I am talking about a single transistor that drives both complementary voltage followers that need current to be driven.
It is a nonsense to think that output stage does not need current, or that it's input resistance is linear and frequency independent.
So, VAS as the name of the stage is absurd. This absurd name calls for absurd solutions.
Quite interesting,
Wavebourn, show us some example of your concept😉
At how much idle current you run your "VAS"
Did already, several years ago. 😉
Now I live in tube amp forum.
1A is usually enough.
Wait, also I don't use class AB output stage! It is either A, or C, depending on output power requirements. Class AB is the worst possible output stage (I mean, for audio applications using solid state devices).
Walker patented similar approach many years ago, it was used in Quad amps. People who don't know that output stage works in class C, the amp has no typical VAS, instead of it it has driver that works with big idle current driving load through resistor, like the sound. But orthodox engineers who can't escape from prison of presuppositions can't accept that.
1 amp! Wow! And I get by with 60-100ma and everyone else here uses... substantially less. It is well understood by some S21 characteristics of audio power transistors used as emitter followers do a darn good job reflecting the load to the stage before. With the last gain stage (called VAS here) then drivers and outputs the reactance seen by the last gain stage is plenty. Wavebourn, is this why 1 amp is used? To overcome the bad (reflected) load from emitter follower stages? Must say I totally agree with Wavebourn post#142- reality check please!
Simulation guys tend to trip over a lot of faulty theory. Guys with a test bench and real circuits trip over blown parts instead. Then learn how not to...
Here is a handy link to the 2SC5200 data sheet-
http://www.fairchildsemi.com/ds/FJ/FJL4315.pdf
1 Amp I use in Tower - V that consumes 300W idle (-50V + 100V rails), with 2x200W class A+C output. Tower-II was in class A and consumed 1200W.
Frontend of Tower-V uses 6N1P and 6P15P tubes loaded on SS gyrators. Fast and accurate error correction as the result, using nested feedbacks.
Frontend of Tower-V uses 6N1P and 6P15P tubes loaded on SS gyrators. Fast and accurate error correction as the result, using nested feedbacks.
A+C not push-pull? why not using class A instead of A+C.
Isn't that class C switches like class B?
Isn't that class C switches like class B?
A local feedback/feedforward nested EC network used as the "driver" stage will for the most part eliminate the reflection from the output transistors of the load onto the VAS. This allows for the use of a common base (gate) trans-impedance voltage gain stage without consequence. And of course with the inherent benefits thereof.😉
CBS240,
the common base stage provides superb separation, the common collector stage provides poor separation. Feedback promotes not eliminates reflections.
the common base stage provides superb separation, the common collector stage provides poor separation. Feedback promotes not eliminates reflections.
As I said before, my class A amp consumes 1200W, and is very good for winter. For summer class A+C is better. It uses feed-forward linearization of the follower, that's why it has 3 rails, for augmented source follower (or emitter follower) loaded on counter-modulated current source. Yes, it is single ended: source follower on top, and voltage to current converter on bottom, to minimize variations of voltage and current through the follower.
Class C thanks to nested feedbacks and feed-forward drive takes over class A very smoothly. It is complementary pair of BJTs. It is not switches like on/off; when voltage drop on resistor between bases and emitters is enough to charge bases emitter currents go up providing bootstrap back to bases. Actually, it works quite smoothly.
Nothing eliminates reflection, actually. Even tube's grid current depends on load. But it is possible to minimize their effect, if always remember about them and think about power amplification, not voltage/current separately in separate stages.
Last edited:
Did I? 😀
Sorry, as I said I live in tube forum now; it is quite different atmosphere there: people are used to confirmations and support, but don't expect arguments so much like in SS forum... I see, I have to go back there now... 😉
Wavebourn,
I posted this in the other thread:
In physics, using terms like "eliminate", "remove" and "cancel out" is a bit unfortunate, even if meant symbolically, "transform" is more adequate. Like any electrical properties, distortions cannot be "removed" by any means, they can only be transformed at the expense of higher complex distortions. The change made by the distortion reduction and error correction techniques is distortion, since by definition, any change (except magnitude) to the signal waveform is distortion.
I posted this in the other thread:
In physics, using terms like "eliminate", "remove" and "cancel out" is a bit unfortunate, even if meant symbolically, "transform" is more adequate. Like any electrical properties, distortions cannot be "removed" by any means, they can only be transformed at the expense of higher complex distortions. The change made by the distortion reduction and error correction techniques is distortion, since by definition, any change (except magnitude) to the signal waveform is distortion.
I agree with you; but proper selection of active components, their working points, and assistance to them can reduce distortions without multiplying order of transfer functions.
My current passion is solid state assisted tube amplifiers.
My current passion is solid state assisted tube amplifiers.
I think this is a bit too dogmatic.
Error correction in digital technology for instance can restore a damaged signal to it's original value, perfectly.
In analog, error correction or distortion reduction can bring the signal closer to it's original.
While in principle doing anything to any signal may be defined as distorting, this 'distortion' can make the signal closer to (or perfectly so) the original.
Calling such a process 'distorting' is a bit misleading; it really is what you want.
jan didden