It is impossible to make a 3-way loudspeaker with using only natural rolloffs of the (8-ohm) drivers, not because the load would be too low (2.67 ohm), but because there are no such drivers. You can not make a 3-way loudspeaker without a crossover.
Three 16-ohm (fullrange) drivers parallely conected makes a loudspeaker with 5.33 ohm nominal impedance - hardly a problem for any amplifier.
and here I was thinking you had modified DIY style... dang it!
I tried connecting the woofer to the woofer, but it didn't work very well.
Ah, ... but you did.You said so the very moment you stated
because, you know, "3 way speaker" implies use of crossovers.
Won't even mention that you are posting in the
subforum
And everybody understood it that way, such as:
etc.
By the way, don't shoot me, I'm your friend, my post was directed to asking fellow Members to answer your main question as straight as possible (which I did
) without wandering into : non flat impedance , other ways to connect the components, TS parameters, Spice simulation, actual measured SPL data, different impedance components, etc.All fine and good in due time, and deserving their own juicy threads, but irrelevant to you if you first don't have your original question answered.
You bet, there's some people who asks something and then gets angry at the correct answers.
Oh well.
Back to your last question: putting 3 different speakers in parallel is almost never done, because it's extremely wasteful.
The amp will have to drive a much lower impedance load, so it will overheat and maybe have current handling problems, the speakers will waste power delivered to them outside its best efficiency range and small light speakers (midrange and tweeter) will be destroyed by power levels easily handled by the woofer.
So it's a bad idea on many counts.
Why did I say "almost never"?
Because there is one very specific application where it's done with good results: guitar speaker cabinets.
There it's relatively common practice to combine different voiced speakers (say a relatively scooped Greenback with a muscular mids V30) to getb a special combined curve which pleases the Musician.
But there they are usually same sized speakers (say 12") , roughly same range, just with different "voice", reasonably matched power handling (or it's acceptable to let the weaker one define total cabinet power) and parallel or series impedance is no big problem for amps which usually have different impedance taps.
But none of this applies to your (presumably) Hi Fi cabinet, was just an example of "the exception of the rule".
I know, I also use that trick all the time.
But please consider 2 relevant points:
1) all electromagnetic speakers, but very specially woofers, have a built in series inductor, and its value comes not only from the actual inductance of the voice coil (which physically is a coil of copper (or aluminum) wire wound around an ... iron
core : the polepiece) but also from the moving mass, which appears reflected as an extra inductor.If that inductor (and I don't enclose it between "" because it's quite real, not virtual) is enough and approppriate for the Xover frequency chosen ... just use it, why add an external one?
But that doesn't mean "there's no inductor" at all.
It can be measured and does its job.
Among which is to start decoupling the woofer from the amplifier above the crossover point.
Which is the concept I wanted the OP to understand
2) "The other" component, in this case a Tweeter, also has a series decoupling crossover element, at least a capacitor.
The main point is that both speaker elements "are not in parallel, halving the cabinet impedance" , which the OP seemed to believe at the beginning.
Hope everything is clear now
There is no doubt that one COULD make a 3-way loudspeaker system without a crossover, making use of the natural roll-off of the drivers. However, getting anything remotely like flat frequency response from such a system I agree would be exceedingly difficult. It might be an interesting design to try. Who knows, there might be something to learn from making the attempt.
No way! At least using drivers existing on Earth.
Not sure for drivers from planet Unobtinium.
Last edited:
...There's an italian maker who sells loudspeakers that says he makes "crossoverless" LS but, when asked, he said he puts a capacitor for protection on the tweeter path.
Arte Acustica
Arte Acustica
I know, but "sending dust" to the eyes of the audiophiles is an old technique too. For cheapness is it ok for 3000 € the two way "tower" ?
Or would you like to spend >15000 € for the " dolly-basketball-style " one ?
Ah, purity ...
(I think they're the only speakers I've heard at a demonstration 3 years ago, not bad in a fully analogue chain...ah, purists! )
Maybe one could find an analogy in the force/momentum formula and the pistonic radiation which limits the band..
Or would you like to spend >15000 € for the " dolly-basketball-style " one ?
Ah, purity ...
(I think they're the only speakers I've heard at a demonstration 3 years ago, not bad in a fully analogue chain...ah, purists! )
Maybe one could find an analogy in the force/momentum formula and the pistonic radiation which limits the band..
Last edited:
Please reread the full OP.
Read it line by line, because they are not simultaneous, they have a sequence.
He starts with a question.
He does not wait for an answer but:
In the next phrase, he turns it into an assumption at the beginning of the phrase, so he can post a new question at the end of that same phrase.
See for yourself:
First phrase, yes, it's a question:
Second phrase, an extension of the first, he just makes clear what he means by "total":
Third phrase, a question:
The drivers are connected in parallel ?
He does not wait for an answer.
Fourth phrase, starts with an assumption which is necessary for the ending question to make sense and be answerable.
"If" does not mean doubt but he's setting the parameters of the problem to be solved, so it's not a question but an assumption or a statement:
And this is the final question:
My point is that in the second part of his post he defines them as parallel and asks about the numerical result.
The result would not be right or wrong, wouldn't even be possible, if he had not defined first the parameters of the problem.
Whew, everyone needs to take a chill pill here!
A vague question always seems to start these sorts of debates. A vague question deserves a vague answer.
https://www.youtube.com/watch?v=L0PZvvt36DA
The video is aon Laplace transform circuit analysis. The beauty is that this answer is correct no matter what the OP means
A vague question always seems to start these sorts of debates. A vague question deserves a vague answer.
https://www.youtube.com/watch?v=L0PZvvt36DA
The video is aon Laplace transform circuit analysis. The beauty is that this answer is correct no matter what the OP means
Last edited:
I suppose you get an idea of impedance from looking at a circuit, the drivers and the bass loading, with experience.
But it's better to model it. Boxsim will work it all out for you, except for the phase angle:
Downloads
The red curve at the bottom. Try putting some of these designs into the projekte folder:
boxsim-db.de | Boxsim Projektdatenbank
Low impedance and high phase angle at any point is the combination to avoid. Amps don't like it.
But it's better to model it. Boxsim will work it all out for you, except for the phase angle:
Downloads
The red curve at the bottom. Try putting some of these designs into the projekte folder:
boxsim-db.de | Boxsim Projektdatenbank
Low impedance and high phase angle at any point is the combination to avoid. Amps don't like it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.