Thanks, good to know it!
I have just seen the same shunt vs folded in the original Rod Colemans' thread.
Perhaps you could use a depletion MOSFET like LND150 in place of the tube as in this example:
I would use a single bidirectional TVS diode (SA18CA) instead of the double zener but that's just me trying for a lower part count.
Removing C2 will make it far more linear with the expense of gain - this was designed for guitar use originally and other changes can be made.
I would use a single bidirectional TVS diode (SA18CA) instead of the double zener but that's just me trying for a lower part count.
Removing C2 will make it far more linear with the expense of gain - this was designed for guitar use originally and other changes can be made.
Last edited:
Thanks kodabmx, I would like to avoid high voltages just for the line stage.
IMHE the LND150 works better if driven fom the source by a pmosfet (in source follower configuration) and the gate connected to a voltage divider from drain to ground to apply local feedback. Like the UNSET.
IMHE the LND150 works better if driven fom the source by a pmosfet (in source follower configuration) and the gate connected to a voltage divider from drain to ground to apply local feedback. Like the UNSET.
It might well.
I've never used this circuit I linked.
I did however make a simple depletion MOSFET (IXTP08N100D2) line amp. One device, 300V B+ and for 2V in you got 100V out - unless you put a bypass cap... then you got way more gain and like 45% THD (I'm guessing - the waveform was ugly).
Still, You can just adapt the same principle. Use lower voltage depletion MOSFET (or jFET?) on a lower voltage supply.
I don't know a lot about solid state parts, but if they make a small SIT, that would work well.
Why not? You can use LTSpice. That means you have access to a PC. You only need a sooundcard with microphone input, an attenuator/potentiometer to protect soundcard input and REW software to measure. For this type of circuit with high -40dB to -80dB harmonics, a cheap usb soundcard is more than good enough.
How Jfets produce harmonics strongly depends on chosen quiescent point as described in NP H2 article. Tubes will behave similarly but then you have choice of triode and pentode. Used without NFB, each and every device will produce a certain difference in characteristic if you listen closely enough. One of the reason why device matching as precisely as practical is highly encouraged for stereo reproduction.
Last edited:
Dear Roberto, I have looked at your circuit, and I describe what I see: you have a FET stage, by your own words "replacing a triode" as in: "Thinking about the classic triode", not sure why you later claim the opposite, driving a cascode connected bipolar.
That you use a PNP and wire it upside down does not change the main working principle.
As of my "distortion" comments, they also apply.
You talk about "harmonics generation" and even simulate them .... guess what´s the other name for that 🙂
Another detail: a triode will work very different whether it straight drives a resistor, or a cascode, so in that case Q1 would have the importance you assign it ... if it were used after a real triode that is; but a FET works WAY more like a pentode, quite independent from Drain ("plate") voltage, so Q1 loses importance.
To boot, it´s very linear by itself, so it will add very little flavour, if any at all.
And in any case, none of my commnets is personal, just some text analysis; If I see a contradiction I mention it, in good faith.
That you use a PNP and wire it upside down does not change the main working principle.
As of my "distortion" comments, they also apply.
You talk about "harmonics generation" and even simulate them .... guess what´s the other name for that 🙂
Another detail: a triode will work very different whether it straight drives a resistor, or a cascode, so in that case Q1 would have the importance you assign it ... if it were used after a real triode that is; but a FET works WAY more like a pentode, quite independent from Drain ("plate") voltage, so Q1 loses importance.
To boot, it´s very linear by itself, so it will add very little flavour, if any at all.
And in any case, none of my commnets is personal, just some text analysis; If I see a contradiction I mention it, in good faith.
Thank you indra1, I didn’t know it! I will search the software!
Yes, that’s the reason I’ve chosen the U440, because it already has two matched fets in it.
The purpose of Q1 is to make J1 work on a vertical loadline (Drain voltage is constant) and it is the key point both for the bandwidth and the analogy with the CCS loaded triode: the load line is perpendicular to the curves.
Indeed during winter holydays I will substitute the Source resistor with a trimmer to play with the working point. Reducing the quiescent current gives a stronger 2H. Not yet explored.
I am not the lion and all the project I sent here are well described, built and measured.
Then the discussion is always welcome if it has a logic.
That you virtual circuit hasn't
Walter
Please waltube, avoid to contaminate further my threads just to spread your unexplainable bitterness, or try to understand the circuits and actively participate to the discussion. Thanks in advance.
The 500 ohm output impedance figure catches my eye. Was that value reported by the simulation software? To be 500 ohms, the impedance of Q1's collector would have to be 1K, which is significantly lower than I would have expected.
Ciao Osvaldo,
the reason why it is here is explained in the title and in the first post: I liked how it sounds, and the simulation of the behaviour seemed similar to some tube SE amps and what I read about them (the falling of the harmonic, etc...). So I opened the thread to ask to tube SE experts if the harmonic content can be compared to a good tube SE (considering the followind unshown class D amp working at very low THD).
the reason why it is here is explained in the title and in the first post: I liked how it sounds, and the simulation of the behaviour seemed similar to some tube SE amps and what I read about them (the falling of the harmonic, etc...). So I opened the thread to ask to tube SE experts if the harmonic content can be compared to a good tube SE (considering the followind unshown class D amp working at very low THD).
The circuit in #1 can work very well. The performance depends on a quiet and stable DC supply, and careful choice of the JFET, but it has the great flexibility of gain vs. output impedance vs. drain current that is not easily matched in other configurations.
I have used almost the same circuit to make a complete RIAA LP preamp, using Shunt Cascode JFET 1st stage, and triode second stage. It has been in used for many years with a Lyra Lydian Beta and later, a Denon DL103. Perfectly quiet to the ear, huge stereophonic image, and a distinct impression of being in the recording space.
Bravo Roberto - good work!
I have used almost the same circuit to make a complete RIAA LP preamp, using Shunt Cascode JFET 1st stage, and triode second stage. It has been in used for many years with a Lyra Lydian Beta and later, a Denon DL103. Perfectly quiet to the ear, huge stereophonic image, and a distinct impression of being in the recording space.
Bravo Roberto - good work!
The purpose is to emulate a Tube; harmonics shown expect to be same or similar as those produced by a tube, sound is expected to be tube like, we are talking triode and pentode behaviour and curves, in my book that makes it fitting the Forum spirit and worth discussing here .... where else?
To boot: "Tube Experts" are called to chime in 😉