I have experienced something similar. In my mods of Philips and Marantz cd players, my experience has been this:lumanauw said:
- 3 units of CD63 (SE and mk II), not matter how extensively modded (power supply cleanup, better coupling caps, good stable opamps, etc) would not produce the same depth as Philips CD931 (2 units), CD950 (2 units), even a humble Marantz CD48 and very humble Philips CD750.
The soundstage width and depth on all the units started to improve mostly when power supplies to chips were carefully bypassed with low inductance film caps and the rectifier diodes were changed to schottky. However the CD63s responded much less in depth of image, even though lateral soundstage improved considerably. (Yes, I have implemented the ground loop track mod on the oscillator circuit of the CD63. The CD931 has a similar loop, and that was re-routed too.)
Mr Evil said:
Yes tube amps produce distortion, and so do solid state amplifiers. Tube amps can be operated with zero negative feedback because tubes are inherently linear devices. Transistors require huge amounts of negative feedback to tame because they are not linear devices. This leads to sonic differences between the two classes of amplifiers and within the classes themselves.
By saying above that "you can see that it may be easy to pick out" you are agreeing with my comment that your original statement of "because the speaker and room dictate to such a large degree what one hears that signatures of the electronics are inaudible" is in error. Recall I used the term "shortsighted".
A personal preference of tube amps or sand amps is one's right. I'm not kicking vacuum in your face, so please don't kick sand in mine.
You seem to be taking this the wrong way. At no point did I express any bias for or against any particular type of technology (vacuum vs semiconductor). What I said was that the amount of colouration imparted by the speakers/room is far, far greater than that imparted by hi-fi electronics.rcavictim said:
To clarify further, you should know that I would exclude any amplifier that introduces significant distortion of its own from 'hi-fi', which by definition implies low distortion. That is in no way a dig against tubes, it's just a matter of definition.
Whereas the difference between a fairly cheap semiconductor amplifier and the most expensive is minimal at best, the difference between any two sets of speakers, even two well designed ones, can be very obvious. In fact I think tube amps and speakers are in many ways similar, insofar as they tend to both be designed to produce a pleasing sound rather than one which is true to the original.
The lack of metric for defining the 'perfect' loudspeaker leads to many design philosophies and thus many different sounds. In contrast, the ideal (hi-fi) amplifier is very easily descibed and a modern amplifier can come pretty close (certainly close enough that the difference is tiny compared to those parts of the audio chain that come before and after). That isn't to say that it is not worth trying to improve amplifier technology, but I think it would be ill-advised to spend lots of money on an amp purely to obtain an audible improvement. Better to spend the money on a comfy chair
Room acoustics is a killer. Most people don't know how important it is, or are unable to do anything to rectify the situation because of practical concerns. In fact I think it a shame that a look around most hi-fi shops will reveal a whole host of very expensive eletronics but zero aids for room acoustics, which could potentially yield much greater benefits at much lower cost.
1970's tubes were more linear than 1970's transistors [think about 2N3055 ] but tubes have not evolved like transistors. Some modern transistors show flat gain from 100mA to 5A or even 10A
As a refference look at MJ21194 datasheet from On-Semi :
A single pair of MJ21194/MJ21196 operated as a class B follower sourcing 100Wrms@8ohms with +-50V rails and driven from 50ohms impedance produces 0.8% THD max with unmatched gains and 0.08% THD max with matched gains and this is without global feedback
Today [year 2004, not 1970] we can consider tubes as the non-linear devices and bipolar transistors as the linear ones
] but tubes have not evolved like transistors. Some modern transistors show flat gain from 100mA to 5A or even 10AAs a refference look at MJ21194 datasheet from On-Semi :
A single pair of MJ21194/MJ21196 operated as a class B follower sourcing 100Wrms@8ohms with +-50V rails and driven from 50ohms impedance produces 0.8% THD max with unmatched gains and 0.08% THD max with matched gains and this is without global feedback
Today [year 2004, not 1970] we can consider tubes as the non-linear devices and bipolar transistors as the linear ones
Mr Evil said:Whereas the difference between a fairly cheap semiconductor amplifier and the most expensive is minimal at best, the difference between any two sets of speakers, even two well designed ones, can be very obvious. In fact I think tube amps and speakers are in many ways similar, insofar as they tend to both be designed to produce a pleasing sound rather than one which is true to the original.
The lack of metric for defining the 'perfect' loudspeaker leads to many design philosophies and thus many different sounds. In contrast, the ideal (hi-fi) amplifier is very easily descibed and a modern amplifier can come pretty close (certainly close enough that the difference is tiny compared to those parts of the audio chain that come before and after). That isn't to say that it is not worth trying to improve amplifier technology, but I think it would be ill-advised to spend lots of money on an amp purely to obtain an audible improvement. Better to spend the money on a comfy chair
Room acoustics is a killer. Most people don't know how important it is, or are unable to do anything to rectify the situation because of practical concerns. In fact I think it a shame that a look around most hi-fi shops will reveal a whole host of very expensive eletronics but zero aids for room acoustics, which could potentially yield much greater benefits at much lower cost.
one more reasonable human being was added to the forum
"thumbs up" and welcome to the forum !!!
Eva i didnt know that 2n3055 is that badEva said:
Eva said:1970's tubes were more linear than 1970's transistors [think about 2N3055
As a refference look at MJ21194 datasheet from On-Semi :
A single pair of MJ21194/MJ21196 operated as a class B follower sourcing 100Wrms@8ohms with +-50V rails and driven from 50ohms impedance produces 0.8% THD max with unmatched gains and 0.08% THD max with matched gains and this is without global feedback
Today [year 2004, not 1970] we can consider tubes as the non-linear devices and bipolar transistors as the linear ones
Eva,
Boulderdash! A silicon transistor consists of diode junctions. Semiconductor diode junctions are NON LINEAR devices. This was a fact in 1970 and this has not changed as of today. Their transfer function is that of approaching a square law, not a linear function.
Vacuum tubes were inherently linear devices in 1930 and guess what... they remain so today.
Although the non-linearity of transistors hasn't changed in 40 years it seems the advent of the home computer and the internet has made circular discussions more circular.
Show me how you can get low distortion and linear response from a non-linear device WITHOUT applying feedback!
Third channel and PMPO
The third channel was used in the seventies... they just used the red wires from stereo amplifier... this used the out of phase signals from one channel related the other.... the phase errors are added in this fake third channel... do not know if this is the case you are discussing..... but makes one beautifull effect...but that one was one trick... a fake third channel.
The PMPO, is Peak Musical Power Output.... something stupid, that multiplies by 10, now a days 20 and i could see some factories that is multiplying by 30 the left plus rigth channel power added, and measured in 10 percent distortion levels.... the power have a very big number to influence people.... the power is bigger then the consumed power written in the amplifier back panel...normally, the RMS power maximum is around a half...or if you prefer 60 percent of the consumption power..... half the energy is transformed in heat and the other half is transformed in audio.... depends on the amplifier's efficiency..... that variates from 50 to 65 percent class AB circuits...beeing class A, everything will change a lot, of course!
In the true, the value is always bigger than the real value...you can divide by two...because you have a lot of things that make energy consumption too, LEDs, lamps, motors and displays....those use energy and that energy is appearing in the back panel inscription...there, you will see the lamps consumption together audio consumption. Also, the maximum power is normally clipped power...this way, not the power you will use... so...things is less.
In my country this PMPO power is called Potencia Musical Para Otarios..... Potenza Musicale per idioticci, Potensa Musical para idiotas, Puissance Musicale pour l'idiots, Musical power to foll people....normally much bigger than the mains consumption...this is crazy... impossible to create energy!..... you always loose...not energy creation... this is really crazy!
regards,
Carlos
The third channel was used in the seventies... they just used the red wires from stereo amplifier... this used the out of phase signals from one channel related the other.... the phase errors are added in this fake third channel... do not know if this is the case you are discussing..... but makes one beautifull effect...but that one was one trick... a fake third channel.
The PMPO, is Peak Musical Power Output.... something stupid, that multiplies by 10, now a days 20 and i could see some factories that is multiplying by 30 the left plus rigth channel power added, and measured in 10 percent distortion levels.... the power have a very big number to influence people.... the power is bigger then the consumed power written in the amplifier back panel...normally, the RMS power maximum is around a half...or if you prefer 60 percent of the consumption power..... half the energy is transformed in heat and the other half is transformed in audio.... depends on the amplifier's efficiency..... that variates from 50 to 65 percent class AB circuits...beeing class A, everything will change a lot, of course!
In the true, the value is always bigger than the real value...you can divide by two...because you have a lot of things that make energy consumption too, LEDs, lamps, motors and displays....those use energy and that energy is appearing in the back panel inscription...there, you will see the lamps consumption together audio consumption. Also, the maximum power is normally clipped power...this way, not the power you will use... so...things is less.
In my country this PMPO power is called Potencia Musical Para Otarios..... Potenza Musicale per idioticci, Potensa Musical para idiotas, Puissance Musicale pour l'idiots, Musical power to foll people....normally much bigger than the mains consumption...this is crazy... impossible to create energy!..... you always loose...not energy creation... this is really crazy!
regards,
Carlos
rcavictim :
This is the datasheet for a common 'High Performance Audio Power Pentode' tube : http://www.amplimo.nl/download/el34.pdf
Transfer characteristics figures are at the end of page 2
And this figure shows transfer characteristics of a 300B power triode usually found in high end class-A non-global-feedback $3.000 stuff : http://www.emissionlabs.com/datasheets/300-302-320-c.gif
Now, please, could you tell to all us where is the linearity on these graphs? I only see curved lines, not a single straight line!!
It's pretty obvious that tubes with more than one driving electrode are allways very non linear, so I will discard these
Now you may argue that triode tubes are very linear in certain current/voltage regions and I would fully agree, but all bipolar transistors are also very linear in certain current/voltage regions
This is the datasheet for a common 'High Performance Audio Power Pentode' tube : http://www.amplimo.nl/download/el34.pdf
Transfer characteristics figures are at the end of page 2
And this figure shows transfer characteristics of a 300B power triode usually found in high end class-A non-global-feedback $3.000 stuff : http://www.emissionlabs.com/datasheets/300-302-320-c.gif
Now, please, could you tell to all us where is the linearity on these graphs? I only see curved lines, not a single straight line!!
It's pretty obvious that tubes with more than one driving electrode are allways very non linear, so I will discard these
Now you may argue that triode tubes are very linear in certain current/voltage regions and I would fully agree, but all bipolar transistors are also very linear in certain current/voltage regions
to rcavictim:
Hmm, i think EVA has studied electronics theory, as she is working
as some electrical engineer for developing PSUs... (if i am not wrong)
So, she knows much more about electronic devices than us together...
And, i agree with her.
To destroyerX:
This "third channel" still exist, it's nowadays driven by a separate
amplifier, but is basically the same. It delivers the "difference" between
L/R channel, at its best only these signals that would cancel out each other.
It's called rearchannel with dolby prologic...
The simplest surround-decoder consist of some resistors...
Mike
Hmm, i think EVA has studied electronics theory, as she is working
as some electrical engineer for developing PSUs... (if i am not wrong)
So, she knows much more about electronic devices than us together...
And, i agree with her.
To destroyerX:
This "third channel" still exist, it's nowadays driven by a separate
amplifier, but is basically the same. It delivers the "difference" between
L/R channel, at its best only these signals that would cancel out each other.
It's called rearchannel with dolby prologic...
The simplest surround-decoder consist of some resistors...
Mike
I think in this case the amp needs to be as precise as possible.
I would say high bandwidth to keep delay constant, means phaseshift
increases linear with frequency.
I am not sure, but for preserving the original phaseshiftinformation
for the brain, any harmonics introduced by the amp should be false
information added.
So additionally all phaseshiftcompensating caps need to be perfect
adjusted to the circuit.
Can't say how my amp performs for this, i only have mono-prototype...
Mike
I would say high bandwidth to keep delay constant, means phaseshift
increases linear with frequency.
I am not sure, but for preserving the original phaseshiftinformation
for the brain, any harmonics introduced by the amp should be false
information added.
So additionally all phaseshiftcompensating caps need to be perfect
adjusted to the circuit.
Can't say how my amp performs for this, i only have mono-prototype...
Mike
Hi, Epupa Epops,
Is there any posiblity to built an ultralow distortion with single differential? Or it is always with complementary differential, like you and John Curl likes?
Question no.2. What do you think of amp configuration of X or SuSy? Is this better than complementary differential in eliminating distortion?
Is there any posiblity to built an ultralow distortion with single differential? Or it is always with complementary differential, like you and John Curl likes?
Question no.2. What do you think of amp configuration of X or SuSy? Is this better than complementary differential in eliminating distortion?
With single differencial is it possible, look for example at patent from Halcro or at original connection from Robert Cordell. SuSy from Pass : I have any experience with it, but it is one stage moreover to high OL gain amp, so I should be careful ( stability ). X configuration is long time examine, so there problem with stability probably will be not, but there is problems with higher distortion at higher levels.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.