And here is the phase plot of that model. If you want wilder, just reduce
the resistance in series with the LC elements.
Some say that when one posts something on a public forum, whatever he posts becomes public property and is free to be used. I'd still like your permission to use your load model, of course, with proper citing of whose it is, i.e by quoting you as the author.
As a long time (since 1974) owner of AR speakers, I am a bit obssessed with complex loads, I've had many an opportunity to hear them seriously tax many an amp. On the other hand, my own current speakers are not really suitable for load tolerance testing because they are an unusually easy and clean load to drive like no other speaker I have ever used, too easy to be practical.
That is in substance what I also said: with a crossover added and peculiar combinations of signal, anything is possible, in any direction, which means it isn't very useful from a general perspective, because you can prove or disprove anything and its contrary.
Here, there were some simple issues to answer, like "can a reactive network added to an ohmic resistance result in a lower impedance modulus", or "can the vector sum of three arbitrary currents exceed their simple arithmetic sum", and the answer is no in both cases.
If the study, design, and measurement of audio components and circuits was made exclusively using music signals as stimulus, things would become much more complicated and muddled, and the results would be much less deterministic: that is what subjectivists do, and in general, their results do not shine when scrutinized objectively
Not exactly.
https://stanford.edu/~boyd/papers/pdf/multitone_low_crest.pdf
Shows how to generate multitones with a low crest factor. This is an area of math that has received more attention in he past 30 years due the vast expansion of communication networks, which typically try to maximize the power in many signals that are amplified concurrently.
So you deny that any continuous time-varying signal can't possibly be composed of sines and cosines?
Interesting, and all this time I thought the findings of Fourier were credible! ;-)
I have never found Class-D amplifiers to my liking, this might be one of the reasons - Ferrite_bead_effect_on_Class-D_amplifier_audio_quality
Dan.
Dan.
mobile phone and hifi amp are very different beasts.
Can't see any problems here NAD Masters Series M22 power amplifier Measurements | Stereophile.com
Can't see any problems here NAD Masters Series M22 power amplifier Measurements | Stereophile.com
I've not come across a commercial classD amp where the output inductor's being pushed to 4X its current rating at peaks as they're doing in that paper. Doesn't make sense - of course its going to distort.
Granted, that Nad is latest generation and measures very well.
Earlier gear that I have heard including IcePower and Sony SRP-X500P on first listen sound clean, clear etc but ime longer term listening drives me out of the room.
I don't know the reason, maybe it's the nature of the low level noise, I need to revisit to put my finger on what it is that I find disagreeable.
Dan.
Earlier gear that I have heard including IcePower and Sony SRP-X500P on first listen sound clean, clear etc but ime longer term listening drives me out of the room.
I don't know the reason, maybe it's the nature of the low level noise, I need to revisit to put my finger on what it is that I find disagreeable.
Dan.
Trying random combinations of signals forever is not very useful. What is important is to understand the load and how it's response impacts the power delivery thing, the amp.
A simple swept sine testing of a speaker tells us exactly how it will respond to a non time varying sine. By looking at a typical reflex speaker response, you can easily see the impedance peak as well as the phase crossing zero at that same frequency. While it indicates the steady state sine impedance response of the system at every frequency, it does not clearly show the transient response of the system to that same frequency. At the peak impedance frequency, the system requires several cycles of energy buildup to reach the conversion efficiency shown in the sine response graph, and as it builds up, the impedance will rise as well.
What this states is simple. The amp is never tasked to provide continuous sines for most types of music, but combinations of sines which are temporally modulated, and we call music. Much is known about how to get that information to the amp, but many generalized assumptions have been used to model the speakers.
While everybody has their own taste with respect to their speaker, like reflex, horn, t-line, 4th order, all are still based on engineering generalizations. As I've stated, each type presents different transient load impedances to the amp, and generalized engineering loads may not be sufficient to model the system's total response.
If one thinks only of the impedance as that obtained by swept sine, one will incorrectly assume a V/I response consistent with a resistive load. Remember, one side of the impedance peak of a speaker is inductive in nature, the other side capacitive. One side has dI/dt issues, the other dV/dt. Too high a dI/dt into inductive requires more voltage, too high a dV/dt into capacitive requires more current. The amp has to work on both sides of the resonance.
So far, consideration of a speaker load as a sine measured impedance has allowed us to get very far. However, if a customer hears something different or somehow "amiss" with his combination of equipment, yet the engineering "assumptions" used to create the system says everything is ok, have we satisfied the customer? In this specific discussion, I've pointed out concerns with the generalized engineering models used; specifically how the transient use of a system engineered for steady state may not be sufficient.
John
That mkes two of us, Dan. I find that they usuall do bass very well, but lack qulity as we go up the scale. They definitely don't like high frequency content.
In my case, various manufacturers none of which I had heard of, names were totally unknown to me, and I wasn't impressed enought to even try to remember them. One was a Dutch product, that much I do remember. Hypex or something like that.
At this time, not interesting to me, my class AB gang is still ahead of them in terms of overall quality balance (frequency and power wise). Since my regular amp is an H/K PA-2400, rated at 170W/8 Ohms, the sheer power could never impress me.
At this time, not interesting to me, my class AB gang is still ahead of them in terms of overall quality balance (frequency and power wise). Since my regular amp is an H/K PA-2400, rated at 170W/8 Ohms, the sheer power could never impress me.
You just need to size the output inductor core appropriately.
That article is very misleading. If it gets into the hands of the hi-fi press, we will never hear the end of it - like feedback is bad, solid state is harsh etc.
No Doubt the 'Sandman' will be sitting there nodding his head sagely saying 'I told you ferrites were no good'
Many here, who use their ears, and trust their personal judgements often come up with what really works, and what sort of works, you might say. Class D amps have always been problematic with me, for example, except to bass, (maybe). Others have heard the same thing. At the moment I am working on a Class D amp, so any input from you all would be appreciated. No, not the digital, just the analog part, but I am now focusing on Class D as a design. Thanks in advance. Hear no difference people, need not apply. '-)
- Status
- Not open for further replies.