About amp "immediacy"

[nosaj]

Ok say we put a meter in line with the clamp feedback circuit. how do you mitigate the loading effect of the meter and nullify the input impedance so that it doesn't change the circuit for what you tryng to measure? Also when you switch meter in and out of said circuit how long should it take to settle what your liking for?
Nosaj Et

[Mike Sulzer]

Instead of provoking all out warfare like you are doing, why don't you try to setup his feedback example and see if it works like he says. Are you afraid to discover he was right? There are lots of ways to be a coward.

How can you set up a "feedback example" that has not been fully specified? I suppose that it is purposely vague so that no one can.

[Mike Sulzer]

Yldouright is selling new sharper/faster tools

No, if he is selling anything, it is garbage.

[yldouright]

@Mike Sulzer and all the other doubting Thomases

1. Pick any gain device of your choice for a simple amp circuit that can do ~40dB of OLG in SOA.
2. Setup a switch for a ~20dB feedback loop to be introduced.
3. Feed your choice of signal and monitor the voltage with both analog and digital meters.
4. Switch in the feedback and note the time taken for the voltage to settle.
5. Factor out the specified settling time of the measuring devices.

I'm assuming that you'll be able to manage your device selection and keep it out of oscillation, etc.

[nosaj]

So how do you account for loading the feedback witha meter?
Nosaj

[Helmholtz]

It is ridiculous to make such measurement with simple meters. Even analog meters are much too slow. It needs a high speed scope to measure such short time delay.

[Mike Sulzer]

@Mike Sulzer and all the other doubting Thomases

1. Pick any gain device of your choice for a simple amp circuit that can do ~40dB of OLG in SOA.
2. Setup a switch for a ~20dB feedback loop to be introduced.
3. Feed your choice of signal and monitor the voltage with both analog and digital meters.
4. Switch in the feedback and note the time taken for the voltage to settle.
5. Factor out the specified settling time of the measuring devices.

I'm assuming that you'll be able to manage your device selection and keep it out of oscillation, etc.

You still have not specified what you are doing. Is this dc and ac feedback or just ac? if there are any large capacitors involved, is there an issue with the voltage across them changing when switching from one state to the other? An actual schematic of the test might help.

RG might have some other issues, but this is what comes to mind.

But surely you must have enough common sense to realize if you have a result that does not agree with standard expectations, you check every possibility before declaring that there is a problem with the theory of E&M and circuits. There is no way anyone should believe what you are saying without at least the equivalent of a Phd thesis.

[rjb]

4. Switch in the feedback and note the time taken for the voltage to settle.

It's been approximately a million years since I've done any kind of circuit analysis, but:
By suddenly switching in the feedback, aren't you slamming the input with a signal approximating a step function (or maybe a stream of delta functions due to switch bounce)?
And don't "classic" EE principles predict that just about any real-life circuit is going to "ring" under such circumstances?

Before answering, it sure would be nice if you could post a pic of your experimental setup, and the measurements and calculations that led to the "35 mph" statement.

-rb

[Leo_Gnardo]

It is ridiculous to make such measurement with simple meters. Even analog meters are much too slow. It needs a high speed scope to measure such short time delay.

NOW we're talking! Thank you Helmholtz, the voice of reason.

Memory scope at that. The fastest you can get.

[yldouright]

There is no way anyone should believe what you are saying without at least the equivalent of a Phd thesis.

I believe all I say was submitted and rejected as a PHD thesis and that student eventually went to work for some private interests that do government contract work. I want to make clear that this is not my discovery and everything you hear from me is handed down from others closer to the origin so I don't want to take an authority position here. I recognize the giant gulf of understanding between myself and some members here when the subject is classic academic E&M. You should view my posts like I'm the child who says the Emperor has no clothes. If some of you happen to see this too, you will probably be much better able to explain how this came to be than I can.

It is ridiculous to make such measurement with simple meters. Even analog meters are much too slow. It needs a high speed scope to measure such short time delay.

Maybe if the signal path is made long enough or have a conduit sufficiently resistive to slow things down enough to record? The idea would be to observe the time expected if the signal was in fact traveling instantaneously compared to actual and then sort out what's unexplained.

[nosaj]

Oh so your asking people to make a circuit to support your ideas? Thats just as ridiculous as your other ideas.
Nosaj

[R.G.]

@Mike Sulzer and all the other doubting Thomases
1. Pick any gain device of your choice for a simple amp circuit that can do ~40dB of OLG in SOA.
2. Setup a switch for a ~20dB feedback loop to be introduced.
3. Feed your choice of signal and monitor the voltage with both analog and digital meters.
4. Switch in the feedback and note the time taken for the voltage to settle.
5. Factor out the specified settling time of the measuring devices.
I'm assuming that you'll be able to manage your device selection and keep it out of oscillation, etc.

Let's compare theories. I'll tell you how existing theory explains this kind of observation, and then you tell me how your new, improved theory explains it.

There are time constants inside all feedback amplifiers. These time constants are either inherent in the active devices, or deliberately designed to maintain stability (go look up Nyquist stability criterion). Any time you mess with the feedback network, you introduced a transient into the amplifier. An external transient in a feedback network causes the amplifier to correct for this transient, including any setting time on the internal time constants. It's notable that existing theory predicts this pretty exactly. To existing theory, rebalancing time is to be expected. No surprise. In a feedback system that uses its feedback to maintain DC conditions, dinking with the feedback path will cause the amplifier to rebalance its DC conditions as well.

That is, an amplifier responding to sudden changes in its feedback path is no reason for conjuring up new theories of the universe, nor for trashing the entire existing understanding of electronics. The existing understanding of electronics does in fact predict the measurable results remarkably well.

Now - you tell me (and us) (a) what your theory of electronics is, and (b) how its set of rules for how electrons act would predict (accurately, please) the actions you described in 1-5 above.

Please - tell us the alternate rules which explain electronics action, and how to use them to design. Just the facts, please.

Be aware that I'm going to toss Occam's Razor and Hume's Maxim back at you.

Edit: for extra credit, tell us what speed those "analog or digital meters" would have to be to measure the settling time in a VHF radio transmitter.

[Chuck H]

Geez! Enough with trying to back up the speed of electrons moving through a conductor being 35MPH already. It's just not true. Originally it was implied regarding signals!?! Imagine how long it would take to actually hear the note at the other end of something like a guitar pickup. The affect would be overtly obvious. Like a third of a second!?!

And as noted, any standard bench meter, analog or digital, is going to be too slow to keep up with the shift in the above test. Why is feedback a component of the test? Is it because the signal path for the feedback is suffering phase error due to the signal path length? Further, if there is a noteworthy delay in the response of the meters that were telling an accurate story about actual circuit conditions then GNF wouldn't even work at all because managing those phase shifts would be impossible (or at least sound horrible).

[nickb]

Geez! Enough with trying to back up the speed of electrons moving through a conductor being 35MPH already.

Drift velocity is more like 1mm /sec. With a 3M lead you won't even get your first note out before the drummer finishes the entire first set and has left for the bar.

[netfences]

What happened to my humble thread? Why can't you guys just open up a new thread about this?