No, experiments don't show it to be correct because your approach and results are incapable of doing so.
Guys, don't forget this very important fact: Chris' circuit makes the same predictions as SY's model does given the "boundary constraints".
If SY's experimental results show his model to be correct, they also show Chris' circuit to be correct. But that's a contradiction. They cannot both be correct.
Thus, SY's results do not show his model (or Chris' or mine) to be correct. SY's experiments are, shall we say, impotent.
No and no, respectively. Unless you're also claiming that if, for example, I model a circuit as a Norton source and you model it as a Thevenin source, the models can't both be correct. The term "model" seems to confuse you.
Just a constraint (I like Dave's euphemism).
If the constraint is met, the circuit acts exactly like my Thevenin model for any arbitrary pair of loads (at least so far- I don't deny that theoretically someone could suggest a pair of loads which will cause my model to give incorrect results, it's just that no-one has yet). If the constraint isn't met, the model doesn't work, as I've stated repeatedly.
If the constraint is met, the circuit acts exactly like my Thevenin model for any arbitrary pair of loads (at least so far- I don't deny that theoretically someone could suggest a pair of loads which will cause my model to give incorrect results, it's just that no-one has yet). If the constraint isn't met, the model doesn't work, as I've stated repeatedly.
Yes, Chris' circuit gives the same predictions as your model when the "boundary constraints" are met.
Yes, they cannot both be correct because Chris' circuit has very different source impedances than your model.
The point is, your experimental approach and results are impotent. They do not uniquely determine a model. That's the fatal flaw in your approach. That's what makes it not even wrong*.
Last edited:
Unfortunately, Chris's model does not give the correct answer for at least one pair of loads (see my upcoming LTE in Linear Audio). Mine does for ALL pairs of loads, unless you've finally come up with a pair for which is gives incorrect results. I kinda doubt that, but if you can do it, by all means do it.
If accurate prediction and experimental verification are "impotent," then give me more impotent! Misusing slogans from actual physicists is not a particularly good substitute for evidence.
If accurate prediction and experimental verification are "impotent," then give me more impotent! Misusing slogans from actual physicists is not a particularly good substitute for evidence.
Leadbelly, the model employs equations which predict impedances - "experimental data." And then it is claimed that the model is the only arbiter of the data's corrrectness.
If you want to talk about physical measurements, they are taken assuming the model is correct - assuming that you connect plate, cathode and ground together to measure short circuit currents. But this is the very assumption that is in contention.
Data which in this case are derived and depend on the use of a model employing procedures or equations which are in contention cannot be used to prove the model itself is correct. At the very best, it can show that the model does not contradict itself. It cannot show that the model is right.
Ptolemy's model of planetary orbits said that the planets travel in interconnected circles of different diameters rotating at different speeds. "Show me orbital data," he might have said, "and I will find you a pattern of circles that describes it." His system is not self-contradictory, but most of us today accept that it is wrong. His "circles" are the incorrect experimental results of his model of faulty assumptions.
Now that I've answered your question, please respond to mine:
Take me through an analysis of the P-K, K-gnd and P-gnd impedances of the Cathodyne using Thevenin's theorem. It'll give you different answers from SY's approach.
Reconcile the simple current mirror circuit in my latest Linear Audio LTE with SY's article's claim that in a Cathodyne whose grid is driven by a square wave and whose P & K loads are identical parallel R-C networks, that equal rise times imply equal drive impedances.
If you want to talk about physical measurements, they are taken assuming the model is correct - assuming that you connect plate, cathode and ground together to measure short circuit currents. But this is the very assumption that is in contention.
Data which in this case are derived and depend on the use of a model employing procedures or equations which are in contention cannot be used to prove the model itself is correct. At the very best, it can show that the model does not contradict itself. It cannot show that the model is right.
Ptolemy's model of planetary orbits said that the planets travel in interconnected circles of different diameters rotating at different speeds. "Show me orbital data," he might have said, "and I will find you a pattern of circles that describes it." His system is not self-contradictory, but most of us today accept that it is wrong. His "circles" are the incorrect experimental results of his model of faulty assumptions.
Now that I've answered your question, please respond to mine:
Take me through an analysis of the P-K, K-gnd and P-gnd impedances of the Cathodyne using Thevenin's theorem. It'll give you different answers from SY's approach.
Reconcile the simple current mirror circuit in my latest Linear Audio LTE with SY's article's claim that in a Cathodyne whose grid is driven by a square wave and whose P & K loads are identical parallel R-C networks, that equal rise times imply equal drive impedances.
Yes DF96, I think what you say is reasonable. But I also think that the problem that occurs is that the constraint as applied violates Thevenin's theorem. Thevenin deals with measuring short circuit currents when any two nodes are shorted. SY's boundary conditions lead him to short together three - Ground, Plate and Cathode. The results are different from those obtained with a straight forward application of Thevenin- shorted currents or voltage differences involving two nodes only.
SY, I recommend you cheese it with the personal deprecations. We wouldn't want you to get tossed!
Anyone can look at the Thevenin theorem and see that only two nodes are involved. It is truly and completely beyond me how you or anyone else can claim that it can deal with three nodes at once.
Since it is so clear to you, how about a detailed explanation here of how to apply it so? Step by step. Number them for subsequent reference. Lay it all out. You have nothing to fear if you are right.
And of course, as always it is your choice to choose to be snide and flippant. But if you do so, you'll lose the oportunity to be informative. Which do you want to be?
What you don't realize about your two loads is that the lack of any pair that proves your model wrong shows only that your model is not self-contradictory. It does not prove it is right. If your model demands that light be observed only through rose colored glasses, you will see only rose colors. Your model will not be self-contardictory, but it's still wrong. This simple logic seems to escape you.
And I do hope your forthcoming LTE will deal with the rise time problem in your article. Because you're not dealing with it here.
Anyone can look at the Thevenin theorem and see that only two nodes are involved. It is truly and completely beyond me how you or anyone else can claim that it can deal with three nodes at once.
Since it is so clear to you, how about a detailed explanation here of how to apply it so? Step by step. Number them for subsequent reference. Lay it all out. You have nothing to fear if you are right.
And of course, as always it is your choice to choose to be snide and flippant. But if you do so, you'll lose the oportunity to be informative. Which do you want to be?
What you don't realize about your two loads is that the lack of any pair that proves your model wrong shows only that your model is not self-contradictory. It does not prove it is right. If your model demands that light be observed only through rose colored glasses, you will see only rose colors. Your model will not be self-contardictory, but it's still wrong. This simple logic seems to escape you.
And I do hope your forthcoming LTE will deal with the rise time problem in your article. Because you're not dealing with it here.
There are two Thevenin sources. One. Two. Are you with me so far?
The "rise time problem" was your rewording of experimental verification, not mine. How's that pair of loads coming along?
Do you even read what I write SY?
There are no two such loads. Congratulations. Your model doesn't contradict itself. But that's a pretty low level of performance. It would be nice if the lack of such loads could prove that it is correct. But it doesn't.
You're avoiding the problem of your rise time experiment by quibling with my wording?
There are no two such loads. Congratulations. Your model doesn't contradict itself. But that's a pretty low level of performance. It would be nice if the lack of such loads could prove that it is correct. But it doesn't.
You're avoiding the problem of your rise time experiment by quibling with my wording?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.