No, it linearize-it. The target value is the DC resistance of the lowest R speakers.
Musical notes are not pure sinus. They are fundamentals and harmonics. It seems obvious it is better the amp can treat all of them the same way (same Voltage/current ratio).
So they incorporate at least impedance compensation networks, a way or an other.
I use a two way speaker, with a horn. The resonance of the horn driver is at least one octave lower than the crossover frequency.
And my filter is 24db/oct.
As the efficiency of this driver is 10bB upper than the bass driver, there is a attenuator in the high filter (oh, yes, again a serial resistance ;-).
Difference on the medium is obvious, there too, with and without the 'motional compensation'. Compensation network removes 'horn' hardness. Both in the active filter and passive versions.
End of my inputs on this out of topic subject.
I have a big file with all informations. I can not post it here because of size. It is in PDF form and i am to stupit to make a zip.
No, i do not need additional components. Actually the circuit is rather simple.
No, i do not need additional components. Actually the circuit is rather simple.
Thanks Christophe!
I am on your side even if other want to discount what you are saying. I imagine what Joachim is doing is just moving the circuit into the XO network but I am looking to use active networks and then that solution does not work. I would still have the passive conjugate compensation and have the active filter network be before the amplifier circuit. It is hard to understand why the simple flattening of the impedance curve is so hard to understand and how that would linearize the load. It would be like saying that the bandwidth or the slew rate wasn't important in an amplifier as long as it makes output from 20 to 20khz and falls on its face at 30 khz because we can't hear that! What would that do to your in bandwidth response curve and phase shift, silly isn't it.
I am on your side even if other want to discount what you are saying. I imagine what Joachim is doing is just moving the circuit into the XO network but I am looking to use active networks and then that solution does not work. I would still have the passive conjugate compensation and have the active filter network be before the amplifier circuit. It is hard to understand why the simple flattening of the impedance curve is so hard to understand and how that would linearize the load. It would be like saying that the bandwidth or the slew rate wasn't important in an amplifier as long as it makes output from 20 to 20khz and falls on its face at 30 khz because we can't hear that! What would that do to your in bandwidth response curve and phase shift, silly isn't it.
Sorry, but what I write is based on my experience It may be pure chance but it is in perfect accordance with those who have made the electronics to accomplish its greatest steps.
Well, it is to you to prove the effects you expect.
No need for a B&K apparatus (by the way, I own two B&K mics to remain connected to the real world).
Nowadays, a sound card, a microphone and some software replace it perfectly.
Usually called conjugate matching, I think.
I am very familiar with another concept giving far more observable effects, it is called "motional feedback".
I read that as well as the other comments hundreds of times.
The too known demonstration of a driver with the voice coil connected to a very low impedance only concerns an effect on the main resonance only, it is always forgotten.
If a really high damping what is required, I use a negative impedance output (I experienced it) it's far more spectacular than just shorting the voice coil.
Material of cone resonances, breakups and microphonic effects away from the main resonance are better handled by current drive, there are some nice discussions about it here.
Last edited:
Forr,
If you have a B&K microphone I would assume that you do understand, and I know of no other mics that have the flat frequency response of the B&K out to 100khz, no cheap plastic diaphragmed mic can come close to the accuracy of the metallic diaphragm of a B&K. So yes you can do some measurements with a sound card and cheaper mic but not nearly as accurately or to anywhere near the frequency range. Yes I know you can put a compensation curve into a program to make it appear that the frequency response is flat and accurate but it just isn't the same and any other mike will change with the humidity, temperature and atmospheric changes.
And a conjugate network is not the same thing as placing a simple resistor across the output terminals of a speaker. LCR does not equal R.
ps. Christophe, don't worry you don't have anything to prove to me. If they don't want to check for themselves so be it. Instead of let's try something, it is you prove it first. You have already done that for yourself and I know others who have done the same. They actually get paid for designing speakers and networks, I'll leave it at that.
If you have a B&K microphone I would assume that you do understand, and I know of no other mics that have the flat frequency response of the B&K out to 100khz, no cheap plastic diaphragmed mic can come close to the accuracy of the metallic diaphragm of a B&K. So yes you can do some measurements with a sound card and cheaper mic but not nearly as accurately or to anywhere near the frequency range. Yes I know you can put a compensation curve into a program to make it appear that the frequency response is flat and accurate but it just isn't the same and any other mike will change with the humidity, temperature and atmospheric changes.
And a conjugate network is not the same thing as placing a simple resistor across the output terminals of a speaker. LCR does not equal R.
ps. Christophe, don't worry you don't have anything to prove to me. If they don't want to check for themselves so be it. Instead of let's try something, it is you prove it first. You have already done that for yourself and I know others who have done the same. They actually get paid for designing speakers and networks, I'll leave it at that.
Last edited:
???
Never seen an attenuator in front of a compression driver ????? Are-you serious Joachim?
In fact they don't need any dumping from the amplifier because they are damped a lot by the air compression and the phase plug, and every competent speaker designer knows they show less distortion when driven in current (high impedance sources).
Don't you understand what i wrote ?
A modern driver with impedance compensation does not care about voltage or current drive. No, i do not understand you.
Impedance compensation as in a Faraday ring? It does help tremendously but there will still be a low slope impedance rise.
Yes, compression drivers show less distortion with current drive because the impedance curve is all over the place. Driving them with a high impedance output amp requires extremely complex frequency response compensation.
NO I HAVE NEVER SENN AN ATTENUATOR. Of cause i HEARD about it.
You know, i am blind. Of cause you do not know that so i can not complain.
NO I HAVE NEVER SENN AN ATTENUATOR. Of cause i HEARD about it.
You know, i am blind. Of cause you do not know that so i can not complain.
Kindhornman, i said IN THE WORKING RANGE, i am not a native speaker.
Who whould use that Satori driver over 2 kHz ?
Who whould use that Satori driver over 2 kHz ?
Joachim,
It is amazing what some on these forums try to do with speakers. The full range guys are something else. That driver looks promising for some uses, though I would not use an overhung design, I consider that a non linear device, no way is the magnetic system completely equivalent on both sides of the gap, never happens no matter how you try and the stray fields from the overhung coil make it non linear in that regard.
I love the comment about the Egyptian papyrus, that is truly funny. And I will not use a rubber surround, I do have my reasons for that but won't get into that in this forum.
It is amazing what some on these forums try to do with speakers. The full range guys are something else. That driver looks promising for some uses, though I would not use an overhung design, I consider that a non linear device, no way is the magnetic system completely equivalent on both sides of the gap, never happens no matter how you try and the stray fields from the overhung coil make it non linear in that regard.
I love the comment about the Egyptian papyrus, that is truly funny. And I will not use a rubber surround, I do have my reasons for that but won't get into that in this forum.
I am in the advanced user group of Klippel. I have a Klippel Distortion Analyser since 2001.
We even cut away 90% of the rubber suspension and the distortion of the spider remained the overriding feature.
This " foam suspension is better " is bullsh......http://www.klippel.de/uploads/media/Klippel_Nonlinearity_Poster.pdf
We even cut away 90% of the rubber suspension and the distortion of the spider remained the overriding feature.
This " foam suspension is better " is bullsh......http://www.klippel.de/uploads/media/Klippel_Nonlinearity_Poster.pdf
I never mentioned a foam surround. I could take that drawing and run it in a magnetics program and I have no doubt it is none linear by design. Sorry we do not agree but that is how speaker design goes.
And I agree that spiders are a problem but will keep that work to myself right now. But do not discount the acoustic signature of surrounds.
I just find it funny that one type of cellulose fiber is superior to another. It is the combination of the entire compound, and most will have reticulated synthetic fibers in the mix plus the binders used. It is the entire matrix that is important. I do think that paper type cones can be excellent, I just happen to be working with composites, but not something that is commercially available anywhere else in the world, it is my own creation.
And I agree that spiders are a problem but will keep that work to myself right now. But do not discount the acoustic signature of surrounds.
I just find it funny that one type of cellulose fiber is superior to another. It is the combination of the entire compound, and most will have reticulated synthetic fibers in the mix plus the binders used. It is the entire matrix that is important. I do think that paper type cones can be excellent, I just happen to be working with composites, but not something that is commercially available anywhere else in the world, it is my own creation.
Last edited: