No, I am not. There is a big difference between a LED in cathode of a triode stage, and bootstrapped one by a global feedback loop, like I have shown. And please don't start here attacks on personalities like you did on a Prodigy forum. Let's stick to the technical discussion.
But if you want to talk about personal things, I would allow myself to say, that I personally don't like the name of the topic. I don't see any "cons", I see different approaches. All components can be used, despite of "cons" when used not properly, so I would rather rename it into "How to use LED bias".
I specifically chose CONS because i wanted to know about those who tried LED bias and felt that it did not suit there amp's performance. If i had worded like you wanted those with a different POV would not feel the need to enter the discussion and make their experience known.
There is so much personality and ego tied up into building an amp or a tube circuit of ones own design that it becomes like political parties. With all sides blindly following and religiously holding on to their beliefs. I like to explore both halves of the argument and decide which is best for my needs.
So, with all due respect, no need to rename my thread. cheers, 808
So don't beat around the bush. Which way did you try that worked the best for you? Did you try both resistor on the cathode to LED and vice versa. And please explain why you liked the one or other. And of course how to determine the value of the resistor.
Well; for distortions obviously "proper" usage would be an opposite to what would be needed for less distortions. However, according to my experience, for shaping of a guitar sound saturated tubes better to be used. And it is better and more musical when the amp clips softly, for better sound expression up to a wild overdrive. I modeled a similar behavior with diode-resistive matrix in feedback loop of an opamp, and it sounded pretty toobey.
As you say, this is a very human tendency, and tends to show up in every sphere of human activity. Politics, for example, has famously been full of lies and misjudgments for thousands of years.
But in this case, there is a way out: the scientific method. Don't trust the idea, trust the math, and trust the experiment above the math.
In the case of this particular discussion about LED bias, the scientific method would work something like this: first, generate a hypothesis (such as "Increasing LED bias current lowers distortion in an LED-biased common-cathode triode gain stage.")
The hypothesis sounds plausible at first sight, but the scientific method tells us not to believe in it - instead, test it. Test it by as many means as possible: Is there a sound mathematical basis to support it? Is there experimental evidence to support it?
In this case, the answers are no, and no. Merlin's experimental data contradicts the hypothesis. And there is no sound mathematical basis for it either: lowering the dynamic resistance of the LED certainly lowers the voltage swing across it, as Wavebourn says. But this only means that there is now a larger voltage swing between the grid and cathode of the valve - and, as it turns out, the valve is less linear than the LED, so having a larger voltage swing across the less-linear element causes more distortion, not less.
Conclusion: the hypothesis "Increasing LED bias current lowers distortion of an LED-biased common-cathode triode gain stage" is false. It doesn't matter which wise and capable person supported which side of the argument; in the end, it is the truth that matters.
A side-effect of the scientific method is that it bruises egos, sometimes brutally. But it gave humanity the knowledge that took us from wooden bridges, human sacrifices, and terrible childhood diseases to interstellar probes, high quality home audio reproduction, and the relative luxury of the present time. In science and engineering, reality matters more than ego. (Something we cannot say about politics, music, acting, advertising, or most other human activities.)
So please let's try and remember that...every one of us has been wrong many, many times in our lives. It's no big deal. No need to hold on to it and make a big thing out of it.
Incidentally, I was wrong about LED bias too - I expected it to cause additional distortion due to the non-linearity of the LED itself. But measurements made by Sy, Merlin, and others show I was wrong. Reality trumps belief, so I'm chucking my former belief out of the window.
-Gnobuddy
Now i finally get it (i think). Increasing LED current reduces the swing across the LED, which in it's turn reduces the cathode feedback (just like a cathode decoupling cap would do). But at the same time the amplification goes up for the same input signal. I wonder if the distortion would be equal if the output signal would be equal for the two cases we have been discussed ?
Sorry, that wasn't a very thoughtful question I just asked in the previous post. They never really turn off, they just pass less current and the relationship is linear so at 1 microamp the slope resistance is a thousand times what it would be at a milliamp. So what are the characteristics of a *dead* LED? Say one is providing the cathode bias and it dies. Or one of a string of LEDs providing the cathode bias dies- are they like xmas tree lights where they all go out? Does it just cease passing current, so the slope resistance becomes very large and the valve goes into cutoff?
Merlin had attached a graph of distortion vs output voltage, for resistive bias, LED bias, and the same LED with additional DC current bled into it, to one of his posts in this thread. It showed (slightly) increased distortion over the entire output voltage range in the case where the LED had additional DC current flow. I believe that graph is from Merlin's book "Designing High Fidelity Tube Preamps".
I note that attachment is now missing. So is the PDF I had attached to one of my posts (where I worked out the math showing that the dynamic resistance of a diode is inversely proportional to the diode current.) A screenshot I'd attached to another of my posts is also missing.
I do not know if the moderators decided to remove those attachments, or whether they were lost accidentally while the thread was under moderation.
Perhaps Merlin will post that image again, if we ask him nicely?
-Gnobuddy
Thank you, Jazbo, but after refreshing, clearing the cache, clearing cookies, logging out and back in, even closing the browser and re-starting it, I still don't see those attachments. For example, the pdf I had attached to my post #163.No attachment was removed, you may want to refresh your browser.
To make doubly sure, I even opened the same page in a different browser (Chrome, instead of my usual Firefox.) No joy - the attachments are still missing.
-Gnobuddy