You should never forget, any additional part is using up energy and stores it. So there is no advantage in using many parts in a crossover, if not absolutely neccessary.
The designer of the Alice 3.5A has intentionally used some neagative construction details to make it look like a LS 3/5a. He knows he did something wrong in the construction and has to to correct for it in the x-over. Because of that the x-over got so complicated. He is know to love such challenges...
"Donhighend" is an award winning speaker designer with decades of experience under his belt. He didn't start designing speakers last year.
Also, he uses any tool you can think of to measure and simulate, on top of his experience. He is a non compromise type of guy, living loudspeakers.
You are still using simple online calculators for x-over components. These work only with real resistors. A speaker is no real resistor, but changes its resistance with frequency. So these calculators are NO simulations!
Even good simulations only work if you feed them measured data of your driver in your speaker inclosure on your baffle. There is no "yes, but...".
Stealing values form existing crossover shematics doesn't work, it may only give you a very corse idea what a similar chassis may need in a compareable situation. A 1:1 copy will always be wrong.
If a crossover is used in a car, the values are absolute useless. Inside a car the conditions are so reflective, you can not compare them to a hifi speaker at all. You got to include the cars interior in the x-over, as you can not change the room.
The designer of the Alice 3.5A has intentionally used some neagative construction details to make it look like a LS 3/5a. He knows he did something wrong in the construction and has to to correct for it in the x-over. Because of that the x-over got so complicated. He is know to love such challenges...
"Donhighend" is an award winning speaker designer with decades of experience under his belt. He didn't start designing speakers last year.
Also, he uses any tool you can think of to measure and simulate, on top of his experience. He is a non compromise type of guy, living loudspeakers.
You are still using simple online calculators for x-over components. These work only with real resistors. A speaker is no real resistor, but changes its resistance with frequency. So these calculators are NO simulations!
Even good simulations only work if you feed them measured data of your driver in your speaker inclosure on your baffle. There is no "yes, but...".
Stealing values form existing crossover shematics doesn't work, it may only give you a very corse idea what a similar chassis may need in a compareable situation. A 1:1 copy will always be wrong.
If a crossover is used in a car, the values are absolute useless. Inside a car the conditions are so reflective, you can not compare them to a hifi speaker at all. You got to include the cars interior in the x-over, as you can not change the room.
Yes, except for non-resistive parts, which won't dissipate energy 😉
I'd be remiss if I didn't add that dissipating energy can be a healthy part of dealing with resonances.
PS This is a construction using the same drivers, just done right, not pressed into the limitations of the LS 3/5a:
https://www.lautsprechershop.de/hifi/wavemon_146_en.htm
You will see that by constructing a "correct" cabinet and baffle, you can cut the parts count of the crossover in half.
It would be a matter of a listening test to find out which ones bass response sounds better (to the individual listener in the individual room).
The vented cabinet will always profit from a high pass to reduce useless cone motion, while the capacitor of the LS 3/5a clone has it integrated.
PS "done right" may be a matter of taste and ability of the designer. The WaveMon was done by an evenly able person as the clone.
https://www.lautsprechershop.de/hifi/wavemon_146_en.htm
You will see that by constructing a "correct" cabinet and baffle, you can cut the parts count of the crossover in half.
It would be a matter of a listening test to find out which ones bass response sounds better (to the individual listener in the individual room).
The vented cabinet will always profit from a high pass to reduce useless cone motion, while the capacitor of the LS 3/5a clone has it integrated.
PS "done right" may be a matter of taste and ability of the designer. The WaveMon was done by an evenly able person as the clone.
@ Allen, in a practical world your are right, but if you are a scientist, there is no part that doesn't turn some electric energy into thermal energy.
Not even a capacitor or length of copper wire.
Not even a capacitor or length of copper wire.
but of course, the scheme pulled out of the drawer was only to understand decipher what type of cut he used. it is obvious that I am not going to take inspiration from that crossover to tune these drivers of this very modest project based on a recovery of drivers. I repeat, as I configured it does not sound bad, now it is a question of perfecting the invasiveness, however not excessive, but there is ... of this mid-woofer at 16ohm.
a doubt remains: the polarity inversion in the 2nd order, only on the tweeter or also on the mid-woofer?
later I will update you with the modification that I will test shortly.
edit:
I'll tell you in advance that I'm going with the 3rd order on the woofer and therefore I won't have any doubts about whether to invert the polarity or not.
a doubt remains: the polarity inversion in the 2nd order, only on the tweeter or also on the mid-woofer?
later I will update you with the modification that I will test shortly.
edit:
I'll tell you in advance that I'm going with the 3rd order on the woofer and therefore I won't have any doubts about whether to invert the polarity or not.
Last edited:
special treatment on the woofer with the 3rd order: almost perfect, now the voice comes out well and without the phase inversion even the bass hit is much better.
of course it's not just because I used the 3rd order, I also lowered the cutoff and brought it to 2850.
L1 = 1.4
C1 = 6.8
L2 = 2.8
in my personal experience it is a good step forward, but there is still one thing: to create a Zobel network, so I can also experiment to understand the effect on the woofer, to make the Zobel I have to insert two parameters - Le and Re - what would be the best way to obtain them?
of course it's not just because I used the 3rd order, I also lowered the cutoff and brought it to 2850.
L1 = 1.4
C1 = 6.8
L2 = 2.8
in my personal experience it is a good step forward, but there is still one thing: to create a Zobel network, so I can also experiment to understand the effect on the woofer, to make the Zobel I have to insert two parameters - Le and Re - what would be the best way to obtain them?
https://www.diyaudio.com/community/...overs-without-measurement.189847/post-2587236
If you don't have the data sheet, you can copy similar ones and tweak it. There's not much that can go wrong.
If you don't have the data sheet, you can copy similar ones and tweak it. There's not much that can go wrong.
I made a reasoning, tell me if I was wrong - looking at the page you linked me I noticed that the woofer was an 8ohm, so I doubled the values and this came out:
Re = 11
Le = 0.8
--
C1 = 6.6
R1 = 13.8
I'm alternating the Jeff Rowland and the Carver M1.5t, with the polarity inversion on the Tweeter, not on the Woofer.
I still have to understand why it sounds better than before 🙂
Re = 11
Le = 0.8
--
C1 = 6.6
R1 = 13.8
I'm alternating the Jeff Rowland and the Carver M1.5t, with the polarity inversion on the Tweeter, not on the Woofer.
I still have to understand why it sounds better than before 🙂
In a 2nd order crossover, only the electrical polarity of the tweeter is reversed.
This is done to ensure that the woofer and tweeter are acoustically in phase at the crossover frequency.
P.S. Other factors can come into play, so it may be worth experimenting with the electrical polarity of the tweeter.
it's the first time I see this thing, and it's the first time I do a Zobel 🙂
how do the values I used seem to you? the listening is much better
🙂
you see that the first time at the beginning of the project I was wrong - I had also inverted on the woofer. then if you remember we had decided on your suggestion to use only the inductance and at that point I had put the normal polarity back. now with 3rd order and the zobel on the woofer normal polarity. I think I got myself mixed up in some tests, now everything is ok.
What is important is that the filter balances the sound. If you listen and you like what you hear, you can keep those values. If you change your mind it is OK to try different values.. but try not to make the resistance very low for the amplifier.
do you mean that the resistance with too high values, puts the amp in difficulty right? like the Zobel - better if I lower the value of R1 and try to make a Zobel as if the woofer was an 8ohm?
I'm also ready to install audio simulation software with microphone and dedicated sound card. Run away while you can!
😀
I'm also ready to install audio simulation software with microphone and dedicated sound card. Run away while you can!
😀