Showing some respect for both the designers and the design is primordial.jacco vermeulen said:
This seeming masterpiece has surely undergone more research, trial and error and cooperative brainwork than we ever can imagine.
The 350khz bandwidth and carefully selected components sure must have been part of the fun and challenge to build the Blowtorch.
I see the end result of tree talented people who have pushed themselves to the limits of their knowledge to achieve this goal. Every single component including the wiring scheme must have a reason to be where, how and why it is there.
Stunning controversial beauty. All too much criticized with envy and the un-will to look and step beyond boundaries.
For a moment, I very much would have liked to be the Cabernet that has been drunk when it played his first notes of music.
Also, its not because a design is discussed here that the author is obliged to release all the info, schematics and building plans to the community.
/Hugo
Hello,
To use DC servo or not ?
A DC servo for many designers limits itself simply to the use of an OP (with some caps and resistances) connected between input and output.
John Curl, him, uses, for his DC servo, various very complex and very cunning configurations of several OP as in PLD1500 and PLD-2000 Parasound preamplifiers. Already, in the 1980s, the schematic of Dennesen JC-80 showed a very particular DC servo with 1 single OP and 1 dual OP .
There is and there will be always, good and bad audio conceptions, and good and bad DC servos.
For me, John Curl, with Nelson Pass, Charles Hansen and some the others, belongs to the elite of the audio designers.
Darry
To use DC servo or not ?
A DC servo for many designers limits itself simply to the use of an OP (with some caps and resistances) connected between input and output.
John Curl, him, uses, for his DC servo, various very complex and very cunning configurations of several OP as in PLD1500 and PLD-2000 Parasound preamplifiers. Already, in the 1980s, the schematic of Dennesen JC-80 showed a very particular DC servo with 1 single OP and 1 dual OP .
There is and there will be always, good and bad audio conceptions, and good and bad DC servos.
For me, John Curl, with Nelson Pass, Charles Hansen and some the others, belongs to the elite of the audio designers.
Darry
Bob Crump of T.G. Audio visited. Bob showed the CTC Blowtorch preamp, deserving of a meeting by itself. CTC stands for Curl-Thompson-Crump, indicating the design is a collaboration with John Curl and Carl Thompson.
The case of the unit (actually two, one being a power supply) is milled from a solid billet of aluminum to minimize resonances and can be built to the customer's needs, with options such as a phase switch. A phono stage similar to the Vendetta Research (designed by Curl and Thompson) will be available soon as a built in option. It is sold direct only. It should offer serious competition to other no compromise designs.
www.chicagoaudio.org/newsletters/cas0499.html
John Curl, a well known electronic designer, notes in "L. A. Audio File", July 1998.
"........We finally get into the 90's. Digital is everywhere, analogue seems to be retiring into the backwash of obsolete technology, and computers are getting more interesting. The early 90's are hard for phono stage designers, and I gave up making Vendetta Research phono stages. There was little wrong with the Vendetta and we had an ongoing operation, but nobody was buying them! My techs went to Clearcom (to make professional intercoms) and I quietly closed up shop. It is now 1997, and everybody wants a Vendetta phono stage. It is ironic, but this is the way audio world sometimes works.... This is a sad situation, but many of these manufacturers are more interested in making money than giving the audio community a shot at making better audio products in general..."
The case of the unit (actually two, one being a power supply) is milled from a solid billet of aluminum to minimize resonances and can be built to the customer's needs, with options such as a phase switch. A phono stage similar to the Vendetta Research (designed by Curl and Thompson) will be available soon as a built in option. It is sold direct only. It should offer serious competition to other no compromise designs.
www.chicagoaudio.org/newsletters/cas0499.html
John Curl, a well known electronic designer, notes in "L. A. Audio File", July 1998.
"........We finally get into the 90's. Digital is everywhere, analogue seems to be retiring into the backwash of obsolete technology, and computers are getting more interesting. The early 90's are hard for phono stage designers, and I gave up making Vendetta Research phono stages. There was little wrong with the Vendetta and we had an ongoing operation, but nobody was buying them! My techs went to Clearcom (to make professional intercoms) and I quietly closed up shop. It is now 1997, and everybody wants a Vendetta phono stage. It is ironic, but this is the way audio world sometimes works.... This is a sad situation, but many of these manufacturers are more interested in making money than giving the audio community a shot at making better audio products in general..."
Now, folks! Let's not start 'second-guessing' the power supply as well! 
I think of the power supply as doing several functions. First, it must separate the channels from each other. IF you don't use a separate regulator for each channel, you will have extra crosstalk on this design. Therefore, each circuit board and each stage (in my design) has a series open loop regulator. There are 8 series regulators in the CTC preamp in the picture.
The other chassis contains 4 regulators Two series, and two shunt. It also has a passive PI network to block RFI and keep it off the ground return.
Now, after almost 40 years, why have I done things this way?
Well, 40 years ago, usually the ONLY regulator would be a single series type for the entire preamp. This was true with AMPEX audio pro recorders, or Dynakits.
Mark Levinson introduced the first +/-15V potted module active linear regulator to audio in the LMP-2 preamp. I used this approach, BUT I had to quiet the inherent noise of the regulator (zener stabilized remember) with an active low noise cap multiplier. This seemed to work great for years, BUT I realized that it introduced considerable (xtalk) between the two channels, because of the relatively high output impedance of the open loop regulator, especially because Mark added a 2 ohm resistor to the output for current protection.
The JC-2 had a problem with imaging. It was great for mono, but stereo was not as good as some other products. I traced this partially to the power supply buffer.
For Vendetta Research, I had to use a more sophisticated approach, in order to make it extremely low noise, and not contribute a sonic 'character' to the circuit. I found that ANY aluminum or tantalum cap that was used directly across the input circuit power supply would change the sound. Therefore, I had to find a quality film cap that would work OK. The caps are prominent in the picture of the CTC. This also forced me to use all fet followers, and remove any bipolar devices. Does anyone know why?
I think of the power supply as doing several functions. First, it must separate the channels from each other. IF you don't use a separate regulator for each channel, you will have extra crosstalk on this design. Therefore, each circuit board and each stage (in my design) has a series open loop regulator. There are 8 series regulators in the CTC preamp in the picture.
The other chassis contains 4 regulators Two series, and two shunt. It also has a passive PI network to block RFI and keep it off the ground return.
Now, after almost 40 years, why have I done things this way?
Well, 40 years ago, usually the ONLY regulator would be a single series type for the entire preamp. This was true with AMPEX audio pro recorders, or Dynakits.
Mark Levinson introduced the first +/-15V potted module active linear regulator to audio in the LMP-2 preamp. I used this approach, BUT I had to quiet the inherent noise of the regulator (zener stabilized remember) with an active low noise cap multiplier. This seemed to work great for years, BUT I realized that it introduced considerable (xtalk) between the two channels, because of the relatively high output impedance of the open loop regulator, especially because Mark added a 2 ohm resistor to the output for current protection.
The JC-2 had a problem with imaging. It was great for mono, but stereo was not as good as some other products. I traced this partially to the power supply buffer.
For Vendetta Research, I had to use a more sophisticated approach, in order to make it extremely low noise, and not contribute a sonic 'character' to the circuit. I found that ANY aluminum or tantalum cap that was used directly across the input circuit power supply would change the sound. Therefore, I had to find a quality film cap that would work OK. The caps are prominent in the picture of the CTC. This also forced me to use all fet followers, and remove any bipolar devices. Does anyone know why?
On an audio DIY forum, it is natural that there are quite many intervening electronicians, passionately involved in audio, who constantly scrutinize any piece of schematics, ideas, topologies, circuits, components, etc... When they are told that, for a product, some options were choosen for technical good reasons (this is the basic rule af any design, isn't it) they expect these reasons should be made clear.
If the so-called good reasons stay mysterious - it is the prefect right of the engineer not to unveil all his secrets - these too curious people have a very sceptical reaction, and their unkindness can be on the par with the one treating them as having silly ideas : they wonder why such claims are made in the forum, suspect some bluff and consider them just as free advertisements for a commercial product.
~~~~~~~ Forr
§§§
If the so-called good reasons stay mysterious - it is the prefect right of the engineer not to unveil all his secrets - these too curious people have a very sceptical reaction, and their unkindness can be on the par with the one treating them as having silly ideas : they wonder why such claims are made in the forum, suspect some bluff and consider them just as free advertisements for a commercial product.
~~~~~~~ Forr
§§§
Hi, Charles, John and others...
John would kill anyone wanting to let anyone know of his own secrets ? Just keep quiet !!! The Vendetta man is very quiet now... No 44 or 38 Special... But with some reflection, he uses full 110 V mains... or Blowtorch to let him know about ...
Frankly, your both emulation (also with Nelson Pass and some others) is a full example of what we have to do : just to overpass our own certainties, guesses, and theory basics... Just for competition and fun...Very difficult to explain as I am an European guy, but the fulll truth ! We do it all the time here, and we are NOT all natives from the same country and have NOT the same native language... But we discuss and share ideas and findings.
Charles, you introduced an interesting folded cascode path to analyze further... (Just ask John if he doesn't think so !). When you FIRST think of learning from others and NOT to criticize all things, it is much easier to achieve, as brains are very opened to newer ideas (even ludicrous ones !). Theory don't say all, or all parts and topologies would be equal... Eh, Forr... Would you comment also about sonic nasties of caps into the DC offset correction path ? Full waste of time IMO !!!
An old short story in Europe I heard 40 years ago (serious one !) : beatles can't, mathematically, fly.... But beatles don't know maths and fly... What is wrong ? Same with audio... We are NOT an the ending point of studies, but at the very beginning... Hope that anybody will help wuith beatles understanding (anyway, experiences on bugs will also help !).
So, in audio fields, all electronics rules don't act fully but the very basic ones... I have been on John's steps for years, and he always returned to this basics from different ways...
BTW, Charles, I have an equipped room for you here at home, but will be a long run from USA (and a bittcloser to Sicilia and Corsica (perhaps) close John's cousins...
All the best, guys.... Still one of the best laughs I had on this forum for months ...
BTW, I will send you both a push-pull trick to comment (sorry, other guys,... Just want NOT to be sent to Texas !)...
jbaudiophile
John would kill anyone wanting to let anyone know of his own secrets ? Just keep quiet !!! The Vendetta man is very quiet now... No 44 or 38 Special... But with some reflection, he uses full 110 V mains... or Blowtorch to let him know about ...
Frankly, your both emulation (also with Nelson Pass and some others) is a full example of what we have to do : just to overpass our own certainties, guesses, and theory basics... Just for competition and fun...Very difficult to explain as I am an European guy, but the fulll truth ! We do it all the time here, and we are NOT all natives from the same country and have NOT the same native language... But we discuss and share ideas and findings.
Charles, you introduced an interesting folded cascode path to analyze further... (Just ask John if he doesn't think so !). When you FIRST think of learning from others and NOT to criticize all things, it is much easier to achieve, as brains are very opened to newer ideas (even ludicrous ones !). Theory don't say all, or all parts and topologies would be equal... Eh, Forr... Would you comment also about sonic nasties of caps into the DC offset correction path ? Full waste of time IMO !!!
An old short story in Europe I heard 40 years ago (serious one !) : beatles can't, mathematically, fly.... But beatles don't know maths and fly... What is wrong ? Same with audio... We are NOT an the ending point of studies, but at the very beginning... Hope that anybody will help wuith beatles understanding (anyway, experiences on bugs will also help !).
So, in audio fields, all electronics rules don't act fully but the very basic ones... I have been on John's steps for years, and he always returned to this basics from different ways...
BTW, Charles, I have an equipped room for you here at home, but will be a long run from USA (and a bittcloser to Sicilia and Corsica (perhaps) close John's cousins...
All the best, guys.... Still one of the best laughs I had on this forum for months ...
BTW, I will send you both a push-pull trick to comment (sorry, other guys,... Just want NOT to be sent to Texas !)...
jbaudiophile
OK, I should make it more clear: Bipolar transistors have 3 major noise generators. The one that I am concerned about here is the shot noise generated by current flow across the base-emitter junction. IF the base-emitter junction is NOT effectively shorted at audio frequencies, then the base current shot noise will dominate and make the regulator noisy. IF you cannot effectively short circuit the base-emitter junction at audio frequencies by using large value caps from both emitter to ground and also from base to ground, then we have a problem. A fet is different. It doesn't have any significant current flow at the input, so it doesn't get noisy, IF there is not a large capacitor at the output of the 'open loop' follower regulator. Once I heard differences in electrolytic caps in my Vendetta input circuit (thanks, Peter Morcrieff for showing this to me) over 20 years ago, I had to design them out. The 0.1 polystyrene cap on the output is not doing much at audio frequencies, and it sounds good as well, for some reason.
john curl said:
John,
Thanks for the input.
Good to see this discussion get interesting.
WRT BJT shot noise in regs I think some perspective would
be in order here. The shot noise will only be a problem if
the noise current at the base has a significant impedance to
react withand cause a noise voltage. Even quite a few hundred
ohms at the base will still result in very quiet noise performance.
As such there are many ways to get a low Z base ref point
and some not using electro caps although they are obviously the
easiest solution.
One advantage of BJT regs is they have inherently lower OP Z
(before FB) than fets.
Having said all of the above, I have not directly compared an
OL fet vs BJT reg together sonically.
Cheers,
Terry
Originally posted by lineup
At that time was not easy to find higher value film caps.
FET can be biased with higher value resistors and
these can make an effective filtering together with lower value film caps.
Originally posted by john curl
IF you cannot effectively short circuit the base-emitter junction at audio frequencies by using large value caps from both emitter to ground and also from base to ground, then we have a problem. A fet is different. It doesn't have any significant current flow at the input, so it doesn't get noisy, IF there is not a large capacitor at the output of the 'open loop' follower regulator. Once I heard differences in electrolytic caps in my Vendetta input circuit (thanks, Peter Morcrieff for showing this to me) over 20 years ago, I had to design them out. The 0.1 polystyrene cap on the output is not doing much at audio frequencies, and it sounds good as well, for some reason.
======================
What does not have a significant current flow,
does have have higher value resistors,
or am wrong?
If noone else will say it, I say it:
you were very close to the truth, lineup
today we have better film caps to use, large, higher values
and are free to use JFET or BJT nice lownoise transistors
At that time was not easy to find higher value film caps.
FET can be biased with higher value resistors and
these can make an effective filtering together with lower value film caps.
Originally posted by john curl
IF you cannot effectively short circuit the base-emitter junction at audio frequencies by using large value caps from both emitter to ground and also from base to ground, then we have a problem. A fet is different. It doesn't have any significant current flow at the input, so it doesn't get noisy, IF there is not a large capacitor at the output of the 'open loop' follower regulator. Once I heard differences in electrolytic caps in my Vendetta input circuit (thanks, Peter Morcrieff for showing this to me) over 20 years ago, I had to design them out. The 0.1 polystyrene cap on the output is not doing much at audio frequencies, and it sounds good as well, for some reason.
======================
What does not have a significant current flow,
does have have higher value resistors,
or am wrong?
If noone else will say it, I say it:
you were very close to the truth, lineup
today we have better film caps to use, large, higher values
and are free to use JFET or BJT nice lownoise transistors
Very interesting thread ! So if I understood well some posts here, in particular those of John Curl, Terry Demol and Lineup, taking also into account my own tests and listening on power supplies, something like a reference voltage followed by an RC filter like 1 MEG/1 uF film cap then a fet follower then a bipolar follower whose emitter output is shunted by something like 10 000uF could be a nice supply 
.
.- Status
- Not open for further replies.