There is just one exception when "impedance flattening" might be helpful. If we make impedance flattening of each single driver in the multi-way speaker, to flatten the driver impedance, then the design of the passive crossover circuit may be much easier. However, this is not the case that Joe is permanently and repeatedly pointing at.
Parallel network is also useful both around resonance and high frequency impedance rise for multi amp current drive system.
Sure, however for the current drive it is obvious and self-explanatory and this was not Joe's point. It has to be repeated that Joe spoke about standard amplifier with voltage output as a voltage source and about impedance flattening of the speaker impedance (parallel to amplifier output) that would decrease speaker distortion. This is a nonsense, in case of normally designed solid state amplifier. I am afraid that this will be an endless discussion. I will only admit measurements at acoustical side and if flattening helps, then also measurements of the amplifier. Further debates for debates without a single proof are useless and only show incompetence.
Yes Richard, I know, sometimes you take yourself a little too seriously.
Trivia question for today:
What do Steely Dan and the Dept of Defense have in common?
'Skunk Baxter' of course.
YouTube
Yes I understand. This measurement really helps in making design choices for my low DF hybrid multi amp system. Again thank you, you have no idea how much help this information gives me. 🙂
I had enough of that, I thought I can lend a bit of help to him. I stop at once when the truth of the matter became quite evident.
Last edited:
We already discussed caveats of acoustic distortion measurements before, several times. Room reflections plus directionality. As the measurement was done close to the woofer (15cm), for the reason to minimize room effect though not overloading the microphone, useful and meaningful frequency range is heavily limited by cone diameter and driver acoustical interactions. Measurement of distortion at distance 1 - 2m in the room makes no sense. However, I can make you sure that even the "full bandwidth" distortion measurement, regardless its imperfections due to microphone position, is again almost same with and without the "flattening" resistor.
Crossover schematics is unimportant here, the main info is the change in impedance as described before, because this is what was suggested - that "impedance flattening" at amplifier or speaker terminals should reduce speaker distortion. This did not happen, not only in the 300Hz range, but also not in 20kHz range.
Instead, I am posting a test schematic for you. Placement of resistor (R1 or R2) has had no effect on results measured.
Attachments
I have been unable (or unwilling) to dredge though the last hundred pages or so so I may now speak out of context. 40+ years ago when I first started building speakers I learned the turnover frequency of an LC filter is changed by it's termination which a correctly designed Zobel at the driver output stabilizes. This implementation of a Zobel network can have a significant impact on the sound of a system. Is this not the usage people have been talking about here?
Howie
No, this is completely different story. That's the problem of the forum discussion. The matter suggested here by one member is that an additional R or R-C, parallel to amplifier out, resulting in "speaker impedance flattening", should have decreased the speaker distortion, as the promoter assumes that the amp has troubles to drive complex load.
Zaph|Audio - ZRT - Revelator Tower is a good example as he shows the conjugate network on the tweeter to make it behave. In this case its a known bad behaviour on the tweeter side so no contention I would hope.
I assume this is what Howie is talking about, you design an L/C crossover network for a nominal termination and if the driver has a large resonant deviation you need the Zobel to tame it. This is common AFAIK.
An LC crossover does a reasonable job of dealing with moderate variations without compensation, if it is designed for what it sees at the knee. It is somewhat self-contained in its behaviour.
Good point, and likely most applicable to 4th and higher order crossovers. Unfortunately the vast majority of reasonably-priced speakers and many expensive ones have a 2nd order LC for either the LF or HF or both. Consider at resonance driver impedance can increase by factors of 4 or more. Under this type of variable loading these networks can show 10% or greater corner frequency modulation as a result of driver impedance variation with frequency.
Audibility of this is certainly up for debate, With smaller speakers (4-6" LF drivers) I have found the audibility greater than for larger drivers...and have not yet investigated why. Maybe some day when I 'retire' I will...
My 2¢ worth,
Howie
Most of what Joe says can be falsified just by using electric circuit theory. In that sense, much of what he says about speakers is 'not even wrong'.john curl said:
Not true. I believe in gravity waves, although I have never personally detected one. I believe in electrons, although I have never seen one. Can you guess why I believe in such things in spite of my lack of personal experience? And why I do not believe in some of what you or Joe say?
Scientific intuition needs to be trained. Even if you eventually overthrow the existing knowledge you first have to know what it is in order to know what is wrong with it. In most cases this is not what happens; instead the existing knowledge is extended so the old knowledge becomes a special case. Quack science rarely bothers to actually learn the existing knowledge; instead it just rejects what it does not understand.Joe Rasmussen said:
Nobody has disputed that equation. It is what you infer from that equation which is incorrect.
I note that you are back to using simple real arithmetic to handle complex quantities. It helps us when you display your ignorance of basic engineering maths so clearly.
😀
😀
Just sayin' yes, a very serious business. So, I have a very different perspective than others, I suppose.
Back to my question. And, If it was positive fb as JC thought it looks like, how could that reduce THD? The amplifier would need a neg fb to reduce THD. So maybe , just maybe? there is a phase change at resonance which provides that neg fb. at least partial (90 degrees).
And. If you added a C appropriately such that the freq and Z allowed an additional 90 degrees, you will get further/deeper reduction in THD.
🙂
-RM
😀
Just sayin' yes, a very serious business. So, I have a very different perspective than others, I suppose.
Back to my question. And, If it was positive fb as JC thought it looks like, how could that reduce THD? The amplifier would need a neg fb to reduce THD. So maybe , just maybe? there is a phase change at resonance which provides that neg fb. at least partial (90 degrees).
And. If you added a C appropriately such that the freq and Z allowed an additional 90 degrees, you will get further/deeper reduction in THD.
🙂
-RM
Last edited:
You're the first person from Livermore that I know brags about having higher clearance, everyone I met that commented on clearance told me I was lucky to not have to deal with all the extra rigamarole. Then again you might have retired before all the electronic protocols won't into place which made everything so much more "fun".
Depends on what you do there. Even a secretary has to have a Q clearance if she supports workers in some classified building/area. The fun was in the work being done which required the Q clearance. Protocol is something else. Dont know about how much more they have added. But, protocol creep continues, i am sure. Mostly it is about how information is handled and transferred. I had minimal to do with that... other than not talking to anyone not involved or who didnt have a need to know.
-RNM
-RNM
Last edited:
Richard, I think that your years with LLL is underrated here, by those who appear to envy you. I know that Jack Bybee had similar restrictions to his travel, back in his time of doing work for the government. This new thing of his is confidential, not classified, because it does not involve his old work.
As I stated earlier, I went back though the chapters on direct radiators in both Olsen and B-------- (can't spell it from memory) and they match what Joe put up on the differences between voltage and current drive yesterday. What this implies with adding further passive parts, I do not know. But, as it was criticized immediately by the usual 'critics' it shows that they have not read up on loudspeakers yet.
As I stated earlier, I went back though the chapters on direct radiators in both Olsen and B-------- (can't spell it from memory) and they match what Joe put up on the differences between voltage and current drive yesterday. What this implies with adding further passive parts, I do not know. But, as it was criticized immediately by the usual 'critics' it shows that they have not read up on loudspeakers yet.
- Status
- Not open for further replies.