The 'how' is everything to do with it, and the cause of amplifier misbehaviours/changes according to non defined (reactive) loadings ie load dependencies. OLBW, SR, FR, DF(Output R/Z), PSU/OPS current capability etc etc all factor into the final real world result. A test protocol that discriminates and displays amplifier load dependencies over a range of defined representative simulated loads (or some kind of defined real loudspeaker loading and a defined range of cable types and lengths) would be very useful and give weight to at least some subjective opinions. It could also be used to derive the values of the networks required to 'straighten out' the impedance curve of any particular loudspeaker and confirm the results. An easy method to quantitatively remove amplifier/cable/loudspeaker load dependencies out of your system by means of passive networks across the loudspeaker terminals would be audio nirvana surely.
What is to attack about feedback....it is what it is.
Dan.
So you do agree that introducing an additional resistance to Re will increase Qes?
So you will also agree that introducing amplifier Rout will increase Qes and so will lower damping? (Grown-ups have defined damping ratio as being in an inverse relationship with Q).
What is there not to understand?
Exactly! Perfectly understandable. Now please tell the others.
The change in electrical Q changes the alignment and the alignment defines the damping, it has nothing to do with any DF. In fact, the so-called DF cannot be used in any way to predict the response shape, just state the output Z and let us figure out the rest. No brain surgery required, right?
So there are perfectly understandable mechanisms at work and we do not need any contrived DF number.
My turn to ask you a question, can the amplifier tell where the divide between its own output Z and the load Z? In other words, can the amplifier in fact only see a single impedance at any one frequency?
If the answer is no and yes, then you have proved DF to be less than relevant.
Please let me know about that one, I am asking in a courteous manner.
Model rockets for me. Estes and Centauri (anyone remember Centauri? They got acquired eventually). Why do you need these extra special "Photoflash" D-cell batteries to fire the igniters? They cost so much more! What does G. Harry Stine's book mean when it recommends you build your own ignition system using a "relay", what is THAT? Why do Nichrome igniters stay cool when the continuity bulb is in series, but glow hot and fire the motors when I mash the launch button which shorts the bulb? What is this trigonometry stuff that lets you figure out the rocket's peak altitude? Why does the radio transponder payload module have so many transistors and what are they doing?
I remember the first time I put one of my 'big' motors on and watched the rocket vanish. That'll learn me.
Of course being young 'how hard can it be to make rocket motor fuel'. So I tried to make gun powder. It did burn but not enough thrust to make it off the launch pad
I didn't buy the igniters I used matches with a strand of wire round them.
Of course being young 'how hard can it be to make rocket motor fuel'. So I tried to make gun powder. It did burn but not enough thrust to make it off the launch pad
I didn't buy the igniters I used matches with a strand of wire round them.
Seems we all started our my engineering careers with model trains. Me too, I bought a door bell transformer and four diodes and hardwired my first rectifier
You have made some interesting measurements, I know about them, because I too have done those kind of measurements and many more like them, more complex than yours. I could post them here, but I shall wait and do it in concert with other collaborators.
But it is the conclusions can be made from them, that is what is so intriguing.
Why does that tweeter measurement of yours show lower distortion under current drive compared to voltage drive. What is the mechanism that explains that? Is it because current drive is better than voltage drive? No! Or is it due to something else? Yes!
I am definitely in or near the nominal voltage drive camp, except I would say that just because you are using a voltage source does not mean that it is voltage drive. But current drive is current drive, no doubt there. But it does not explain why the lower distortion - and the answer is actually in plain sight.
Ask yourself questions about the current phase angle, then you would be close to understanding why the current driven tweeter had less distortion. The key here is that there is only one answer, and it is so surprising as it ought to be obvious - and there are now some serious people who know the answer, not just yours truly.
We are not allowed to say anything is new because that would make it a 'claim' and then you get assaulted. But in fact, there is nothing new under the sun, only a better understanding of what might well be there right in front of our faces!
At the risk of introducing a new concept into the tired circular arguments thought this might be of interest.
Ideal bridge rectifier GB
This has a possibility to be a better bridge. At least within limitations. Real measurable improvement with no woo and foo.
(aside any bets on when copies will appear on ebay?)
Ideal bridge rectifier GB
This has a possibility to be a better bridge. At least within limitations. Real measurable improvement with no woo and foo.
(aside any bets on when copies will appear on ebay?)
I'd put it differently, slighty more EE-oriented:
The electrical damping the cone movement undergoes is proportional to (BL²)/Re_eff, with Re_eff being the effective terminating impedance which is Re+Zout (Zout of the amp, and if |Zout|<<Re its influence ceases making the DF discuss moot for the most part).
The larger this quotient is, the more "self-controlled" the speaker reacts to external exitation, which includes its own errors and distortion. Low Qes (<0.3 or so) is degenerative feedback of sorts, the terminal voltage is mostly comprised of back-EMF voltage and when the amp applies a low (or even negative) impedance voltage on the terminals it means it controls velocity pretty much directly. BTW, that's why we see the 6dB/oct slope, coming from the pole splitting of the second order system highpass of a CB or free-air driver.
With current drive, no electrical damping, now the system is pure force-controlled as the current is directly related to the force that results. The current driven speaker doesn't even know what its cone movement is, you can block it with no change.
Yet both systems can be EQ'd to the same target and will show the same system response which dominantly "defines how it sounds". To first order, their output, seen in frequency and time domains, is exactly the same, under any reasonable operating conditions. However, looking closer at larger amplitudes and excursions, and looking on the distortion fine-print and things like excursion overload recovery, DC shifts, etc there can be significant differences and that explains the perceived differences.
In that light, going from a Zout=0.01R amp to a 0.1R ("lousy DF") one is surely insignificant to the fine-print stuff. OTOH, if it affects alignment considerably this is audible but then it is just the effect of a tiny linear frequency response change. When this small FR change is corrected for by pre-EQ, the differences are gone, and the 0.1R amp is not percieved to have any less control over the driver than the other one with 10x the DF value.
Joe, DF is a figure of merit which allows one to calculate the Zout of an amplifier if so desired. This Zout can be used to calculate effective Qes and this new Qes can be used to calculate the resulting response shape. No magic.
I don't quite understand your underlined question about seeing and telling amplifiers, but the words like 'output Z', ' load Z' and 'impedance' all have well defined meanings that are readily understood and which can be used to predict both the behaviour of the amplifier as that of the loudspeakers.
I don't quite understand your underlined question about seeing and telling amplifiers, but the words like 'output Z', ' load Z' and 'impedance' all have well defined meanings that are readily understood and which can be used to predict both the behaviour of the amplifier as that of the loudspeakers.
Claims without evidence are challenged that is only natural and the way it should be. In audio testimonials by happy customers are not evidence.
Joe, I thought you were going come back with some new material. The mechanisms for distortion reduction with current drive have been fairly well discussed for a long time. If you disagree please present the physics and math that prove them wrong. Actually measuring the distortion under different drive techniques and presenting hard data is not rocket science.
Last edited:
Sounds a bit pompous, wouldn't you say? Hope you have a sense of humour? But good to see you get that of your chest and... ?
Quoting text books is fine by me. To go any further and extend our understanding, that somehow becomes a 'claim' and here we go again. It is a verboten, even when I don't make any claim, often it magically becomes one?
But may I say something that might add to all that: Look closely at what the current does to the amplifier and something the amplifier cannot prevent when it is a voltage source. I know I am teasing, but I have no other choice, but as Pavel's measurements (and mine) proves that the current reflects the harmonic distortion of the speaker and distorts the current of the amplifier. Please, just give it some thought, drivers are current devices and Pavel has shown (and have I) that the current of the amplifier can be distorted because a voltage source can only control the voltage and must relinguish control over the current. Does this make any sense to you? If the answer is 'yes' to that, then you must make that compatible with other sound established matter. Alas, DF is not one of them. OK?
But thanks for the lesson, but do you think I really needed it? Maybe you felt a need to give it, then good for you!
Mark, I very well remember Centauri. My favorite rocket of theirs back then (the 70's) was named the 'SST', which was an Concord looking booster that carried a small piggybacked glider plane. At apogee, the glider would release and fly down, while the booster came down via parachute. Back in middle school, before we all discovered girls, my friends and I used to carry both Centauri and Estes catalogs around in our book bags. Much wasteful classroom day dreaming took place.
Balsa wood, sanding sealer, Testors paint, parachute wadding, rocket motors, launch control electrical - a nerd's paradise. And when those old style igniters failed to fire the motor, which they frequently seemed to, the suspense filled slow sneaking up on the unlaunched rocket like it was an live grenade. The blissful innocence of that time.
Last edited:
Those two sentences say opposite things. Which one do you believe?Joe Rasmussen said:
Clearly I have failed to explain this simple effect in a way which you can understand, but I don't know how to explain it in a simpler way. Woffling about two voltage sources is a diversion.
It helps if you understand the questions you are asking. This is often a useful precursor to understanding the answers. I have explained what DF means.
Given that it has a definition, and from it an estimate of the electrical damping can be calculated, I cannot imagine what you mean by "no meaning". Clearly you use the word "meaning" to mean something different from the rest of us. That makes discussion difficult. Maybe you have private definitions of words to add to what may be a private version of physical facts?
Yes, please show us the maths. Preferably in your own words; that would be better than merely quoting someone else who you may have misunderstood. My mind is sufficiently open to allow new facts in, but not so open that the existing facts fall out.
It is good that you have chosen your words carefully. That makes it harder for you to disown them when I say that your statement is the opposite of the truth. Every amplifier adds damping, it cannot decrease damping.
That is because it is true.
Your confusion is obviously of a profound kind. I don't know what to say to put you straight.
No actually it doesn't. Pavel is driving the speaker with a constant current it is not distorted or "reflecting" the speakers distortion or whatever you mean by that. Possibly Pavel could drive the tweeter with a conjugate matching network (speaker looks like a real resistive load) and show that the distortion does or does not go away.
What's going on can be expressed fully in engineering terms.
Last edited:
- Status
- Not open for further replies.