This thread provides a summary of the title audioxpress article. Those who have not read it will find the attached summary useful. Since the attachment provides some refinements if the article's content, those who have read the article might also benefit from reading the attachment. Comments are welcome.
Last edited:
I know some people look askance at a triode model which consists of a voltage source in series with a resistor. The good new is that for component values of V and R respectively, such a circuit performs indistinguishably from a current source of value V/R in parallel with the resistor R. So why not introduce the Oracle equation using a triode current source? It could be done. But the voltage divider equation then changes from Rload / (Rload + Rx) to a "busier" parallel resistance expression of Rload * Rx / (Rload + Rx). And the voltage source expression e then changes to a "busier" e / Rx current source. The two newly introduced Rx values cancel, and we're back to the original voltage source model.
I will read the article in AudXpress, which edition did it come with? I get the mag at work but haven't had the time to read them very thoroughly the last year or so...I don't know about the other members here, but I think we are generally heavy on the practical aspect of tube design and not so affected by deeper theory. Perhaps why no comments yet. ?
I like any article about tubes, but here I feel the article takes my beloved electrons trapped in a bottle full of vacuum and turns them into math equations. I admit I tend to ignore theory and math as soon as there are more than half an equation in a paragraph. I am amazed I survive as an EE in the lab I work, but thankfully we need practical engineers too. I much enjoy the visual solving using the good old datasheets and trace my lines across the curves. I am not really sure what the article is trying to do? I think present a better model than what is common? But since I have no intimate relation to the models I use I do not know if the models in the article are better or worse or the same. In my case the models in LTspice are just fine for the type of circuits I simulate.
In my case I think the article needs a header with points telling what it is about and what is trying to achieve, or what key take-aways it gives the reader. Sorry, I think anyone more into simulations will get your article much better than me, but at least you know why I have not placed a comment on it. Untill now...
I like any article about tubes, but here I feel the article takes my beloved electrons trapped in a bottle full of vacuum and turns them into math equations. I admit I tend to ignore theory and math as soon as there are more than half an equation in a paragraph. I am amazed I survive as an EE in the lab I work, but thankfully we need practical engineers too. I much enjoy the visual solving using the good old datasheets and trace my lines across the curves. I am not really sure what the article is trying to do? I think present a better model than what is common? But since I have no intimate relation to the models I use I do not know if the models in the article are better or worse or the same. In my case the models in LTspice are just fine for the type of circuits I simulate.
In my case I think the article needs a header with points telling what it is about and what is trying to achieve, or what key take-aways it gives the reader. Sorry, I think anyone more into simulations will get your article much better than me, but at least you know why I have not placed a comment on it. Untill now...
So? Blind-spots are rampant. Ignore these folks.
I missed the AE article and have NO idea who this "oracle" is or what that argument is about.
The pictures and equations look correct (at a glance) and should be a handy reference for anybody developing improved tube models. That is not a common or frequent past-time. I've dabbled a few hours over a few decades so I'll save your essay for the next time I get excited.
Thanks for your comments. You and I are birds of different feathers, SemperFi. I love the math!
From the article in the October 2022 issue, "My motivation for writing this article is to propose a simple, straightforward means of enhancing the understanding of specific triode circuits and to settle certain controversies, some of which are well known and long standing." and "I hope that readers will see some value in it and that it might be useful in settling some of what The Valve Wizard calls '… heated debate in the pages of audio magazines for decades.' "
As the Valve Wizard says above, there is a long history of misunderstandings of the simple Cathodyne. The last part of the article details a number of these.
The attachment in this thread is not meant to stand alone. It is an annex to the article. Reading it without reading the article first will be problematic.
In general, the LTSpice triode models work well. But spice models are like a little bit of magic. They can tell you what the circuit does, but they cannot give you an understanding of how the circuit works. That's where the math comes in. I do not know of a better way to convey that understanding.
Thanks for your comment, PRR.
The whole point of the article with the name of this thread in the October 2022 addition of audioXpress was to focus on blind spots like that one (especially as they relate to a Cathodyne) and to provide as simple as possible an explanation of how that circuit works. There is a long history to this present day of misunderstanding Cathodynes. And some of those misunderstandings are promulgated by well-known and respected authors.
The article discusses what I've called the "Oracle equation." This non-controversial equation relates a triode's small signal grid and anode voltages and is a gold mine of other information. It can be re-arranged to give output voltages and impedances of Cathode Follower, Voltage Amplifier, and Cathodyne stages. All this without invoking Thevenin's Theorem or Feedback Theory, or anything more complicated than the resistive voltage divider equation. This is because given the load L that is the part of a stage across which its output voltage is developed, we can always derive an equation of the form output voltage = K * grid voltage * L / (L + Rx). Here K is some unitless constant and Rx is a resistance. The output impedance is simply L in parallel with Rx.
The posting in this thread was never meant to stand alone. It’s an annex to the article which references this thread. I was apparently wrong in assuming that most people who came across this thread would have come here from the article. If you have an opportunity to read it, you might find it interesting.