john curl said:
John could you develop more precisely this point and what precisely do you call symmetric/asymmetric x-talk please.
john curl said:
From this point of view bjts would seem best appropriate since they have a rather lower parasitic capacitance than fets/mosfets, however you use them in preamps output stages (input stage are folded cascoded) so I guess that despite this drawback they still perform better than bjts, why? Channel separation (gate bias) ?
P.S. Plan you to come to the AES exhibition in Paris?
Pavel
What I wrote wat not meant as a judgement of your taste in performance. Nor was it a discribtion of my own preferences. but you asked for a simple good sounding circuit. I belive that resolution is a great part of everything It is one of the issues i hunt in my designs. Resoulution to me is the absence of resonanses. Resonanses in materials, in membranes, in kabinets. I belive that the casework and the sorrounding that electronics is kept in means a whole lot. So for me mechanics plays an important role, also in electronics. One clue to the perfrormance of the Torch might well be superior mechanics. I belive that also the execution of PCB's, selection of resistors, input circuit switching, and the great joker the volume control ,is just as important as low distortion. If all's equal distortion might play a role. but it is under no circumstance the deciding factor. Tubes can sound wonderful and so can SS. I think the rightness of tones. The trasient and the way the sustain and decay develops is the true judge. I my expirence people all react with a nice happy smile when they hear something extra ordinary. It dosen't take any training, my wife can judge it instantanious. But then she can judge anything instantanious.
What I wrote wat not meant as a judgement of your taste in performance. Nor was it a discribtion of my own preferences. but you asked for a simple good sounding circuit. I belive that resolution is a great part of everything It is one of the issues i hunt in my designs. Resoulution to me is the absence of resonanses. Resonanses in materials, in membranes, in kabinets. I belive that the casework and the sorrounding that electronics is kept in means a whole lot. So for me mechanics plays an important role, also in electronics. One clue to the perfrormance of the Torch might well be superior mechanics. I belive that also the execution of PCB's, selection of resistors, input circuit switching, and the great joker the volume control ,is just as important as low distortion. If all's equal distortion might play a role. but it is under no circumstance the deciding factor. Tubes can sound wonderful and so can SS. I think the rightness of tones. The trasient and the way the sustain and decay develops is the true judge. I my expirence people all react with a nice happy smile when they hear something extra ordinary. It dosen't take any training, my wife can judge it instantanious. But then she can judge anything instantanious.
Miib, you are on track with the component selection and the box construction. If you tap the side of a Blowtorch preamp with your finger, you get something closer to a 'thump' rather than a tap. This is because the Q of the cabinet is much different than sheetmetal or plastic boxes. This seems to be important. We use the VERY EXPENSIVE TKD pots that are only obtainable from Michael Percy (at least here in the USA). They cost over $160 each! We need two for each preamp. Still, they sound darn good.
No, the success of the sound is not exclusively locked up in the line amp board.
Now, those who complain that they will never hear one of these preamps. Well, make or buy the best that you can and enjoy it. I can't afford to buy a Bugatti auto either (1,000,000 Euros) so I drive a 22 year old Porsche 944. I can still admire the technology behind a modern super-car.
JCMD, I mean by asymmetrical x-talk that right channel goes to left channel differently than left goes to right. No, I won't be in Paris.
No, the success of the sound is not exclusively locked up in the line amp board.
Now, those who complain that they will never hear one of these preamps. Well, make or buy the best that you can and enjoy it. I can't afford to buy a Bugatti auto either (1,000,000 Euros) so I drive a 22 year old Porsche 944. I can still admire the technology behind a modern super-car.
JCMD, I mean by asymmetrical x-talk that right channel goes to left channel differently than left goes to right. No, I won't be in Paris.
Unlike other members here, I'm not EE, so what I wrote below might be wrong. I cannot supply any differential/integral mathematical equation that will support it either 
I just wrote based on my analysis of my experiments.
Many times I've been asked : putting a preamp will degrade or give us better sonics? Isn't that preamp is another "degrading device" for the audio chain, ie : putting preamp will degrade the sound?
By the same above thinking this will result that the "best" preamp is one single potentiometer/resistor voltage divider?
I think that it is not necessarily so.
In the above statement, the asking person have 1 fundamental assumption (not realized), that the combination between signal source and the power amp input will give 100% signal transfer. Both signal source impedance and power amp's input impedance is matched 100%.
What if the combination between signal source and power amp input impedances (driving capability) is only resulting only 70% of the signal will be reproduced by the end of power amplifier's speaker outputs? And if we put a preamp between the signal source and the power amp, and this preamp will match the impedances better, then we will have 95% of the audio signal will be there in the amp's output terminal, rather than only 70% condition if we are not using preamp? Wouldn't it make a preamp is a "hero" than "evil" like in the assumption case?
This kind of "not realized assumption" is common to people when thinking about how things "would" work out.
Feedback is a good concept. What is not good is how the feedback is implemented. Unfortunately the main barrier still comes from nature, the transistor's behavior. I bet if someday we have a linear maginification transistor, without (i+jw) properties (only real number, (i)), everyone will be using feedback without hesitation about how the sonic results will be. Everyone will apply tons of feedback without doubt. Ooops....bad example. With linear transistor, who will need feedback?
We got a car and a road. This is one special car. You can replace the engine like LEGO / kid's toy. You can put 500cc engine (this one will run 10km/h) in it or put 5000cc V12 (this one will run 250km/h), whatever you like.
The analogy is 500cc engine equal to no-global feedback topology and 5000cc V12 engine is high global feedback topology.
Now the road. There will be 2 kind of road. Smooth asphalt road (like high-way) and bad+brick sized stones road (like in the remote, non touchable area).
Feedback (250km/h engine) will run very-very good on the car IF the road is smooth asphalt road. If you put this 250km/h engine for the bad road, you will damage the car, maybe you won't reach finish because your car will be flying without parachute first
For the bad road, the 10km/h engine will bring you safer to the destination than the 250km/h engine.
What is the relation of car analogy with feedback/non global feedback?
Back again to the first story, people usually have "not realized assumption" that any cct will receive any magnitude of OLG(=feedback) and would work out. They don't realize that there are "bottleneck(s)" in the cct that won't permit electrons to flow more. One example is TIM phenomena existed because there ARE bottleneck(s) in audio path. You cannot put more force to pass 1 bottleneck, putting more force will result in more destruction instead of getting better results. This is the "bad" road.
Most of the bottlenecks are in the base area. Whether it is the base/gate drive condition or base-emitor / base-collector transfer function. Transistor STILL have capacitance (and non linear magnification) there.
From my experiment, feedback is not evil, what is evil is the road that this feedback must pass. Many bottlenecks can be healed by better base drive/base condition and transistor magnification "preconditioned" first. Better cct. With better road, we can put more feedback. High global feedback will not work (result in good sonics) if the "road" is not preconditioned/treated first.
Still, nature gives us the ultimate bottleneck. Internal capacitance and non-linear relation of transistors. While we can trick/preconditioned the cct, we still cannot trick the internal properties of 1 transistor.
I just wrote based on my analysis of my experiments. Many times I've been asked : putting a preamp will degrade or give us better sonics? Isn't that preamp is another "degrading device" for the audio chain, ie : putting preamp will degrade the sound?
By the same above thinking this will result that the "best" preamp is one single potentiometer/resistor voltage divider?
I think that it is not necessarily so.
In the above statement, the asking person have 1 fundamental assumption (not realized), that the combination between signal source and the power amp input will give 100% signal transfer. Both signal source impedance and power amp's input impedance is matched 100%.
What if the combination between signal source and power amp input impedances (driving capability) is only resulting only 70% of the signal will be reproduced by the end of power amplifier's speaker outputs? And if we put a preamp between the signal source and the power amp, and this preamp will match the impedances better, then we will have 95% of the audio signal will be there in the amp's output terminal, rather than only 70% condition if we are not using preamp? Wouldn't it make a preamp is a "hero" than "evil" like in the assumption case?
This kind of "not realized assumption" is common to people when thinking about how things "would" work out.
Feedback is a good concept. What is not good is how the feedback is implemented. Unfortunately the main barrier still comes from nature, the transistor's behavior. I bet if someday we have a linear maginification transistor, without (i+jw) properties (only real number, (i)), everyone will be using feedback without hesitation about how the sonic results will be. Everyone will apply tons of feedback without doubt. Ooops....bad example. With linear transistor, who will need feedback?
We got a car and a road. This is one special car. You can replace the engine like LEGO / kid's toy. You can put 500cc engine (this one will run 10km/h) in it or put 5000cc V12 (this one will run 250km/h), whatever you like.
The analogy is 500cc engine equal to no-global feedback topology and 5000cc V12 engine is high global feedback topology.
Now the road. There will be 2 kind of road. Smooth asphalt road (like high-way) and bad+brick sized stones road (like in the remote, non touchable area).
Feedback (250km/h engine) will run very-very good on the car IF the road is smooth asphalt road. If you put this 250km/h engine for the bad road, you will damage the car, maybe you won't reach finish because your car will be flying without parachute first
For the bad road, the 10km/h engine will bring you safer to the destination than the 250km/h engine.
What is the relation of car analogy with feedback/non global feedback?
Back again to the first story, people usually have "not realized assumption" that any cct will receive any magnitude of OLG(=feedback) and would work out. They don't realize that there are "bottleneck(s)" in the cct that won't permit electrons to flow more. One example is TIM phenomena existed because there ARE bottleneck(s) in audio path. You cannot put more force to pass 1 bottleneck, putting more force will result in more destruction instead of getting better results. This is the "bad" road.
Most of the bottlenecks are in the base area. Whether it is the base/gate drive condition or base-emitor / base-collector transfer function. Transistor STILL have capacitance (and non linear magnification) there.
From my experiment, feedback is not evil, what is evil is the road that this feedback must pass. Many bottlenecks can be healed by better base drive/base condition and transistor magnification "preconditioned" first. Better cct. With better road, we can put more feedback. High global feedback will not work (result in good sonics) if the "road" is not preconditioned/treated first.
Still, nature gives us the ultimate bottleneck. Internal capacitance and non-linear relation of transistors. While we can trick/preconditioned the cct, we still cannot trick the internal properties of 1 transistor.
David -
I'll ask you another question. In case you sit on board of an airplane, would you choose an airplane operated without temperature sensors, motion sensors, air speed sensors, humidity sensors, height sensors, fuel level sensors (i.e. non-feedback drive) or will you prefer a feedback drive?
Happy flight in a non-feedback plane!
P.S. Some of the designs really remind me Sopwith Camel
I'll ask you another question. In case you sit on board of an airplane, would you choose an airplane operated without temperature sensors, motion sensors, air speed sensors, humidity sensors, height sensors, fuel level sensors (i.e. non-feedback drive) or will you prefer a feedback drive?
Happy flight in a non-feedback plane!
P.S. Some of the designs really remind me Sopwith Camel
Regarding pots - I remember well the story when my breadboard prototype of prestandard preamp (with 2,- USD plastic Piher pot)defeated renowned highend preamp rated at 8.500,- USD, and I also remeber faces of highend studio people when I showed them what was inside my casework ...
Seems we start to skate on alchemy ice here.
Seems we start to skate on alchemy ice here.
Haha.. those analogies, cars, aeroplains.. well, we don't need anything in a simple small aeroplain, aerodynamically stable plain need only a Otto motor, throttle and rudder, the feedback is the pilot himself!
John,
going back couple of days or so when cables wher up on the topic, what would you think of cables with plastic conductive shield?
Cheers Michael
John,
going back couple of days or so when cables wher up on the topic, what would you think of cables with plastic conductive shield?
Cheers Michael
- Status
- Not open for further replies.