That was really cool!!!
Somebody hates laptops almost as much as I do....😀
I love the "and some electronics on top to keep it from blowing up".
That instills confidence, especially when it's about a foot from the head.
jn
The people who really have no time for computers are those who've tussled with them, every day, for the last 40 years ... 😉
Yes, my bad. You commented on Bateman and treatment of impedance as a real number. Davis does the same thing, so misremembered.
Well it's impossible to measure if it's not, and it isn't.
Under the artificial conditions you defined
No, it requires proper interpretation as to what it means. It does not mean a real 1R as you use in your models. But apparently you either can't offer an explanation or don't even see the issue.
Yes it is.
Too bad the massive rail gun project has hit some snags such as each firing destroys the required alignment.
You've made at least three huge conceptual mistakes in your understanding, JN. First is to neglect the effect of audio bandwidth restriction on risetime. Second is to model speaker impedance dips as though they apply across an infinite bandwidth. Thirdly is to model impedance dips as though they are resistive - ie to use as 1R, rather than a reactive or active load model amounting to |Z| =1R. Collectively this thoroughly trashes your model.
Work out what it means to achieve |Z| = 1R when the voice coil resistance is near 8R. Then apply that load model to whatever analysis floats your boat TL, LCR, multi stage LCR all give the same result using audioband risetimes. And you'll immediately see your model and argument for cable latency c 5uS falls apart because 1R resistive load is seriously not realistic. Really.
If 5uS is 'geological', the typical current latency in the impedance dip might be considered 'cosmological' - work it out.
DF96 alluded to the importance of using complex impedances earlier in this thread, and he's right. IMO if one has to use a nominal resistive impedance 8R is as good as anything.
JN, if you really either can't do the analysis or run a simulation (which is near trivial) to confirm this, that's OK I'm sure if you ask nicely someone might do it for you. But it's best first if you at least try to research and understand the concepts in play here. Sooner or later the penny will drop.
Last edited:
To be clear as to the Otala and Hutterman reference in Davis paper, what they claim to show is that maximum value of peak current reported for a speaker can be 6.6 times larger than that of an 8R resistor, and say that such results imply that for short periods of time an amplifier should be able to drive, with full output voltage swing and without appreciable distortion, loads equal to a resistor of 1ohm - which IMO is true in terms of current drive capability BTW.
Davis briefly restates this in his paper in different terms.
However, that is not to say that load impedance is ever 1R resistive - that depends on instantaneous load current phase wrt applied voltage - and hence the critical matter of interpretation. Load impedance |Z| = 1ohm does not mean a 1R resistive ohm. Ohm sweet ohm.
FWIW O&H also present that current peaks typically last a few hundred microseconds or are caused by the simultaneous parallel excitation of several drivers of multiway system, by the summation of cancellation currents originating from the energy stored in the mechanical and electrical reactances of the circuit, and by impedance transformation effects in the crossover network. Which I believe to be plausible, and is personally how I rationalise |Z| being below the resistance of the voice coil for short periods.
Davis briefly restates this in his paper in different terms.
However, that is not to say that load impedance is ever 1R resistive - that depends on instantaneous load current phase wrt applied voltage - and hence the critical matter of interpretation. Load impedance |Z| = 1ohm does not mean a 1R resistive ohm. Ohm sweet ohm.
FWIW O&H also present that current peaks typically last a few hundred microseconds or are caused by the simultaneous parallel excitation of several drivers of multiway system, by the summation of cancellation currents originating from the energy stored in the mechanical and electrical reactances of the circuit, and by impedance transformation effects in the crossover network. Which I believe to be plausible, and is personally how I rationalise |Z| being below the resistance of the voice coil for short periods.
Last edited:
Again, "5 uSec is geologic""...can't happen.
""5 uSEc is trivial to calculate""
Now it's impossible to measure if it's not..
MAKE UP YOUR MIND. It's fun watching you argue with yourself..
Artificial, as in 20 feet of belden #12 awg and a speaker that can dip to 1 ohm?
Seriously??...
Learn what you are talking about. Your confusion is unbelievable, yet you continue arguing about something you do not understand..
Model a speaker impedance across an infinite bandwidth?? That's an entirely new strawman for you..Make some more things up..this is fun
You do understand what impedance means, no? Think about what a 1 ohm reactive load pulls with respect to current as opposed to a 1 ohm pure resistance. You need to look up phasor notation...did you ever learn it? It's standard content in a ScB course..
The discussion has always been about dI/dt dude.. Please learn.
Collectively, you have tossed a misconception, a flat out lie, and a misunderstanding of what reactance means. You blew it big time.
You need to learn phasor notation, you need to learn some control system theory, you need to learn some of what I learned in my rotational energy conversion course..
Google isn't going to pull you out of this ditch, dude.
Your thinking remains in the stone ages. You need to lean more if you want to discuss this rationally.
Why should I run a simulation using software developed by others?? You did, and yet you do not understand the ramifications. That has been one of my biggest gripes with the current state of education, everybody thinks simulations are the end all, but most haven't a clue as to what the problem is.
You very clearly are a perfect example of the problem. Rely on the tools without understanding the concepts.
hmm, loads equal to a resistor of 1 ohm...but, certainly not equal to 1 ohm.
How do YOU think an 8 ohm load can pull current 6.6.times larger than an 8 ohm resistor? What the heck do you think I've been talking about all this time?? At least your starting to think..
Again, what do you think this discussion is about, dude. It's dI/dt.. Look up phasor notation.
Again, what do you think I've been talking about all this time..
Welcome to the party pal.
BTW, none of that was the takeaway to this entire discussion..
jn
You simply misunderstand what 'dip to 1 ohm' means, in the context of a resistive load for your model - it doesn't mean a resistance BTW.
Now you're getting warm. So go work out what happens to current in the time domain when your 'step change' or audioband risetime applied voltage meets a reactive load with |Z| = 1 ohm and resistance of voice coil is 8R..........then contrast with the resistive load of 1R in your model, and you'll see.jneutron said:
Pretty warm now all you have to do is the calc or simulation......(usually this is where we seem to stall BTW)....jneutron said:
Yup that's warm - now apply a correct understanding and see what happens.jneutron said:
???jneutron said:
Because it's quick, easy and it works?jneutron said:
Doh, all gone cold again......jneutron said:
Now isn't that just the question ?jneutron said:
Last edited:
Just do it, JN and sweep for a range of plausible cable inductances. Not only will you realise your latency model is junk but you'll discover more interesting effects to wax lyrical about.............really.
So you have no idea what phasors are, do you? It was part of a core course in an undergraduate EE. You know, phasors, angles, currents, voltages..put one together for a 3 way system, show us how Otala and crew actually got that 6.6 factor. Then figure out clamping and foldback schemes needed to protect the output devices. The fact that you do not understand what a reactive load means to the cable current also tells me a whole lot..
Go back to the problem statement, and read what I said the pure resistive model provides. Go ahead, I'm patient..
What you really need, besides actually taking some courses on this stuff, is to get the hardware together and actually build and test something.
Hey, I know. Why don't you build the unit Cyril did, using his schematic, and his parts list, then watch the reflections caused by a short, an open, a resistor, and a reactive load on a 4 meter long cable. You know, transmission line theory?
Or, you can go and google the words I've given you.. tech I learned 40 years ago.
Or, you can continue to sim without understanding..
Oh, btw. You do know that 8 ohm speakers rarely have an 8 ohm DC resistance coil, right?
jn
If that is really your level of conceptual understanding here JN, this is never going to fly. Your loss.
Yes, often several ohms less. Makes no difference, just use whatever you feel is plausible, interesting part of the effect will be the similar.jneutron said:
Last edited:
But a resistive model does not apply - it does not represent a 'dip' in impedance here, especially when |Z| obtains as low as 1R in the context of Otala and Hutterman. So it is your mistake to use 1R resistive, and your model relies on it to obtain latancy c 5uS you claim.
If you adapt your model to use a plausible non-resistive |Z| = 1R model based on a realistic voice coil resistance, and apply audioband risetimes, you obtain very different results, and which are far more interesting per se IMO.
I understand well the maths/physics of what you've posted, JN - just don't agree. The simulations are impartial and confirm what I post. It's your mistake to assume that because I disagree I don't understand, and you've overlooked several key points that seriously undermine your model for reasons already set out.jneutron said:
Sure, everyone knows phasers are common directed-energy weapons first seen in the original Star Trek. Phasers release a beam of fictional subatomic particles called "rapid nadions".
In the original series the phaser was a PHoton mASER...........
- Status
- Not open for further replies.