It is from the science field, "Eating fat causes fat under skin, eating cholesterol causes cholesterol on vessels, eating cucumbers causes green pimply skin".
Any circuit has a propagation delay. The maximum speed possible is that of light in a vacuum,
about 1ns to travel a foot of distance. In a real circuit this time per foot can increase by tens of percent,
depending on details. The propagation delay in a real circuit is only a few ns at most.
The major source of "delay" in a circuit is actually due to the low pass amplifier's high frequency phase shift.
If the bandwidth of the circuit is at least twice (or more) the signal's bandwidth, then the HF phase shift
(which follows an arctangent curve) is nearly linear with frequency, and so approximates a time delay.
However, this has nothing to do with the true (much smaller) propagation delay, since this "delay" depends
on the circuit's bandwidth.
Last edited:
When did I insist on using 4558 opamps in new production? B|
I just left most of them biased in the console instead of replacing all of them spending money for no reason. 😀
Not much use to explain it, they cannot understand it anyway.
And in another day, another one comes along with his own ideas and you need to explain it all over again. Google 'Sisyphean'
Jan
Last edited:
Can you imagine the difference between feedback RC loop and all that long turns in an inductor? How much meters of a wire?
His passive EQ has significantly longer delay!
Every time I have done the "legwork" or had others do portions that I couldn't do (I am no good at higher math, so that part of your "inltellectually" 😉Spell Ckecker??😉 dig is true)still usually left me with having to, AT MINIMUM, tweak by ear.
And look at all the work on RIAA feedback and the topology still basically sucks because of the fundamental aspect of an input, getting some group delay, then being radically spectrum shifted((more group delay (than just one resistor) AND frequency dependent phase shifted, all this sent back to correct an event that has already left the input, therefor now the event is correcting something different at the input.
Your bike riding analogy is fine for robot servo control.
You can have horse has left the barn audio if you want; not me and many others feel the same as the passive RIAA designs are used in the best designs.
See? How can anybody hope to explain feedback if you don't even understand the concept.
Sisyphean, I say.
Jan
But it's not sent back to the input to correct an event that is not there anymore in a passive design.
You can't see this? Does it take the higher math dummy to hammer this in your head? WOW! just WOW! plane as day right in front of you.
A feedback event that is correcting something that is not present at the input. On a grosser level you get ringing at the output. You like ringing in your audio? I don't not even a little bit.
I think we're seeing is just perfectly fine, there's nothing revolutionary going on here. Now, the emphasis on the worry about a ns of prop delay is very different.
Totally shocking you guys are. You can't grasp a fundamental concept that the higher math dummy can see.
See my comment below on this.
Oh no, not THIS again! I'd seen this argument before on diyaudio, don't remember if it was morinix.
Long rant short, I agree with others' responses.
In response to the first quote, LPs have a spiral groove. Edison cylinders have a helical groove, and as far as I know they were always acoustically recorded, and thus never had any [electronic] EQ applied, much less anything specified by the RIAA.
But if the whatever of LPs affects phase (or whatever else is claimed to be audible) so badly, I'm surprised there's not a greater audiophile demand for cylinders.
How can you be so obsessed with feedback delay while totally ignoring skin effect?
Don't bother getting out your calculator, they both clearly exist as electrical phenomena so they both must affect audio negatively.
Does it take the higher math to understand that Armstrong & Fitzgerald are dead long time ago? How long ago? But their voices still sound in your wires!
Compare with picoseconds of delay!
BTW I'm not the only one. The designer of the Manley Chinook and Steelhead will tell you pretty much the same I am.
Unfortunately it is too late for me to use an art of hypnotic suggestion, as John Curl pointed out... People simply won't believe me... :-(
Sniping comments merely serve to degrade the quality of the discussion, think about it.
To answer an irrelevant analogy, Too much time actually.
It is not possible to ride a bicycle in a perfectly straight line in the presence of perturbations in the form of cross winds, ground surface slopes etc, and even without any such perturbations.
With a highly skilled rider an approximation is possible, but a perfectly straight path is not.
At slower speeds, the bicycle path is a continuous series of smaller errors resulting in weaving path approximating but never actually a straight line path.
At higher speeds, the path is indeed straighter due to 'help' from the gyroscope action of the wheels, but is still never exactly straight.
Now if the task is to follow a curved path, the errors will compound, and according to the speed of the bicycle and the perturbations.
If a further confounder such as alcohol is added into this equation, the accuracy of the feedback/servo mechanism goes right out the window, even to the point of total failure/crash.
Feedback is useful, but it will never instantaneously fully correct an inherently unstable system.
Think through all the perturbations in an electronic/electrical/electromechanical circuit/mechanical system and you will see that exact/total precision is not possible.
Dan.
- Status
- Not open for further replies.