I've yet to hear an amplifier (or DAC) with no added "electronic" coloration, or conversely, able to reproduce the full gamut of tone colors and vividness of live music.
Gary Pimm's Class A amplifier, though, is pretty close. Out of all the amps I've heard, it's certainly in the Top Ten.
Gary Pimm's Class A amplifier, though, is pretty close. Out of all the amps I've heard, it's certainly in the Top Ten.
Last edited:
Lynn.
As I said I understand and remember reading your reasons for using a class A tube amp. I also see no problem with using a hybrid SS/tube amplifier. Where you just can't do that and expect 20 watts to do the job is with a common direct radiator type of loudspeaker. you just run out of power and into major distortion. Now your insistence on zero feedback I think would have many designers in disagreement even if they are tube aficionados. I think you have just learned to like and expect the type of harmonic distortion that you will produce with a non-feedback amplifier, I don't believe that it truly has an inherent superiority to a feedback design, it is just a different harmonic content. Feedback has been used to linearize tube amplification going back to at least the Williamson type of amplifier circuits. I can't and wouldn't argue your tastes in music reproduction, we all like what we have come to expect and what we are exposed to. Your chose is just as relevant as anyone else. I am amazed that the old Altec drivers have remained and are reproduced today, I remember them well though back when we did PA work even with A2's and such we had decided that we liked the old JBL drivers of the time better.
As I said I understand and remember reading your reasons for using a class A tube amp. I also see no problem with using a hybrid SS/tube amplifier. Where you just can't do that and expect 20 watts to do the job is with a common direct radiator type of loudspeaker. you just run out of power and into major distortion. Now your insistence on zero feedback I think would have many designers in disagreement even if they are tube aficionados. I think you have just learned to like and expect the type of harmonic distortion that you will produce with a non-feedback amplifier, I don't believe that it truly has an inherent superiority to a feedback design, it is just a different harmonic content. Feedback has been used to linearize tube amplification going back to at least the Williamson type of amplifier circuits. I can't and wouldn't argue your tastes in music reproduction, we all like what we have come to expect and what we are exposed to. Your chose is just as relevant as anyone else. I am amazed that the old Altec drivers have remained and are reproduced today, I remember them well though back when we did PA work even with A2's and such we had decided that we liked the old JBL drivers of the time better.
Incorrect. If you do the maths, that demonstrates that such is not true; however, most 20 watt amps also have miserable power supplies, which totally cripples the ability of the amp to do the job - sort all that out, and the 20 watter will easily produce deafening sound, with minimal audible distortion, into those types of speakers.
Zero-feedback amplifiers, if competently designed, have a soft clipping characteristic, very much like a limiter for a radio station. This gives them about 3 to 4 dB of progressively compressed headroom where distortion gradually increases.
The Karna has 0.1% distortion at 15 watts, almost entirely 2nd and 3rd harmonic. Steady-state, it can put out 30 watts at 3%, and for transients in the 100 millisecond range, 50 watts or more. Slewing is pretty good too: it can put out full power at 500 kHz, although I didn't have the nerve to try that for more than a few seconds.
The 20-year-old Ariels are 92 dB/meter, and I listen at a distance of 11 feet. Peak sound levels are typically center-dominated (instead of hard left or hard right), so power from both channels is available. The system as a whole is audibly compressing at 105 dB (at the listening position), which is consistent with 40 watts (both channels) steady-state, and 100 watts for brief transients.
Raising the efficiency by 5 to 8 dB is a simple way of gaining headroom while retaining an amplifier that I prefer over all others. If I want to listen to transistors, I switch over to the Marantz MM8003 power amplifier, which is mostly used for home theater duties.
The external crossovers of the Ariels have a set of DPDT input switches that quickly change over from the 2-channel triode system to the 5-channel home theater system. The crossover input selector switches grounds as well as "hot" signals, so the grounds between the two systems are isolated from each other. I can listen to either transistors or triodes, and upgrade the 5-channel transistor side of things if I want ... maybe Class D, maybe a more sophisticated solid-state amp. But it's not a priority.
If the readers of the DIY forum prefer Class AB transistor amplifiers, good for them! It saves a lot of money, and greatly widens choices of potential loudspeakers. If 200 watts or more is on tap, then you can choose electrostats, magnetic-planars, MBL or Ohm radial omnidirectional speakers, or the traditional favorite, direct-radiators with efficiencies in the 85~93 dB/meter range.
The Karna has 0.1% distortion at 15 watts, almost entirely 2nd and 3rd harmonic. Steady-state, it can put out 30 watts at 3%, and for transients in the 100 millisecond range, 50 watts or more. Slewing is pretty good too: it can put out full power at 500 kHz, although I didn't have the nerve to try that for more than a few seconds.
The 20-year-old Ariels are 92 dB/meter, and I listen at a distance of 11 feet. Peak sound levels are typically center-dominated (instead of hard left or hard right), so power from both channels is available. The system as a whole is audibly compressing at 105 dB (at the listening position), which is consistent with 40 watts (both channels) steady-state, and 100 watts for brief transients.
Raising the efficiency by 5 to 8 dB is a simple way of gaining headroom while retaining an amplifier that I prefer over all others. If I want to listen to transistors, I switch over to the Marantz MM8003 power amplifier, which is mostly used for home theater duties.
The external crossovers of the Ariels have a set of DPDT input switches that quickly change over from the 2-channel triode system to the 5-channel home theater system. The crossover input selector switches grounds as well as "hot" signals, so the grounds between the two systems are isolated from each other. I can listen to either transistors or triodes, and upgrade the 5-channel transistor side of things if I want ... maybe Class D, maybe a more sophisticated solid-state amp. But it's not a priority.
If the readers of the DIY forum prefer Class AB transistor amplifiers, good for them! It saves a lot of money, and greatly widens choices of potential loudspeakers. If 200 watts or more is on tap, then you can choose electrostats, magnetic-planars, MBL or Ohm radial omnidirectional speakers, or the traditional favorite, direct-radiators with efficiencies in the 85~93 dB/meter range.
Last edited:
Lynn, I think almost all serious DIY ers who have been around would have a similar view on this. What Gary Pimm enjoys in the US, likely I enjoy here in the UK but achieved differently but with similar basic electronic principles. Certainly your Karna reflects a nice circuit and balance of selected parts, and mine are again, like the Gary,s different than yours and theirs. This type of friendly stuff no doubt happens at the meetings of small audiophile groups i.e DHT, PA , Theatre music
This is good because it does give hope to many new DIYers that there are quite few options open to them. And I like the Ariel replacement speakers of Gary Dahls, which are beautifully made. I would make them different again jsut for fun and the ywould work too. There is a slight rivalry in there with all DIYers and that is human nature. And the guy from Israel , disarmingly makes his own valve amps and his may be exceptional, while he asks for unique circuit diagrams for state of art solid state Gary Pimm amp. This is waht really does keep DIY going and its has needed it in the world recession. I rue the day if and when you could only buy complete manufactured units. Digital may have brought that a little too close.
Lynn
I strongly agree with this view, and would add that one could even progress to making ones own electrostatic or Magneplanar equivalent. WE can if inclined make anything with the now available materials parts and circuits, so DIY ers are in the best position today to do this if so inclined and can reasonably afford it.
Hi Lynn ,
When you say zero feedback, is this truly a zero feedback design without any degeneration ....?
This would put your system sensitivity at listening position @82db/w add in second channel and some room gain 85db /87 db ? 15 watts will not give you much in dynamic head room, of course 8db increase in sensitivity will increase it some fold ...
If you look around we've been trying to pry that out of Lynn since the '90s! Very coy -- but I don't blame him
However there's enough in his own web page; forum postings and product reviews to infer that it will include massed vocals, combined with an orchestra and/or organ, recorded in a way that captures the ambient acoustics of the venue.
If you look around we've been trying to pry that out of Lynn since the '90s! Very coy -- but I don't blame him
However there's enough in his own web page; forum postings and product reviews to infer that it will include massed vocals, combined with an orchestra and/or organ, recorded in a way that captures the ambient acoustics of the venue.
I really don't know why this would be kept a secret. If i knew what recordings he referenced to i would could judge the credibility of his statements or get a better idea where he is coming from. Why would you not blame him?
Last edited:
It's a pretty long list here are a few exceptional recordings I go to to evaluate the music playback in the big system - surely none are "perfect" that's why I find it hard to say absolutes like never or always when referring to playback - all recordings are flawed as far as I can tell so there is no "absolute"
try these
Eiji Oue - Rachmaninoff: Symphonic Dances RR LP
Audio Sympony Check up your donds RCA LP
Fennell Trttico RR CD
T Houston Pressure Cooker Sheffield D to disk
Sheffield "track" record LP & CD
Tor Dietrichdon Global Village CD
Bernie Krause Citadels of Mystery MF LP
John Klemmer Touch MF LP
Ala Carte Brass & Percussion Wildchild CD
Sheffield Track Record LP
Rusted Root - When I awoke CD
Pretenders Isle of View CD
Arch Choir Mapleshade CD
many many more LOL
here are a few exceptional recordings I go to to evaluate the music playback in the big system - surely none are "perfect" that's why I find it hard to say absolutes like never or always when referring to playback - all recordings are flawed as far as I can tell so there is no "absolute"try these
Eiji Oue - Rachmaninoff: Symphonic Dances RR LP
Audio Sympony Check up your donds RCA LP
Fennell Trttico RR CD
T Houston Pressure Cooker Sheffield D to disk
Sheffield "track" record LP & CD
Tor Dietrichdon Global Village CD
Bernie Krause Citadels of Mystery MF LP
John Klemmer Touch MF LP
Ala Carte Brass & Percussion Wildchild CD
Sheffield Track Record LP
Rusted Root - When I awoke CD
Pretenders Isle of View CD
Arch Choir Mapleshade CD
many many more LOL
Pooh, thanks for the list. I will try to make one of my own for the good of the order.
When evaluating equipment, Lynn prefers to use recordings that are difficult for the system to "sort out." I noticed this the first time I visited his home in Portland in the 1990s. We listened to a DAC I brought along with me. The music he put on sounded pretty bad (to both of us) and he blamed the DAC. I thought it was just a bad recording, but later found that he was right. Fortunately, we have much better DACs now.
Two things have always impressed me about Lynn's ability to evaluate equipment by listening. First, he is fast. It doesn't take him long at all to zero in on problems. And he is remarkably skilled at identifying the source of a specific issue in a system. I can hear these issues well enough, but it takes me quite a bit longer and I'm generally not sure where the specific problems originate. He can usually even listen "through" one problem while analyzing another. It was fascinating to see the process unfold during the months we were working on the Aurora amplifier.
I don't think that knowing Lynn's preferred recordings for evaluation would be of any help to the rest of us. But I can honestly say that he knows what he is hearing when he uses them.
Lynn and I are sensitive to different aspects of reproduced sound. He has written about his preferences: dynamics and transparency are especially important to him, and of course the lack of electronic colorations. My first priority is getting the correct tone color of orchestral instruments. I spend a lot of time in the presence of orchestral instruments and have a very good auditory memory, so this is very important to me.
When listening for enjoyment, Lynn and I have different tastes, yet enjoy one another's recordings. And we generally end up with the same verdict when evaluating equipment, but as I already mentioned, Lynn is much faster.
Gary Dahl
When evaluating equipment, Lynn prefers to use recordings that are difficult for the system to "sort out." I noticed this the first time I visited his home in Portland in the 1990s. We listened to a DAC I brought along with me. The music he put on sounded pretty bad (to both of us) and he blamed the DAC. I thought it was just a bad recording, but later found that he was right. Fortunately, we have much better DACs now.
Two things have always impressed me about Lynn's ability to evaluate equipment by listening. First, he is fast. It doesn't take him long at all to zero in on problems. And he is remarkably skilled at identifying the source of a specific issue in a system. I can hear these issues well enough, but it takes me quite a bit longer and I'm generally not sure where the specific problems originate. He can usually even listen "through" one problem while analyzing another. It was fascinating to see the process unfold during the months we were working on the Aurora amplifier.
I don't think that knowing Lynn's preferred recordings for evaluation would be of any help to the rest of us. But I can honestly say that he knows what he is hearing when he uses them.
Lynn and I are sensitive to different aspects of reproduced sound. He has written about his preferences: dynamics and transparency are especially important to him, and of course the lack of electronic colorations. My first priority is getting the correct tone color of orchestral instruments. I spend a lot of time in the presence of orchestral instruments and have a very good auditory memory, so this is very important to me.
When listening for enjoyment, Lynn and I have different tastes, yet enjoy one another's recordings. And we generally end up with the same verdict when evaluating equipment, but as I already mentioned, Lynn is much faster.
Gary Dahl
Last edited:
Thanks , the highest quality system to me is one that connects directly deep into the soul. One that brings the emotion of the musicians direct to the heart just like live - The above recordings above are more for sound then this feeling I seek in my system although some of them can pull it off others are just "sound." Sound is important but there is more to it then that, a more complete envelope. Horns do it best for me :}
When Lynn said he has never heard electronics "sound" like live music I stepped back and thought what recording does he use to make this broad statement? I would like to know so I can check the recording out in my room - I don't really care how he listens or why he says this I just want to hear it in my playback - IOW when evaluating the sound of the system for no coloration what recordings are used?
When Lynn said he has never heard electronics "sound" like live music I stepped back and thought what recording does he use to make this broad statement? I would like to know so I can check the recording out in my room - I don't really care how he listens or why he says this I just want to hear it in my playback - IOW when evaluating the sound of the system for no coloration what recordings are used?
Last edited:
Cone edge resonance damping
Therefor a simplified version for a single 2nd order resonance with center frequency < 1 kHz from the same directional axis:
If such a resonance has a bandwidth < 1/10-th octave and a peak of 3dB, it is very visible on a frequency response curve and CSD but is not audible. Whereas a 1 dB increase over > 1 octave is hardly visible on a frequency response curve and CSD but is very easy to hear. What can easily be seen in measurement curves is not always what can be easily heard and vice versa.
We probably have all tested the damping of an inaudible but measurable resonance with a series notch filter and experienced that if implemented with a polycarbonate capacitor and ferrite core inductor the sound quality is affected over the whole frequency range, although the reduction of the resonance itself was not noticeable.
The same is true for damping the cone edge resonance: the material used for damping can have an impact on the complete sound reproduction. A resonance should not be 'smoothed and cleaned-up' beyond a peak amplitude whereby it is no longer audible.
If I look at the frequency response curve of a paper cone driver and can not see the cone edge resonance, I am already worried that more damping has been applied than necessary with more than likely a negative impact on sound reproduction.
But it all depends on which benchmark is the most important in the driver development:
a) best visual curves on paper for a novice (an important market with substantial sales)
a) best sound reproduction, taking into account the properties of voices and music, the acoustics of a room and human hearing.
The intention was to mention the case of a high loss rubber suspension glued to the backside of the cone, with the glue and width of the contact area optimized for the maximum damping of the cone edge resonance. My apologies that this could apparently also be mistaken for front attachment.
This is complicated and would require a few pages and a lot of references and examples if we want to cover the variables and details.
Therefor a simplified version for a single 2nd order resonance with center frequency < 1 kHz from the same directional axis:
If such a resonance has a bandwidth < 1/10-th octave and a peak of 3dB, it is very visible on a frequency response curve and CSD but is not audible. Whereas a 1 dB increase over > 1 octave is hardly visible on a frequency response curve and CSD but is very easy to hear. What can easily be seen in measurement curves is not always what can be easily heard and vice versa.
We probably have all tested the damping of an inaudible but measurable resonance with a series notch filter and experienced that if implemented with a polycarbonate capacitor and ferrite core inductor the sound quality is affected over the whole frequency range, although the reduction of the resonance itself was not noticeable.
The same is true for damping the cone edge resonance: the material used for damping can have an impact on the complete sound reproduction. A resonance should not be 'smoothed and cleaned-up' beyond a peak amplitude whereby it is no longer audible.
If I look at the frequency response curve of a paper cone driver and can not see the cone edge resonance, I am already worried that more damping has been applied than necessary with more than likely a negative impact on sound reproduction.
But it all depends on which benchmark is the most important in the driver development:
a) best visual curves on paper for a novice (an important market with substantial sales)
a) best sound reproduction, taking into account the properties of voices and music, the acoustics of a room and human hearing.