Reading a lot more. Besides schooling would have been a great benefit for such a rude and ignorant person.
Why are you trying to scholar someone someone if your 'knowledge' just a guess? See, it's not I mind if I'm wrong, I just don't understand why someone so vehemently insists I'm wrong. Please don't see it anything against you but that has happened to me a lot before. What's driving you to be so insisted on to me not knowing xy?
I freely admit if I'm wrong and I am happy to learn something new.
Well, that's not true. If the non-linearity consists of amplitude and/or phase differences, there's not necessarily IMD. Read up on the Klippel papers.
I don't believe that this is really the main point to be made, but rather the relative levels of harmonic and modulation distortion... (see first article below). Modulation distortion is much more audible. Human "just detectable distortion" tests that add-in only harmonic distortion I think miss the point that the main source of audible distortion is usually in the loudspeakers, and it's modulation distortion--that is inharmonic--that we hear so acutely, and usually the lower sidebands of the modulation distortion as evidenced by the human hearing system's masking properties that do a good job masking the upper sideband distortion frequencies.
Chris
I used LTspice to simulate the effect of about 1 % of second harmonic distortion on a signal with 1 Khz and 10 kHz equal amplitude signals and the intermodulation components can be seen in the FFT plot. It shows a relationship between HD and IMD.
I don't doubt the relationship in analog circuits but I was talking explicitly about actual sound reproducing drivers and horns - which differs a lot from pure electronics. I appreciate your attempt to clear things up and I don't doubt the results but that doesn't adress the issue we're discussing right now. Sound reproducing drivers behave a lot differnt from electronics, especially if there are wave guides or horns are involved.
If it's mechanical or electrical is totally irrelevant and rather pedantic.
There are always equivalent structures that can be translated from one to another.
It's far more important to understand the underlying concepts imo, instead of trying to fight over little details.
That makes any discussion incredibly vague and blurry.
Just respond with arguments, instead of creating a hostile, disrespectful and unsafe environment please.
There is nothing rude about my comments, I just respond to the subject.
I can't predict how that translates to your specific culture, language and personality.
So there is absolutely ZERO reason to get personal about technical subjects.
Thank you and I am also gonna leave it at that.
It's my daily work as well as my professional field, so why exactly should I read up on it?
I am reading on all this material on a daily basis basically.
Maybe think about the idea that miscommunication is also a possibility?
That's why people shouldn't get their information from forums, but start reading proper literature.
But apparently that is considered "rude"
Makes no sense.
Whatever, I personally couldn't care less, I thought I would just share some insights.
But as somebody else already mentioned, give this topic a couple of pages.
So I think I am gonna take that advice as well.
I am not here to be convinced by people, I have better resources for that, again no offense.
Yes, frequency make a different but that much.
By principle high frequency harmonics of distortion is much harder to detect since they come close to what we are able to hear.
Here you can see my results...
Uhm, no, it's not irrelevant - high mechanical reasons usually result in decay ridges and probably phase shifts, which electronics do not. And mechanical reasons can't be fixed by applying a filter or change on feedback etc.
Well, you went on your high horse and said I have to read up while I wasn't in the wrong. So either you switch off your knowledge on leaving the office? ..or..
Ah, that must be it! And I thought you saying I have to read up couldn't be any clearer! Ha! How silly of me!
Well, that's exactly how you treated me.. ? But you are right, as long as renowned experts write something in a forum, it MUST be right, ofc. Yeah, makes no sense.
Well, I'd prefer if you'd save the drama and simply admit being wrong. Or just not post. Because that's why these threads end.
Well, maybe try to use these resources.. ? ..no offense..
I have a question : what means AMD (see below) ?
Now personally, if I measure a 1% distortion with a good harmonic decomposition (we call it dégradé harmonique in French), it's already fine...
T
Now personally, if I measure a 1% distortion with a good harmonic decomposition (we call it dégradé harmonique in French), it's already fine...
T
It might be worth remembering that HD and THD and IMD numbers are measurements, and not characteristics of played back music. They are from setting up a relatively simple mathematical circumstance and seeing how badly unexpected frequency components get produced, and in a way that yields a repeatable number. You don't "listen to music at 0.01% THD distortion" or like that, the circumstances aren't the same or even similar to the test.
And in speakers and even some analog electronic circuits the chosen circumstance can "hide" effects of nonlinearity. A good example is looking for components at a frequency that a speaker driver just doesn't radiate much at or that it emphasizes. Distortion numbers give engineers a (rather synthetic) handle on something they can look at while developing or tweaking a device or circuit to indicate improvement or degradation in some way. Preferably without getting confused by the hidden nature of some effects.
In my mind, THD or IMD numbers are primarily helpful as a way as to specify output level capability of a broadband circuit, like "100W 8 ohms at less than 0.1% THD". Thought even that can lose some value in cases like amplifier crossover distortion or output low pass filtering in a class D amp.
And in speakers and even some analog electronic circuits the chosen circumstance can "hide" effects of nonlinearity. A good example is looking for components at a frequency that a speaker driver just doesn't radiate much at or that it emphasizes. Distortion numbers give engineers a (rather synthetic) handle on something they can look at while developing or tweaking a device or circuit to indicate improvement or degradation in some way. Preferably without getting confused by the hidden nature of some effects.
In my mind, THD or IMD numbers are primarily helpful as a way as to specify output level capability of a broadband circuit, like "100W 8 ohms at less than 0.1% THD". Thought even that can lose some value in cases like amplifier crossover distortion or output low pass filtering in a class D amp.