Hello!
I'm new here. This is my second attempt at building my own speakers. My first set of speakers was a replacement for my father's Fishers that were awful 😕 . That was 8 years ago, and they ended up good, I guess. I made a lot of mistakes but they sound good, even great compared to the store bought 12" floorstanding and old Fishers.
Anyway, that's not the reason of this thread. I'm now building a 3-way bookshelf speakers, but there is a DISCLAIMER: where I live, I can't source any drivers, in general. There is no market here for DIY audio, and if I find something it usually cost between 2 and 3x the MSRP.
Now that I have spent days designing the crossover, I came to the conclusion that the midrange that I selected was probably a bad choice for this build, but it was the driver that I had available to buy and could afford. So, I'm trying to make it work.
I have two designs, I'm going to start with the last one that I did, because it's making more sense, I think. I'd love to get your input on what I am doing wrong, and what should I change.
The drivers that I have are:
Woofer: HiVi M6N 89dB 2.83v/1m 8ohm
Midrange: HiVi B3N 81dB 2.83v/1m 8ohm
Tweeter: HiVi SD1.1-A 91dB 2.83v/1m 6ohm
So, in this attempt I first started by lowering the sensitivity of both, the tweeter and the woofer, to match the midrange. I then placed a 3-way 2nd order crossover at 500/4000. Then modified the tweeter highpass to -18db/octave, added a elliptic notch filter to the midrange, and tuned the values of each component. The end result is this:
I'm not so sure about this design, primarly because of the L-Pad on the woofer. If I look at power dissipation, I see 22.5W on R6 (woofer) and only 6.5W on the woofer itself, while the midrange is at 28W, and the tweeter resistor peaks at 55W (!). I think I'm looking at this wrong, mostly the tweeter resistor, that is after the crossover and probably the power won't be this high.
The reference rating is at 8.375, I'm really not sure what this means, but I concluded it has something to do with linearity.
On the other hand, I have my first design. On this one, I made no efforts on matching the sensitivity, at least not on the woofer side, but did lower it on the tweeter. The issue with this design is that it was impossible, for me, to make it work with the crossover at 500/4000. So I started to move the values of each component, to have something that looks right. It ended up looking right, at least on the power response and SPL, but there is something that I believe is simply wrong: there is much overlap between the woofer and the midrange, and between the midrange and the tweeter (and the crossover values are all over the place mostly). The midrange, having a very low SPL compared to the other drivers, is just "adding up" to the response of the other two... and this seems wrong to me.
The power dissipation looks much better thou:
The reference rating on this design is 8.471, and it "looks" good. I was going to build this, at first, but now I'm afraid that even thou it looks it's going to sound right, it wont. The big difference on driver SPL made me ditch this design, but it was much simpler, less components, and with my extremely limited knowledge looked great (until it didn't).
So, I was wondering if someone could help me understand the pro and cons of each design, the major mistakes I'm making that I'm sure there are, and what to tune to make this drivers work together.
Thanks for reading, and regards from Argentina.
I'm new here. This is my second attempt at building my own speakers. My first set of speakers was a replacement for my father's Fishers that were awful 😕 . That was 8 years ago, and they ended up good, I guess. I made a lot of mistakes but they sound good, even great compared to the store bought 12" floorstanding and old Fishers.
Anyway, that's not the reason of this thread. I'm now building a 3-way bookshelf speakers, but there is a DISCLAIMER: where I live, I can't source any drivers, in general. There is no market here for DIY audio, and if I find something it usually cost between 2 and 3x the MSRP.
Now that I have spent days designing the crossover, I came to the conclusion that the midrange that I selected was probably a bad choice for this build, but it was the driver that I had available to buy and could afford. So, I'm trying to make it work.
I have two designs, I'm going to start with the last one that I did, because it's making more sense, I think. I'd love to get your input on what I am doing wrong, and what should I change.
The drivers that I have are:
Woofer: HiVi M6N 89dB 2.83v/1m 8ohm
Midrange: HiVi B3N 81dB 2.83v/1m 8ohm
Tweeter: HiVi SD1.1-A 91dB 2.83v/1m 6ohm
So, in this attempt I first started by lowering the sensitivity of both, the tweeter and the woofer, to match the midrange. I then placed a 3-way 2nd order crossover at 500/4000. Then modified the tweeter highpass to -18db/octave, added a elliptic notch filter to the midrange, and tuned the values of each component. The end result is this:
I'm not so sure about this design, primarly because of the L-Pad on the woofer. If I look at power dissipation, I see 22.5W on R6 (woofer) and only 6.5W on the woofer itself, while the midrange is at 28W, and the tweeter resistor peaks at 55W (!). I think I'm looking at this wrong, mostly the tweeter resistor, that is after the crossover and probably the power won't be this high.
The reference rating is at 8.375, I'm really not sure what this means, but I concluded it has something to do with linearity.
On the other hand, I have my first design. On this one, I made no efforts on matching the sensitivity, at least not on the woofer side, but did lower it on the tweeter. The issue with this design is that it was impossible, for me, to make it work with the crossover at 500/4000. So I started to move the values of each component, to have something that looks right. It ended up looking right, at least on the power response and SPL, but there is something that I believe is simply wrong: there is much overlap between the woofer and the midrange, and between the midrange and the tweeter (and the crossover values are all over the place mostly). The midrange, having a very low SPL compared to the other drivers, is just "adding up" to the response of the other two... and this seems wrong to me.
The power dissipation looks much better thou:
The reference rating on this design is 8.471, and it "looks" good. I was going to build this, at first, but now I'm afraid that even thou it looks it's going to sound right, it wont. The big difference on driver SPL made me ditch this design, but it was much simpler, less components, and with my extremely limited knowledge looked great (until it didn't).
So, I was wondering if someone could help me understand the pro and cons of each design, the major mistakes I'm making that I'm sure there are, and what to tune to make this drivers work together.
Thanks for reading, and regards from Argentina.
Hi @agb86 and welcome!
try the following:
and then reconsider both projects.
try the following:
- include bafflle step influence for the driver responses (will mainly lower effective SPL of woofer and make crossover to midrange easier)
- don't put L-pads in front of all drivers - mostly woofer (and your low SPL mid)
and then reconsider both projects.
Last edited:
Thank you @stv !
I have removed the L-Pads on midrange and woofer, as it makes more sense. I'm playing with the baffle step compensation, but I'm not sure about including it (another big inductor on the design).
I was using air core inductos everywhere, is there much difference replacing the woofer crossover and baffle step inductors to laminated steel core?
With the baffle step at -8dB:
Whitout it:
I'm not sure if it makes much of a difference really.
And I have a great dip in response at 2000Hz but I don't think I can make much about it.
Anyway, thanks a lot. This is making a bit more sense.
I have removed the L-Pads on midrange and woofer, as it makes more sense. I'm playing with the baffle step compensation, but I'm not sure about including it (another big inductor on the design).
I was using air core inductos everywhere, is there much difference replacing the woofer crossover and baffle step inductors to laminated steel core?
With the baffle step at -8dB:
Whitout it:
I'm not sure if it makes much of a difference really.
And I have a great dip in response at 2000Hz but I don't think I can make much about it.
Anyway, thanks a lot. This is making a bit more sense.
I was referring to baffle step (difffraction) to be included in your bearfield measurement or manufacturer response curve (2pi). That will reduce the woofer spl in the low range by around 4-6 dB.
see vituixcad help "diffraction tool" and "merger tool"!
I'm not sure about (audible) difference between air or steel core inductors. Depends on your listening levels (steel core saturation may induce distorsion). And of course there is a price difference!