickness and length to within couple of nanometers. Beutifully made. Flawlesly engineered. Crafted to perfection. Absolutely no question where the money went. Would you pay 3000$ for it ? And will it really make a difference ?
- Status
- Not open for further replies.