Last post I mentioned how pleased I was with Amarra Symphony media player.
Symphony also contains a room/speaker digital correction system, licenced from Dirac.
I guess that is why Symphony is so expensive as I presume they pay licence fees to Dirac.
I trialed Dirac, and while I liked what it did, I didn't like its in-built processor, which I thought stifled dynamics.
I thought maybe it could be a lot better combined with a good player like Amarra which is why I gave it a go.
I reported in my last post how impressed I was with the sound of Amarra in its own right, and now I can report on the integration of the room/speaker correction function , named IRC in Symphony.
It took me quite a while to figure out how to get everything functioning properly- in essence you measure the room, configure a "target" filter, which in theory fixes all your problems , apply the filter to the Amarra player and away you go.
Of course its not that simple, as choosing the right target for the room/speaker and, most importantly , your taste, takes quite a bit of trial and error.
The default suggested target has a slight downward leaning slope towards the high frequency. "Flat" response is actually not desirable at all, tending to sound a bit peaky in most systems.
I used the system first on my Lipinski studio monitors , with active subs underneath, and they actually measured very well. Indeed the suggested target was not much different from the actual measurement, and it was no surprise that with IRC switched in there was not much difference , and not worth using.
Next I tried IRC on my stacked AR-LST vintage speakers. LST stands for laboratory standard transducer , and back in the 70's and 80's they were regarded as highly accurate monitors.
They didn't measure quite as well as the Lipinski's, but still very respectable and would put most modern speakers to shame. The target filter for the AR's had a bit more correcting to do, but i couldn't find the happy medium of correction and retaining the sound I've come to enjoy with the AR's. They are driven by six tube amps, and are something I listen to for enjoyment, rather than critical listening. I could make some fine corrections with the IRC, but overall I preferred the sound unaltered.
At this stage I was ready to give up on Symphony with IRC and just buy the standard Amarra without IRC.
A friend mentioned to me that he had heard an electrostatic set up that used an OTL direct drive set up, and it was the best system he had ever heard, a big call, but I trust this guys ears, and it got me thinking to build an OTL direct drive amp for my own stats.
So I pulled the stats out from storage and set them back up to listen to while I pondered on an OTL design.
I have to say I was quite disappointed with the sound as I've gotten used to the very precise imaging and accuracy of my monitor speakers. Sure electrostatics are clear as a bell, have a big soundstage, but there is always that "electrostatic" sound, the midrange sounds just a little light , male vocals need just a bit more "weight", in comparism to a moving coil speaker. The light weight of the stat speaker certainly gives extraordinary speed and detail, and I guess like all speakers you have to acknowledge they have strengths and weakness's. And if the stat is full range, bass is compromised, and if it is a hybrid, its difficult to integrate the bass driver with the stat panel.
But before I put them away I thought I'd see how they measured in the room with Symphony. Now I suspect it has a lot to do with how stats load a room, but the measurement was not very good. So this would be a good test to see how well IRC would deal with such a problematic result.
It made a dramatic difference, in fact it was a totally different speaker, much better integration, better bass, better mids, better treble, better everything!
This was with the default target, and after a few hours listening I decided to try a different filter, as I was now beginning to better understand the relationship between target and result.
After a couple of hours I had it pretty good, and I was very impressed with my new non- electrostatic sounding speakers!
Further substantial gains were made over a couple of days using the DEQX parametric equaliser to fine tune the final sound to taste . This is very easy to do with the DEQX as one can sit in the listening position and adjust the EQ in real time.
The sound now is nothing short of very impressive, I could rabbit on about it, but I want to finish this blog so I can go and listen!
The interesting thing is that IRC really didn't do very much for accurate speakers, and yet it transformed something that I now realize was incompetent , into something that is far better in most ways than the "competent" speakers, it's certainly more interesting and enjoyable to listen to.
Perhaps conventional, accurate, well designed moving coil loudspeakers are working close to their optimal design and there is little to be gained from digital correction, apart from room influences.
It is a fact that many of the finest regarded speakers are quirky in nature- electrostatics, horn speakers, ribbons, etc - all have many enthusiasts, because they do most things well, but the same enthusiasts would admit that they have weakness's. Digital sound processing (DSP) can possibly make them near perfect, perhaps better than any "conventional" speaker. I certainly believe that electrostatic speakers benefit greatly from DSP, in my experience.
I believe DSP is the future of hi-fi, I'm an old tube guy, and I still believe there will always be a place for a tube pre-amp in any system, but DSP will make tremendous inroads into our listening experience and , importantly, pleasure.
I cant wait to see what the next few years will bring.
Like everything, it takes patience and some understanding to get the best from a system such as Symphony. Its important to understand that you still have to achieve what is subjectively "best" for your ears. There is no "correct" target , and there is no easy solution. But when it works it is sublime.
Tuesday, October 29, 2013
Wednesday, October 9, 2013
Amarra Symphony and DHT tubes.
Amarra Symphony version 2.6 is now easily the best Computer Audio software media player.
I discovered it accidentally while researching DSP systems, and was blown over by the huge improvement in sound quality over its competitors - Audirvana Plus, JRiver etc.
I'm not sure exactly what they are doing that makes it so clearly superior, but it is by far the best digital audio I've ever heard.
"Analoge sounding" is a tired quote, but it is very apt for the sound of Symphony.
If you have ever heard a master tape on a studio quality reel to reel machine you would know that it is the most "direct" sound you can hear. Amarra Symphony has much in common with that level of sound quality.
Some of my best recordings are from the 1960's - tube microphones, tube reel to reel recording, the dynamics are startling. In many ways we have gone backwards in recording technology.
However , digital is getting better and better, and I'm sure it wont be long before those old masters are surpassed.
Of course these days everything is recorded to computer files, and then released as digital files, or burnt to a physical format such as CD or SACD.
Of all the ways to play back these files, I'm convinced that Amarra Symphony is easily the highest quality playback device currently available.
Forget sample rates, bit depths, all the techno babble, Symphony plays music!
It has faults, it's library file management is a long way behind the ease and elegance of JRiver, and its not exactly cost effective, as it is expensive.
However, I think it is worth its cost. I saw this on the Computer Audio forum as a comment on the sound quality of Symphony and I quite liked the analogy:
I hesitate because I've only had a few hours with Symphony, but I opened that can of worms now didn't I. The change is hard to described so I will resort to an analogy (surprise). The experience is akin to taking in a gorgeous view of the ocean from the comfort of your air-conditioned room. Through a perfectly clear floor to ceiling window, it's beautiful. However, when you find yourself standing on the other side holding on to the railing, it's breathtaking. The sun on your face, the heat, the breeze, the smell & roar of the ocean.. a somewhat different experience than that lovely, plush, climate-controlled room.
Amarra paints a beautiful sonic picture, one which has held me captive since first listen. The recent update to 2.6 is fabulous & I truly wondered what more could I ask for my humble system. I sincerely doubt I would have upgraded if it weren't for the new iRC component now offered in Symphony. The SQ is beyond what I expected and for me the change was immediate and palpable. I can only imagine what room correction will add. I can't wait.
Symphony will really make a substantial difference to you audio enjoyment- get it!
So what has this got to do with Directly Heated Triode (DHT) tubes?
After nearly 35 years of building with DHT tubes I know most of the tricks to get them sounding as required. Most DHT preamps, indeed most tube preamps, tend to lean towards the "tube" sound character of warm and romance. And that is what I was doing myself when I released the first commercial 6SN7/300B preamp many years ago.
Over time though it became increasingly more neutral and accurate as better designs and tube loading methods were used.
Although I dont make a 300B preamp now they are still very good sounding , if a little bloomy through the bass regions.
The 300B was replaced with the 101D, and this tube could best be described as "analytical" with very good sound-staging and imaging properties, something I came to consider just as important as the "tone" of a preamp.
My preamps were now a long way from the "warm and romantic" sound of most contemporary tube preamps, and I think that was the main reason for their success.
They still had the "magic" of tubes, that vividness and directness , but were also very accurate and full of resolution.
I used the 101D tube because it was , in my opinion, the only new production tube available that was reliable, available and suitable for preamp use.
Of course there are many other DHT tubes suitable for preamplifiers, and I've tried all of them over the years- the NOS tubes from the 1930's on can be superb (the 71A is a wonderful preamp tube) but they are very hard to find in matched and unused condition.
Lately, in a quest for the ultimate balance in tone, and the 3D effects I value so highly I've turned to Russian DHT tubes. While these are also mostly NOS they are available in quantity, and most importantly are unused and of very high construction and quality.
They've allowed me to build preamps that are of even higher resolution and accuracy, perhaps in terms of absolute performance the very best tube preamps ever built .
(I'm not the only person using these tubes, there are a handful of designers around the world having fun with the Soviet tubes)
But now we have an enigma - you certainly could not call these preamps "warm and romantic" and I suspect many tube lovers would find the sound a little challenging.
Its possible, and relatively easy to "warm" up any design , and some tubes have a tendency to sound slightly warmer - given time I can build a preamp EXACTLY to a customers taste.
But you know what? A preamp that is coloured eventually becomes boring.
A coloured preamp can be very "enjoyable" , but when a really well constructed recording is played through a system that is composed of a collection of equipment that is working optimally , in synergy, at high levels of technical proficiency , meaning accuracy and resolution at the nth degree, then the truth of the recording comes shining through, the musical soul and energy is there for you to hear and be a part of.
You cant get higher than the truth.
This is the way that Supratek preamps have evolved, so if you want a warm "tubey" sounding preamp I can build it for you, and I'm happy to do it.
But , using a superb digital front end like Amarra Symphony, an accurate DHT preamp, with similarly accurate power amps and speakers , you will never be wondering what you may be missing out on.
I discovered it accidentally while researching DSP systems, and was blown over by the huge improvement in sound quality over its competitors - Audirvana Plus, JRiver etc.
I'm not sure exactly what they are doing that makes it so clearly superior, but it is by far the best digital audio I've ever heard.
"Analoge sounding" is a tired quote, but it is very apt for the sound of Symphony.
If you have ever heard a master tape on a studio quality reel to reel machine you would know that it is the most "direct" sound you can hear. Amarra Symphony has much in common with that level of sound quality.
Some of my best recordings are from the 1960's - tube microphones, tube reel to reel recording, the dynamics are startling. In many ways we have gone backwards in recording technology.
However , digital is getting better and better, and I'm sure it wont be long before those old masters are surpassed.
Of course these days everything is recorded to computer files, and then released as digital files, or burnt to a physical format such as CD or SACD.
Of all the ways to play back these files, I'm convinced that Amarra Symphony is easily the highest quality playback device currently available.
Forget sample rates, bit depths, all the techno babble, Symphony plays music!
It has faults, it's library file management is a long way behind the ease and elegance of JRiver, and its not exactly cost effective, as it is expensive.
However, I think it is worth its cost. I saw this on the Computer Audio forum as a comment on the sound quality of Symphony and I quite liked the analogy:
I hesitate because I've only had a few hours with Symphony, but I opened that can of worms now didn't I. The change is hard to described so I will resort to an analogy (surprise). The experience is akin to taking in a gorgeous view of the ocean from the comfort of your air-conditioned room. Through a perfectly clear floor to ceiling window, it's beautiful. However, when you find yourself standing on the other side holding on to the railing, it's breathtaking. The sun on your face, the heat, the breeze, the smell & roar of the ocean.. a somewhat different experience than that lovely, plush, climate-controlled room.
Amarra paints a beautiful sonic picture, one which has held me captive since first listen. The recent update to 2.6 is fabulous & I truly wondered what more could I ask for my humble system. I sincerely doubt I would have upgraded if it weren't for the new iRC component now offered in Symphony. The SQ is beyond what I expected and for me the change was immediate and palpable. I can only imagine what room correction will add. I can't wait.
Symphony will really make a substantial difference to you audio enjoyment- get it!
So what has this got to do with Directly Heated Triode (DHT) tubes?
After nearly 35 years of building with DHT tubes I know most of the tricks to get them sounding as required. Most DHT preamps, indeed most tube preamps, tend to lean towards the "tube" sound character of warm and romance. And that is what I was doing myself when I released the first commercial 6SN7/300B preamp many years ago.
Over time though it became increasingly more neutral and accurate as better designs and tube loading methods were used.
Although I dont make a 300B preamp now they are still very good sounding , if a little bloomy through the bass regions.
The 300B was replaced with the 101D, and this tube could best be described as "analytical" with very good sound-staging and imaging properties, something I came to consider just as important as the "tone" of a preamp.
My preamps were now a long way from the "warm and romantic" sound of most contemporary tube preamps, and I think that was the main reason for their success.
They still had the "magic" of tubes, that vividness and directness , but were also very accurate and full of resolution.
I used the 101D tube because it was , in my opinion, the only new production tube available that was reliable, available and suitable for preamp use.
Of course there are many other DHT tubes suitable for preamplifiers, and I've tried all of them over the years- the NOS tubes from the 1930's on can be superb (the 71A is a wonderful preamp tube) but they are very hard to find in matched and unused condition.
Lately, in a quest for the ultimate balance in tone, and the 3D effects I value so highly I've turned to Russian DHT tubes. While these are also mostly NOS they are available in quantity, and most importantly are unused and of very high construction and quality.
They've allowed me to build preamps that are of even higher resolution and accuracy, perhaps in terms of absolute performance the very best tube preamps ever built .
(I'm not the only person using these tubes, there are a handful of designers around the world having fun with the Soviet tubes)
But now we have an enigma - you certainly could not call these preamps "warm and romantic" and I suspect many tube lovers would find the sound a little challenging.
Its possible, and relatively easy to "warm" up any design , and some tubes have a tendency to sound slightly warmer - given time I can build a preamp EXACTLY to a customers taste.
But you know what? A preamp that is coloured eventually becomes boring.
A coloured preamp can be very "enjoyable" , but when a really well constructed recording is played through a system that is composed of a collection of equipment that is working optimally , in synergy, at high levels of technical proficiency , meaning accuracy and resolution at the nth degree, then the truth of the recording comes shining through, the musical soul and energy is there for you to hear and be a part of.
You cant get higher than the truth.
This is the way that Supratek preamps have evolved, so if you want a warm "tubey" sounding preamp I can build it for you, and I'm happy to do it.
But , using a superb digital front end like Amarra Symphony, an accurate DHT preamp, with similarly accurate power amps and speakers , you will never be wondering what you may be missing out on.
Friday, August 9, 2013
New products! DHT preamps.
Been a while since my last post. Been very busy the last 6 months but still found time to do rainy day R&D and proud to present new products, and the demise of a couple of long standing old faithfuls , which is a bit sad, but the new designs are very exciting. Let me tell you about them......
Supratek was possibly the first audio company to release a Directly Heated Triode preamplifier, the first Cabernet had an option to use either a 300B, PX4 or 45 DHT tube .
Later on I went to a 101D DHT tube in the Dual Cabernet.
The Dual Cabernet was also a revolutionary product as it was the first preamp to provide 2 seperate output stages in the one chassis.
One stereo output was a 6SN7/6H30 line stage with an output impedance of 50 ohms, perfect for driving difficult solid state amps and especially useful for powering active sub-woofers. With its own seperate volume control it was the ideal device for integrating a sub woofer into a stereo system.
The main stereo preamp output used a 101D DHT tube , which is an excellent tube to use in combination with the 6SN7/6H30 , as freed of sub-sonic bass requirements , the 101D is ideal for upper bass to high frequency operation, with the wonderful ease and dynamics of DHT tubes.
There are very few new production DHT tubes suitable for pre-amplification use, the 101D is probably the best all round new production DHT preamplifier tube as it has perhaps the best sound balance of the candidates available.
Which is why other tube audio companies have started using it in the last few years.
Of course there have been many other DHT tubes available since the 1930's and my favorite tube to use in a preamp has for a long time been the 71A tube.
I've been using it for many years in my own preamps, I wish I could have offered it as an option for the Cabernet but unfortunately it hasn't been manufactured for a very long time and it is very hard to get matched pairs in pristine condition, it's taken me years to assemble a collection just for my own use.
During 2011-2013 I started using obscure NOS Russian DHT tubes in my preamps and am excited to have found some tubes that equal the sound quality of the best 1930's DHT tubes.
These are available in limited quantities, are exceptionally well built and sound fantastic!
I've purchased enough stock to build preamps and supply spare tubes, and these tubes have been incorporated into new Supratek designs that replace the circuits in the Cabernet and Cortese preamps.
I'm very pleased to be able to use a DHT tube in the Cortese phono preamp. The high quality Single Ended Class A phono stage, designed for moving coil cartridges, really deserves the best line stage. The fundamental requirements of a line stage are: a suitable gain structure and low output impedance , in order to be able to drive any power amplifier. Input impedance also plays a part.
The 6SN7 line stage used in the Cortese was an innovative and very good sounding design that ticked all the boxes, and although I was very happy with it, I would have loved to use a DHT tube. Unfortunately, there just wasn't a suitable DHT tube available to work within the confines of space and engineering in the Cortese.
Now there is!
The Russian 4P1L DHT tube is ideal as a line stage for the Cortese. Of course it has the usual state-of-the-art design which loads the tube optimally and it incorporates the feature of being able to be used as capacitor or transformer coupled output.
This gives the user a choice of gain and impedance structure, and two different types of "sound" .
Personally I feel both choices are extremely good sounding , and it is just a matter of what is best choice for your system.
The Cabernets are designed solely for line stage use and the engineering possibilities for DHT use are quite exciting.
As noted above the obvious choice for present production pre-amplification DHT tubes is the 101D.
In 2012 the design of the Cabernets changed from directly coupling the 6SN7 driver tube to the 101D, in favour of capacitor coupling , in combination with grid chokes , and the use of exotic Lundahl output transformers.
This form of circuit enables a different type of bias which we label LowR , the primary advantage of which is to negate the need for large capacitance decoupling and bypass caps, the source of "veiling" in many tube preamps.
This greatly enhanced what I consider most important in pre-amplification- the ability to throw a realistic soundstage with realistic 3D imaging, (depth and width), with dynamics that show greater expression of micro and macro layers.
Vividness is a term I've been using to describe the Supratek sound for many years, and with the new DHT designs it is again enhanced.
It has been very exciting to me to be able to source DHT tubes with the quality and sound characteristics of the DHT tubes produced in the "golden age" of tube manufacture and design , the period from 1920- 1960.
While the Western electronics industry veered towards low cost, high production solid state technology that promised high profits, the Russian tube manufacturers refined and developed very high quality tube products that matched and exceeded the best tubes available from the West.
Of all the Russian DHT tubes I've tested and experimented with over the last few years my favourites are the 4P1L and 6P21S, which are both multi-grid low power tubes, however, connected as triodes they are superb preamp tubes with low impedance and ideal gain. To find a tube that sounds as good as the best 71A is very satisfying.
The Cabernet Dual is the premier preamp in the Supratek line up. The 6SN7/6H30 output stays the same , but the 101D tube is replaced by the 6P21S for a fuller sound that seems more musical, yet as detailed and of higher resolution than the 101D.
The 6P21S has a greater range then the 101D, the bass of the 6P21S is deep and tight, yet it also has exceptional mid and treble sound.
Only the highest quality parts and components are used- Lundahl and Magnequest transformers and grid chokes. Capacitors are paper-in-oil, chosen to compliment the sonic signature of the preamp.
The tremendous advantage of the Cabernets shunt regulated power supply remains, and the overkill size of power transformers means the Cabernet electronics perform quietly and effortlessly .
The chassis's are still fabricated from 1.2mm solid copper, far more important than it seems, and covered with high quality chrome or powder coat.
The Cabernet Single is exactly the same without the 6H30 driver output.
(BTW the The 6H30 driver in the Dual Cabernet can also be used as a valve output stage for suitable dacs)
Supratek hasn't been a "conventional" retail audio manufacturer for a long time, I dont depend on it for income and the reason for Supratek's existence is more about a passion for high end tube electronics than a commercial venture.
The (small) profit from preamp sales goes back into R&D- for every 3 or 4 preamps sold there would be at least one prototype built to investigate a new concept or idea.
I have a lifetime's experience of DHT use for preamps, and my passion for DHT preamps is stronger than ever before, even with so called claims for the demise of preamps as they supposedly are not needed with digital devices.
This is rubbish and reflects poor equipment and listening skills, or perhaps the claimants have just never heard a good tube preamp. It is true that good tube preamps are rare, but a good tube preamp will be the heart and soul of a system that takes you to the essence of the music.
Because the Supratek preamps are completely handmade, and it is not a commercial profit making venture , very few are actually built.
I've found it impossible to supply worldwide demand and as a consequence the Chardonnay, Sauvignon, and Grange preamps will no longer be built.
The Chenin will continue for a little while longer and then I will be concentrating on only building Cabernet and Cortese DHT preamps, as I am available.
The reliability and very long life time of the Cabernet and Cortese is something I take very seriously. New models are not regularly introduced and there is a great deal of thought and commitment to longevity. A classic tube product should last 50 years and more.
Directly Heated Triode Preamps are the pinnacle of preamp design and I hope to be able to continue to enrich the musical lives of future owners.
Supratek was possibly the first audio company to release a Directly Heated Triode preamplifier, the first Cabernet had an option to use either a 300B, PX4 or 45 DHT tube .
Later on I went to a 101D DHT tube in the Dual Cabernet.
The Dual Cabernet was also a revolutionary product as it was the first preamp to provide 2 seperate output stages in the one chassis.
One stereo output was a 6SN7/6H30 line stage with an output impedance of 50 ohms, perfect for driving difficult solid state amps and especially useful for powering active sub-woofers. With its own seperate volume control it was the ideal device for integrating a sub woofer into a stereo system.
The main stereo preamp output used a 101D DHT tube , which is an excellent tube to use in combination with the 6SN7/6H30 , as freed of sub-sonic bass requirements , the 101D is ideal for upper bass to high frequency operation, with the wonderful ease and dynamics of DHT tubes.
There are very few new production DHT tubes suitable for pre-amplification use, the 101D is probably the best all round new production DHT preamplifier tube as it has perhaps the best sound balance of the candidates available.
Which is why other tube audio companies have started using it in the last few years.
Of course there have been many other DHT tubes available since the 1930's and my favorite tube to use in a preamp has for a long time been the 71A tube.
I've been using it for many years in my own preamps, I wish I could have offered it as an option for the Cabernet but unfortunately it hasn't been manufactured for a very long time and it is very hard to get matched pairs in pristine condition, it's taken me years to assemble a collection just for my own use.
During 2011-2013 I started using obscure NOS Russian DHT tubes in my preamps and am excited to have found some tubes that equal the sound quality of the best 1930's DHT tubes.
These are available in limited quantities, are exceptionally well built and sound fantastic!
I've purchased enough stock to build preamps and supply spare tubes, and these tubes have been incorporated into new Supratek designs that replace the circuits in the Cabernet and Cortese preamps.
I'm very pleased to be able to use a DHT tube in the Cortese phono preamp. The high quality Single Ended Class A phono stage, designed for moving coil cartridges, really deserves the best line stage. The fundamental requirements of a line stage are: a suitable gain structure and low output impedance , in order to be able to drive any power amplifier. Input impedance also plays a part.
The 6SN7 line stage used in the Cortese was an innovative and very good sounding design that ticked all the boxes, and although I was very happy with it, I would have loved to use a DHT tube. Unfortunately, there just wasn't a suitable DHT tube available to work within the confines of space and engineering in the Cortese.
Now there is!
The Russian 4P1L DHT tube is ideal as a line stage for the Cortese. Of course it has the usual state-of-the-art design which loads the tube optimally and it incorporates the feature of being able to be used as capacitor or transformer coupled output.
This gives the user a choice of gain and impedance structure, and two different types of "sound" .
Personally I feel both choices are extremely good sounding , and it is just a matter of what is best choice for your system.
DHT Cortese 4P1L tube at back
The Cabernets are designed solely for line stage use and the engineering possibilities for DHT use are quite exciting.
As noted above the obvious choice for present production pre-amplification DHT tubes is the 101D.
In 2012 the design of the Cabernets changed from directly coupling the 6SN7 driver tube to the 101D, in favour of capacitor coupling , in combination with grid chokes , and the use of exotic Lundahl output transformers.
This form of circuit enables a different type of bias which we label LowR , the primary advantage of which is to negate the need for large capacitance decoupling and bypass caps, the source of "veiling" in many tube preamps.
This greatly enhanced what I consider most important in pre-amplification- the ability to throw a realistic soundstage with realistic 3D imaging, (depth and width), with dynamics that show greater expression of micro and macro layers.
Vividness is a term I've been using to describe the Supratek sound for many years, and with the new DHT designs it is again enhanced.
It has been very exciting to me to be able to source DHT tubes with the quality and sound characteristics of the DHT tubes produced in the "golden age" of tube manufacture and design , the period from 1920- 1960.
While the Western electronics industry veered towards low cost, high production solid state technology that promised high profits, the Russian tube manufacturers refined and developed very high quality tube products that matched and exceeded the best tubes available from the West.
Of all the Russian DHT tubes I've tested and experimented with over the last few years my favourites are the 4P1L and 6P21S, which are both multi-grid low power tubes, however, connected as triodes they are superb preamp tubes with low impedance and ideal gain. To find a tube that sounds as good as the best 71A is very satisfying.
The Cabernet Dual is the premier preamp in the Supratek line up. The 6SN7/6H30 output stays the same , but the 101D tube is replaced by the 6P21S for a fuller sound that seems more musical, yet as detailed and of higher resolution than the 101D.
The 6P21S has a greater range then the 101D, the bass of the 6P21S is deep and tight, yet it also has exceptional mid and treble sound.
Only the highest quality parts and components are used- Lundahl and Magnequest transformers and grid chokes. Capacitors are paper-in-oil, chosen to compliment the sonic signature of the preamp.
The tremendous advantage of the Cabernets shunt regulated power supply remains, and the overkill size of power transformers means the Cabernet electronics perform quietly and effortlessly .
The chassis's are still fabricated from 1.2mm solid copper, far more important than it seems, and covered with high quality chrome or powder coat.
All Supratek preamps, and power supplies, are made from 1.2mm copper.
The Cabernet Single is exactly the same without the 6H30 driver output.
(BTW the The 6H30 driver in the Dual Cabernet can also be used as a valve output stage for suitable dacs)
Supratek hasn't been a "conventional" retail audio manufacturer for a long time, I dont depend on it for income and the reason for Supratek's existence is more about a passion for high end tube electronics than a commercial venture.
The (small) profit from preamp sales goes back into R&D- for every 3 or 4 preamps sold there would be at least one prototype built to investigate a new concept or idea.
I have a lifetime's experience of DHT use for preamps, and my passion for DHT preamps is stronger than ever before, even with so called claims for the demise of preamps as they supposedly are not needed with digital devices.
This is rubbish and reflects poor equipment and listening skills, or perhaps the claimants have just never heard a good tube preamp. It is true that good tube preamps are rare, but a good tube preamp will be the heart and soul of a system that takes you to the essence of the music.
Because the Supratek preamps are completely handmade, and it is not a commercial profit making venture , very few are actually built.
I've found it impossible to supply worldwide demand and as a consequence the Chardonnay, Sauvignon, and Grange preamps will no longer be built.
The Chenin will continue for a little while longer and then I will be concentrating on only building Cabernet and Cortese DHT preamps, as I am available.
The reliability and very long life time of the Cabernet and Cortese is something I take very seriously. New models are not regularly introduced and there is a great deal of thought and commitment to longevity. A classic tube product should last 50 years and more.
Directly Heated Triode Preamps are the pinnacle of preamp design and I hope to be able to continue to enrich the musical lives of future owners.
Thursday, May 9, 2013
Update Sonic Panel Electrostatics
http://supratekaudio.blogspot.com.au/2012/11/an-inexpensive-top-quality.html
They're really quite capable as singles but stacking two pairs gives even greater sensitivity and a larger dispersion, which gives a bigger , more open sound.
I had some speaker frames made up by a furniture craftsman (John Thompson Designs in Bussellton) to give some weight and solidity to the set up.
The sub/woofer remains the same, an Acoustic Elegance 18" driver which has the ability to remain linear up to well over the crossover point of 400Hz.
I chose this higher point to give a bit more solidity to the crossover region as it did sound a little weaker there with the extra sensitivity of the stats.
Amps remain solid state high power, pre is tube (of course!) active crossover is either digital or a op-amp version I'm experimenting with.
Once I found the optimal crossover points I'll build a tube crossover .
One new feature is the use of a parametric equaliser program incorporated into the computer audio program I use for playback.
Fabfilter Pro-Q used with JRiver Media Centre 18 is really exceptional- easy to use, very versatile, and operating at 64 bit , it sounds very good.
(For Mac it can be used with PureMusic)
I prefer it to the sound of the parametric functions in the digital crossovers I use (DEQX and DBX PA) and its much quicker and easier to use.
Heres a screen shot of one of the settings I used to determine EQ for the large workshop the speakers are presently in. Final settings will be not as dramatic, but a very powerful tool.
The sound of the stacked stats is very good, they have a real ability for a big sound, and for their size have remarkable dynamics and power.
Active operation and EQ is the best option to get them at their best, I would imagine it would be very time consuming to build passive crossovers to suit.
They are too good for my workshop, but enjoyable to listen to while working.
Wednesday, May 8, 2013
"Euphonic" Tubes
There is a perception widely held that tubes have a more "euphonic" sound than solid state devices, that they sound "warmer, lusher, more romantic".
Its a common theme on the audio forums and you see this perception of "euphonics" very frequently, and often from people who have been in audio for a long time.
"Use a tube buffer to soften the sound" or "use a XXXX type of tube for a lovely lush tone".
This type of perception is very misguided and completely wrong.
Let me explain why.
Firstly the term buffer can be applied to either tubes or solid state. The term is correctly used as a non-amplifying device (no gain) which couples two devices so they can work optimally together. If you have an amplifier with a lowish input impedance and a source with highish output impedance , combined together they will probably have a degraded frequency response.
You need a "buffer" to couple these devices together, something that will present an agreeable load to the source and the amplifier.
The most common devices are opamps, usually used on the outputs of CDP, DACs, etc.
The problem with opamps are they are 100% feedback devices, they do a very good job technically , but they sure leave a lot to be desired in respect to sound quality.
Good designers often revert to discrete designs that require much less levels of feedback.
(There are design methods and types of opamps that sound good, but rare)
Or some people say "use a tube buffer- much better than opamps".
A tube buffer that has no gain is usually a design known as a cathode follower, it has just less than a gain of 1 (unity gain) and has high input impedance and low output impedance which is ideal, but like an an opamp it has 100% degenerative feedback.
It is possible to make a tube buffer sound ok, but all the tube buffers tacked on to the back of dacs , preamps etc are no better than the opamps they are supposed to improve on, and in many cases are inferior.
Its easy to tack a tube on the back of a solid state design, but many solid state designers do not understand tube design, and follow a tired formula to get some tube "credibility".
It is quite common to see many tube buffers used at far lower voltages than they need for optimum operating conditions, and the tube is struggling to do its job.
The cathode follower design is commonly used because it allows a tube that can not normally be used for buffering to have a low enough output impedance to do the job.
But remember, like an opamp it is a 100% feedback device.
What determines how well a tube can drive a following stage is its Plate Resistance (Pr)
If a tube has a high plate resistance, and used in a zero feedback design (nice) it will have a high output resistance (Z).
Tubes like the 12AX7, 6SL7, 12AT7, etc have Pr in the range 50-100Kohm and also have very high gain.
They cant be used as buffers as they have too much gain, and no drive ability. They are high gain devices very much limited to specific use.
These devices when used as amplifying/buffering devices really do have the "warm, lush and romantic" sound of tubes, because they are not operating as they were designed for.
Tubes with medium Pr like the 6SN7, 12AU7, 12BH7, etc have Pr around 5K ohms, and are better at driving loads. You could use one of these tubes to drive a reasonable load of 100Kohms, or higher, but not much lower, they struggle with power amps of lowish input Z . Typically they have medium gain, usually too much to use as non-amplifying buffers, and they are not ideal as buffers anyway.
Depending on the situation the sound can have varying degrees of "euphonics".
An ideal input/output ratio is a minimum of 10 . A preamp with output Z of 10K will be just ok with a power amp's input Z of 100K, although the lower the better.
The higher the ratio the better the performance, and generally speaking, the accuracy will increase and tube "euphonics" will be reduced.
Tubes such as the 12B4A, 6H30, D3a, etc have low Pr around 1-2Kohms and are ideal as interfaces between devices. They have extremely low distortion, provide accurate driving conditions and are ideal used as buffers, although depending on the tube and operating points there will be gain.
Some of the DHT tubes are also ideal in this situation, although a bit more finicky to use.
Some really good tubes were designed at the end of the tube era, high gain and low Pr was a difficult, almost impossible task in the early 19th century but by the 60/70's some excellent tubes were designed- the D3a, E810F, E280F and many others had very high gain, and also low Pr, really the perfect amplifying device.
It should be mentioned that the operating points of all tubes is critical to performance, the level of harmonic distortion plays a major part in any tubes sound, for good and bad, and the designer needs to adjust for the performance required.
Many "lush" circuits are the direct result of improper operating points, sometimes done deliberately, eg guitar amps, SE amplifiers.
It should also be noted that feedback is not necessarily a "bad" thing, feedback that is used to fix design or component faults is always counter-productive, but used sparingly and wisely , it can be a powerful and good sounding tool.
Feedback can be especially useful with high and medium Pr tubes as it lowers the Pr and output Z of these tubes.
Naturally I like tubes in my preamps and as the devices on the output stages of my dacs, I think for dac outputs the low Pr tubes are very much ideal and have very much better sound than any opamp or other discrete device (although I do have a fondness for j-fets).
There are no "euphonics" with the low Pr tubes if used properly- they have very low levels of distortion, are accurate, clean and powerful sounding , very much broadcasting the fact that they are completely zero-feedback, which leads to very good phase performance, and that can be heard in excellent sound-staging and imaging.
But how to get the unity gain that is required in this situation without using a 100% feedback "tube buffer" (also referred to as a goats turd) when I typically use a low Pr tube with a gain of up to 50?
Fortunately the dac often needs some amplification, but the best option is to use a step-down transformer which brings the gain down to useable levels, and lowers the output Z even further.
It is imperative to use a high-bandwidth device that is matched to the tube used, and these devices are expensive. I use Magnequest, Lundahl and Jensen transformers, all imported.
There are different methods of loading the plate resistance of the tube, and many different methods of biasing the tubes.
In fact there are so many variables to tube design, it does take many years of experience to sort the best designs from the less desirable.
But the advantages are very worthwhile and it is very hard to listen to an op-amp buffered dac after living with a good tube output stage that is not a 100% feedback tube buffer.
"Euphonics" has its place , but preferably in a glass of wine, or your choice of ecstacy.
Of course there are many who actually prefer the sound of euphonic systems, actively chasing "warm and lush" sounds.
In todays hi-resolution world, that is perhaps illusionary , but each to his own, and I can quite readily enjoy such a system for a limited time
Here is one of my preamps, used as an example of tube selection and the reasons.
The bigger tubes at the back are 71A Directly Heated Triodes. These are actually very low power tubes, used back in the 30's in radios, but with a low Pr and low gain they are ideal for driving modern power amps. DHT's are finicky to use in terms of noise and minimal components, but there are ways around this. They are used with output transformers so gain is very low, to counter this I use a pair of 6SN7 medium Pr tubes (on the outside) to provide gain. They see an easy load to the 71A's and need a simple non-feedback circuit to drive the 71A's.
The inside tubes are E280F's which are used to provide a separate tube gain stage for one of my dacs. The output from the dac chip is fed straight into the E280F and the output can be switched into the line stage. The E280F is chosen as it is a high gain/low Pr tube. It is transformer coupled and makes an excellent compliment to the digital-to-analouge conversion.
Its a common theme on the audio forums and you see this perception of "euphonics" very frequently, and often from people who have been in audio for a long time.
"Use a tube buffer to soften the sound" or "use a XXXX type of tube for a lovely lush tone".
This type of perception is very misguided and completely wrong.
Let me explain why.
Firstly the term buffer can be applied to either tubes or solid state. The term is correctly used as a non-amplifying device (no gain) which couples two devices so they can work optimally together. If you have an amplifier with a lowish input impedance and a source with highish output impedance , combined together they will probably have a degraded frequency response.
You need a "buffer" to couple these devices together, something that will present an agreeable load to the source and the amplifier.
The most common devices are opamps, usually used on the outputs of CDP, DACs, etc.
The problem with opamps are they are 100% feedback devices, they do a very good job technically , but they sure leave a lot to be desired in respect to sound quality.
Good designers often revert to discrete designs that require much less levels of feedback.
(There are design methods and types of opamps that sound good, but rare)
Or some people say "use a tube buffer- much better than opamps".
A tube buffer that has no gain is usually a design known as a cathode follower, it has just less than a gain of 1 (unity gain) and has high input impedance and low output impedance which is ideal, but like an an opamp it has 100% degenerative feedback.
It is possible to make a tube buffer sound ok, but all the tube buffers tacked on to the back of dacs , preamps etc are no better than the opamps they are supposed to improve on, and in many cases are inferior.
Its easy to tack a tube on the back of a solid state design, but many solid state designers do not understand tube design, and follow a tired formula to get some tube "credibility".
It is quite common to see many tube buffers used at far lower voltages than they need for optimum operating conditions, and the tube is struggling to do its job.
The cathode follower design is commonly used because it allows a tube that can not normally be used for buffering to have a low enough output impedance to do the job.
But remember, like an opamp it is a 100% feedback device.
What determines how well a tube can drive a following stage is its Plate Resistance (Pr)
If a tube has a high plate resistance, and used in a zero feedback design (nice) it will have a high output resistance (Z).
Tubes like the 12AX7, 6SL7, 12AT7, etc have Pr in the range 50-100Kohm and also have very high gain.
They cant be used as buffers as they have too much gain, and no drive ability. They are high gain devices very much limited to specific use.
These devices when used as amplifying/buffering devices really do have the "warm, lush and romantic" sound of tubes, because they are not operating as they were designed for.
Tubes with medium Pr like the 6SN7, 12AU7, 12BH7, etc have Pr around 5K ohms, and are better at driving loads. You could use one of these tubes to drive a reasonable load of 100Kohms, or higher, but not much lower, they struggle with power amps of lowish input Z . Typically they have medium gain, usually too much to use as non-amplifying buffers, and they are not ideal as buffers anyway.
Depending on the situation the sound can have varying degrees of "euphonics".
An ideal input/output ratio is a minimum of 10 . A preamp with output Z of 10K will be just ok with a power amp's input Z of 100K, although the lower the better.
The higher the ratio the better the performance, and generally speaking, the accuracy will increase and tube "euphonics" will be reduced.
Tubes such as the 12B4A, 6H30, D3a, etc have low Pr around 1-2Kohms and are ideal as interfaces between devices. They have extremely low distortion, provide accurate driving conditions and are ideal used as buffers, although depending on the tube and operating points there will be gain.
Some of the DHT tubes are also ideal in this situation, although a bit more finicky to use.
Some really good tubes were designed at the end of the tube era, high gain and low Pr was a difficult, almost impossible task in the early 19th century but by the 60/70's some excellent tubes were designed- the D3a, E810F, E280F and many others had very high gain, and also low Pr, really the perfect amplifying device.
It should be mentioned that the operating points of all tubes is critical to performance, the level of harmonic distortion plays a major part in any tubes sound, for good and bad, and the designer needs to adjust for the performance required.
Many "lush" circuits are the direct result of improper operating points, sometimes done deliberately, eg guitar amps, SE amplifiers.
It should also be noted that feedback is not necessarily a "bad" thing, feedback that is used to fix design or component faults is always counter-productive, but used sparingly and wisely , it can be a powerful and good sounding tool.
Feedback can be especially useful with high and medium Pr tubes as it lowers the Pr and output Z of these tubes.
Naturally I like tubes in my preamps and as the devices on the output stages of my dacs, I think for dac outputs the low Pr tubes are very much ideal and have very much better sound than any opamp or other discrete device (although I do have a fondness for j-fets).
There are no "euphonics" with the low Pr tubes if used properly- they have very low levels of distortion, are accurate, clean and powerful sounding , very much broadcasting the fact that they are completely zero-feedback, which leads to very good phase performance, and that can be heard in excellent sound-staging and imaging.
But how to get the unity gain that is required in this situation without using a 100% feedback "tube buffer" (also referred to as a goats turd) when I typically use a low Pr tube with a gain of up to 50?
Fortunately the dac often needs some amplification, but the best option is to use a step-down transformer which brings the gain down to useable levels, and lowers the output Z even further.
It is imperative to use a high-bandwidth device that is matched to the tube used, and these devices are expensive. I use Magnequest, Lundahl and Jensen transformers, all imported.
There are different methods of loading the plate resistance of the tube, and many different methods of biasing the tubes.
In fact there are so many variables to tube design, it does take many years of experience to sort the best designs from the less desirable.
But the advantages are very worthwhile and it is very hard to listen to an op-amp buffered dac after living with a good tube output stage that is not a 100% feedback tube buffer.
"Euphonics" has its place , but preferably in a glass of wine, or your choice of ecstacy.
Of course there are many who actually prefer the sound of euphonic systems, actively chasing "warm and lush" sounds.
In todays hi-resolution world, that is perhaps illusionary , but each to his own, and I can quite readily enjoy such a system for a limited time
Here is one of my preamps, used as an example of tube selection and the reasons.
The bigger tubes at the back are 71A Directly Heated Triodes. These are actually very low power tubes, used back in the 30's in radios, but with a low Pr and low gain they are ideal for driving modern power amps. DHT's are finicky to use in terms of noise and minimal components, but there are ways around this. They are used with output transformers so gain is very low, to counter this I use a pair of 6SN7 medium Pr tubes (on the outside) to provide gain. They see an easy load to the 71A's and need a simple non-feedback circuit to drive the 71A's.
The inside tubes are E280F's which are used to provide a separate tube gain stage for one of my dacs. The output from the dac chip is fed straight into the E280F and the output can be switched into the line stage. The E280F is chosen as it is a high gain/low Pr tube. It is transformer coupled and makes an excellent compliment to the digital-to-analouge conversion.
Sunday, April 28, 2013
AR-LST speakers "activated"
I've sold most along the way, but some of the real classics are of such quality I've kept them in my store.
Around 2008 I bought 2 pairs of the Acoustic Research AR-LST from a little old lady's son after she become too old to manage them. I don't know why she needed 2 pairs, but they were in perfect condition and I paid $4000 for them, which was quite a lot for speakers made in 1975.
I knew of the formidable reputation the Laboratory Standard Transducers (LST) had and loaded all 160 kgs into my car.
The LST's are an unusual speaker with multi-drivers. One 12" bass driver in a sealed cabinet, 4 of one and a half inch midrange dome drivers and 4 of three quarter of an inch dome tweeters.
The cabinet has two offset side panels which contain the midrange and 2 of the tweeter units.
Kind of like a Bose 901 cabinet, but much bigger and meaner!
With the 2 pairs stacked I have a total of 36 drivers playing. I've always believed that small speakers are a compromise and for realistic sound a speaker has to have the physicality to move lots of air, hence my love of big stats, horns and big moving coil speakers.
The crossovers in the LST are quite interesting, 12dB slopes at 575 and 5000Hz. Nice big air coils and quality caps , I only needed to replace the midrange electrolytic cap for a poly cap.
The speaker had an input transformer with many taps to alter the tonal frequency, to adjust for room response and taste. Unfortunately this required a 2500uF electrolytic cap to counter sub bass resonance and I can't say I was too happy about having this in the signal path.
As I intended to actively tri-amp I had no need for level adjustment and rebuilt the crossovers on the back panel of the cabinet, removing the transformer and hardwiring all the crossover components .
The printed circuit boards also went in the bin.
I also put an extra set of speaker connectors for each set of drivers on back panel so that I could either use the passive crossovers or run the drivers directly from the seperate amps I intended to use for bass, mid and treble, utilizing an active crossover, either analogue or digital.
With this set up I could choose whatever types of amps I wanted to use with either the bass, mid and treble sets of drivers.
The AR-LST's came along at a defining moment in amplifier design- the new high power solid state designs had just replaced the tube designs of the 50's/60's and the LST was a difficult load- it needed the high power and current abilities of the solid state amps, but these new designs typically used a lot of negative feedback and were not the nicest sounding amps. Amps like the Crown DC300 were commonly used and the results were not in the LST's favour.
Modern day solid state amps fare much better, but part of this restoration was to enable me to use my tube power amps with this speaker.
Dividing the crossover into 3 separate sections- bass, mid and treble makes the amplifiers load much easier and I was positive my tube amps, which use positive current feedback would drive the LST's very well.
The only question was would they also drive two sets of speakers, in parallel and with the passive crossovers presenting a heavy load.
Fortunately it was not a problem, and the amps have no problem at all in driving the LST's cleanly at high levels.
I'm using a 100watt KT88 ultra-linear amp on bass, an 80 watt KT88 triode amp on mids and a 60 watt 7027 pentode amp on treble, all with positive current current.
I've been using current feedback on all my amps for many years now and it is interesting to see positive feedback now becoming used by quite a few well respected tube designers- its a win/win design that enables tube amps to work exceedingly well with just about any difficult speaker.
I've been a fan of digital crossovers for around 5 years, owning DEQX and DBX active digital crossovers, but I must admit I'm enjoying the passive crossover/ tri-amping approach used on the LST's.
My big reference electrostatic system demands solid state amplification and I somewhat miss the use of an all tube system, but with the LST's use of 6 tube power amps and a tube preamp , I am in tube heaven!
Incidentally I also went away from using a digital crossover in my big reference system, and built the speaker crossover function into the tube preamp, which works very well.
I dont think there is anything wrong with digital crossovers, but getting tubes or passive components to do the job if possible makes it simple, and simple is often better.
Although it is obvious that 6 monoblock power amps is not exactly simplistic!
I'm very happy with the AR-LST sound, it has the big dynamic openness I like, and is completely natural and un-restrained sounding. It is the type of sound you can listen to indefinitely and is simply very enjoyable.
Used in this fashion the LST's almost sound like a cross of moving coil, electrostatic and horn speakers- they have something of the good points of all 3, even a touch of ribbon sound to them.
This is the beauty of active amplification- the finest of adjustments can be made to the levels of each crossover function and the system can be precisely fine tuned to suit the owners room and taste.
Subscribe to:
Comments (Atom)