Lumanauw
"This fact of having a "background" noisefloor can "boost" the details"
I read a little article published in Electronics World which said it has been scientifically proved. But instead of designing a power amp with difficultly modifiable noise, hiss, etc... it is just easier to introduce nicely sounding clarifying defaults at will using a "magic box" at the input of the amplifier.
~~~~~~~~ Forr
§§§
"This fact of having a "background" noisefloor can "boost" the details"
I read a little article published in Electronics World which said it has been scientifically proved. But instead of designing a power amp with difficultly modifiable noise, hiss, etc... it is just easier to introduce nicely sounding clarifying defaults at will using a "magic box" at the input of the amplifier.
~~~~~~~~ Forr
§§§
tlf9999
"People should have the right to be irrational."
Sure. belgian singer Jacques Brel even said :
"The right to be stupid shold be written in the constitution"
Yes, I agree. What I less agree, and contest using polite but hard words, is when people try to convince others to adher to their irrationnality.
~~~~~~~~ Forr
§§§
"People should have the right to be irrational."
Sure. belgian singer Jacques Brel even said :
"The right to be stupid shold be written in the constitution"
Yes, I agree. What I less agree, and contest using polite but hard words, is when people try to convince others to adher to their irrationnality.
~~~~~~~~ Forr
§§§
I've also gone down the Doug Self road of very low distortion and built several very low distortion amps that I've been ultimately dissatisfied with. The words sterile, cold, uninvolving go some way to describing the end result. More recently though I managed to bridge that 'low THD > musically involving' gap. My main tools are a 24 bit audio analyser and 200MHz CRO but they were unable to measure any change. Me thinks there is more to it than just THD signature. The project was a class A preamp / headphone amp. THD at 6KHz was below -118 which is the limits of my analyser as were IMD tests. It's a generic DS design but the EF VAS config favoured by DS has been substituted with a CFP, still with single pole Cdom compensation. Initially the extra VAS gain caused a 70MHz parasitic but I was able to remove it with VAS degeneration. The sound however was still cold and sterile and that character was actually made slightly worse after VAS degeneration. I used an F14 ferrite coil former as a bead on the VAS emitter and hey presto not only did the parasic vanish (as expected) but the treble became delishiously sweet and involving. At this stage I have no way of measuring the settling time or indeed the real world transient behaviour but it's worth investigating further. My design philosophy was ultra high speed and there are other aspects of the design that are a little unique such as special thermal tracking of current mirror collector voltages and DC offset.
I'm not sure to what extent these cct highlights are contributing but for for the time being I'm just enjoying the result.
I'm not sure to what extent these cct highlights are contributing but for for the time being I'm just enjoying the result.
The explanation is supersimple: following non-linear solid state devices will detect and intermodulate. This is nothing good, a kind of alchemy. Yes, distortion emphasizes details. This kind of emphasizing should be strongly avoided.
That's not it. Since the oscillation was at 15MHz, all the IMD will be centered around 15MHz, and I highly doubt that it makes any sonic difference. What's actually happening is that the superimposed 15MHz voltage is shifting the operating point of the finals to a more non-linear condition, and that's what is causing the apparant improvement in the sonics.
This kind of emphasizing should be strongly avoided.
This goes to personal preference. Obviously, the customer who wanted the amp to oscillate did not agree. Yes, it does "offend" one's sense of technical excellence to be producing amps with more distortion than you know you can achieve. However, in the final analysis, amps are for entertaining people with music, not distortion and spectrum analyzers. So who's ultimately "right"?
That's not it. Since the oscillation was at 15MHz, all the IMD will be centered around 15MHz, and I highly doubt that it makes any sonic difference. What's actually happening is that the superimposed 15MHz voltage is shifting the operating point of the finals to a more non-linear condition, and that's what is causing the apparant improvement in the sonics.
This kind of emphasizing should be strongly avoided.
This goes to personal preference. Obviously, the customer who wanted the amp to oscillate did not agree. Yes, it does "offend" one's sense of technical excellence to be producing amps with more distortion than you know you can achieve. However, in the final analysis, amps are for entertaining people with music, not distortion and spectrum analyzers. So who's ultimately "right"?
Hi, Forr,
You can make money here
Why don't make a "black box" preamp, that actually rides the original music with Mhz background? Those who doesn't know electronic will praise it as a "magic detail enhancher". It is not "bass" it is not "treble" it is not "BBE" it is not "loudness", it is something else. Revealing detail on every frequency. Oh no.... On the other hand, PMA wouldn't buy it
You can make money here
Why don't make a "black box" preamp, that actually rides the original music with Mhz background? Those who doesn't know electronic will praise it as a "magic detail enhancher". It is not "bass" it is not "treble" it is not "BBE" it is not "loudness", it is something else. Revealing detail on every frequency. Oh no.... On the other hand, PMA wouldn't buy it
You can make money here Why don't make a "black box" preamp, that actually rides the original music with Mhz background? Those who doesn't know electronic will praise it as a "magic detail enhancher". It is not "bass" it is not "treble" it is not "BBE" it is not "loudness", it is something else. Revealing detail on every frequency.
Yannow, I just might do that. I've seen way wierder stuff being sold for some surprising $$$$$.
Yannow, I just might do that. I've seen way wierder stuff being sold for some surprising $$$$$.
Miles Prower said:
I was playing with that. Adding sinus waves in MHz range to an audio band signal. Depending on circuit non-linearities, new components in audio band were created. It is about multiplication and subtraction, so one cannot refuse possibility of new frequency components in audio band.
15MHz - my CD has similar noise at the output. When filtered by audio buffer circuit, the sound has much better resolution. I do not agree with the idea that HF noise/interference brings anything audibly good.
I don't really know what happens, but I suspect it is far more than just "adding". It has to do with intermodulation, but I don't know what kind.
If you just "adding" Mhz sinusoidal to the music, maybe the result will be different than what I experienced. It is because the oscilating one is the output stage, and it has feedback to the differential. So must be more than just "adding". More like Intermodulation. But Miles Prower makes sense. It is too far from 15Mhz to 20khz. So what is actually happening?
One other strange thing. When the amp is set to not-oscilating, the threshold point of bias pot (VBE multiplier) is around 12 O'clock. Setting the bias less than 12 O'clock, the output is distorted.
But when the setting is in oscilating mode, even when I put the bias pot at minimal, the output is not distorted.
Is that oscilating amps needs less VBE bias than proper one?
If you just "adding" Mhz sinusoidal to the music, maybe the result will be different than what I experienced. It is because the oscilating one is the output stage, and it has feedback to the differential. So must be more than just "adding". More like Intermodulation. But Miles Prower makes sense. It is too far from 15Mhz to 20khz. So what is actually happening?
One other strange thing. When the amp is set to not-oscilating, the threshold point of bias pot (VBE multiplier) is around 12 O'clock. Setting the bias less than 12 O'clock, the output is distorted.
But when the setting is in oscilating mode, even when I put the bias pot at minimal, the output is not distorted.
Is that oscilating amps needs less VBE bias than proper one?
Hi LUMANAUW
Make money with a "magic box" ? Well I am not in the business and France is a so clever nation that a lot of things is done to dissuade you to enter it.
The magic box I planned to built (one day...) will not add high frequency components to the signal. It may include a noise adder but it is mainly based on an attenuator made of a resistor and a biased diode followed by a buffer, after a Ian Hickman's design . This will give better graded components of distorsion than any mythic tube à la 300. The generated distorsion will be added to the initial signal at will and sent to the power amplifier. Then funny experiements will start.
I think Mr Cheever, the one of the famous thesis, should have play with such a box before writing all his stuff about the harmonic content of amplifers.
Anybody with ultra-subjective claims should be submitted to blind tests with magic boxes of different kinds.
~~~~~~~ Forr
§§§
Make money with a "magic box" ? Well I am not in the business and France is a so clever nation that a lot of things is done to dissuade you to enter it.
The magic box I planned to built (one day...) will not add high frequency components to the signal. It may include a noise adder but it is mainly based on an attenuator made of a resistor and a biased diode followed by a buffer, after a Ian Hickman's design . This will give better graded components of distorsion than any mythic tube à la 300. The generated distorsion will be added to the initial signal at will and sent to the power amplifier. Then funny experiements will start.
I think Mr Cheever, the one of the famous thesis, should have play with such a box before writing all his stuff about the harmonic content of amplifers.
Anybody with ultra-subjective claims should be submitted to blind tests with magic boxes of different kinds.
~~~~~~~ Forr
§§§
I do not agree with the idea that HF noise/interference brings anything audibly good.
I don't get it myself. However, we're not talking about you or me. We're discussing
lumanauw's customer:
How do you account for that?
Sure, in one sense, you give the customer what he wants, even if you think it's "wrong" or utterly ridiculous. He's the guy who's paying your bills, after all.
I'd like to see the matter resolved. However, even if we knew exactly what was going on, you'd still have personal preference to deal with.
I don't get it myself. However, we're not talking about you or me. We're discussing
lumanauw's customer:
How do you account for that?
Sure, in one sense, you give the customer what he wants, even if you think it's "wrong" or utterly ridiculous. He's the guy who's paying your bills, after all.
I'd like to see the matter resolved. However, even if we knew exactly what was going on, you'd still have personal preference to deal with.
lumanauw said:
Hello lumanauw.
In what way did you eliminate oscillation? Is it possible that eliminating oscillation you did something that as a "side effect" it gave better (subjective) performance? Perhaps better performance is related to the "cure" for oscillations and not oscillations themselves? Perhaps also "curing" oscillations in a different way would not give you better sound?
Just a thought...
Hi Lumanauw,
Trouble with an amplifier that oscillates even mildly, it's a warning that something very much more severe maybe just around the corner - a temperature increase, different load, spike from the fridge!
You cannot leave an amplifier in such a state as it will surely bite you.
Cheers,
greg
Trouble with an amplifier that oscillates even mildly, it's a warning that something very much more severe maybe just around the corner - a temperature increase, different load, spike from the fridge!
You cannot leave an amplifier in such a state as it will surely bite you.
Cheers,
greg
Miles Prower
"Doug Self published a long polemic against B. Ollsen and the quasi-comp topology. Self had all the numbers to back up his claims against Ollsen. However, I've built them both ways, and Ollsen was right: quasi-comp sounds better AFAIC."
This assertion ("sounds better"... so repetitivly seen) is typical of what I call an inobjective comment. It is absolutely false that it sounds better, it is absolutely true you prefer the sound of Ollsen's amplifer (a huge lot of difference !) I may even prefer it too. Why ? Probably that in most cases, distorsion is prefered, it sounds as more detailed, nicer to our ears.
What is very perturbing is that if the numerrous electronic stages prior to the final power amp generate the same distorsion as the Ollsen amp (or any other claimed to have been subjectively designed) , the final result would probably judged as dirty.
I am not aware of IC's désigned for good sound and not having very good specs and a lot of feedback. And there usually are some IC's at the recording stage. Do they make so much harm to the perceived sound ?
~~~~~~~~~ Forr
§§§
"Doug Self published a long polemic against B. Ollsen and the quasi-comp topology. Self had all the numbers to back up his claims against Ollsen. However, I've built them both ways, and Ollsen was right: quasi-comp sounds better AFAIC."
This assertion ("sounds better"... so repetitivly seen) is typical of what I call an inobjective comment. It is absolutely false that it sounds better, it is absolutely true you prefer the sound of Ollsen's amplifer (a huge lot of difference !) I may even prefer it too. Why ? Probably that in most cases, distorsion is prefered, it sounds as more detailed, nicer to our ears.
What is very perturbing is that if the numerrous electronic stages prior to the final power amp generate the same distorsion as the Ollsen amp (or any other claimed to have been subjectively designed) , the final result would probably judged as dirty.
I am not aware of IC's désigned for good sound and not having very good specs and a lot of feedback. And there usually are some IC's at the recording stage. Do they make so much harm to the perceived sound ?
~~~~~~~~~ Forr
§§§
Hi, Dtrif,
The oscilation is "set" to occur only at full power with square wave. At moderate level, it doesn't oscilate, otherwise how can the power amp stand even 1 minute?
There are some things that actually wrong, but usually DIYer's don't realize. Do you know that plain triple darlington or CFP oscilates at full power? It only shows up when you test the amp with sufficient dummy load and scope. The oscilation maybe small, but it is oscilation. With small/moderate level, this oscilation will not show up.
You can look at the cure, and how this "cure" can be made "adjustable" by searching here about CFP oscilation topic and answers by Nelson Pass.
Hi, Amplifierguru,
Yes, you are right. My opinion also that any oscilation, mild or severe should be not there. What is confusing me, why my friend's ear prefers the oscilating one? Hence this thread
The oscilation is "set" to occur only at full power with square wave. At moderate level, it doesn't oscilate, otherwise how can the power amp stand even 1 minute?
There are some things that actually wrong, but usually DIYer's don't realize. Do you know that plain triple darlington or CFP oscilates at full power? It only shows up when you test the amp with sufficient dummy load and scope. The oscilation maybe small, but it is oscilation. With small/moderate level, this oscilation will not show up.
You can look at the cure, and how this "cure" can be made "adjustable" by searching here about CFP oscilation topic and answers by Nelson Pass.
Hi, Amplifierguru,
Yes, you are right. My opinion also that any oscilation, mild or severe should be not there. What is confusing me, why my friend's ear prefers the oscilating one? Hence this thread
Hi Lumanauw,
An amplifier is always a compensated oscillator.
I have experimented often the things that you say. I believe that is too much attention to the THD and a few to other dynamicses of the amp.
A amp has to be before all a good interface for the speakers. his THD has a more low weight on the final sound of what much believe.
Because " sounds better " a little compensated amp?
Because the classical compensations alter the DF, the output phase response and other parameters.
I agre totally the Graham's DF theories (and first cicle analysis...).
If you uses the "alternatives" techniques of compensation, in accord with the " inverse driven " test, the acoustic performances increases (like no compensate amp).
I have experimented the feed-forward comp. to eliminate the Cdom effects (in my nested-loop opamp, my_ref), with optimize results...
Ciao
Mauro
An amplifier is always a compensated oscillator.
I have experimented often the things that you say. I believe that is too much attention to the THD and a few to other dynamicses of the amp.
A amp has to be before all a good interface for the speakers. his THD has a more low weight on the final sound of what much believe.
Because " sounds better " a little compensated amp?
Because the classical compensations alter the DF, the output phase response and other parameters.
I agre totally the Graham's DF theories (and first cicle analysis...).
If you uses the "alternatives" techniques of compensation, in accord with the " inverse driven " test, the acoustic performances increases (like no compensate amp).
I have experimented the feed-forward comp. to eliminate the Cdom effects (in my nested-loop opamp, my_ref), with optimize results...
Ciao
Mauro
Mauropenasa
diyAudio Member
>An amplifier is always a compensated oscillator.
An oscillator is an amplifier made voluntarily oscillating.
Egg and chicken dilemna ?
A famous electronician's Murphy's law is :
an amplifier always oscillates when one does not want it.
an oscillator never oscillates when one wants it.
~~~~~~~~ Forr
§§§
diyAudio Member
>An amplifier is always a compensated oscillator.
An oscillator is an amplifier made voluntarily oscillating.
Egg and chicken dilemna ?
A famous electronician's Murphy's law is :
an amplifier always oscillates when one does not want it.
an oscillator never oscillates when one wants it.
~~~~~~~~ Forr
§§§
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.