3-way how would you do it?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi everyone. Thinking of a 3 way, if one day you felt like doing it, how would you do it? What would be the things to think about before choosing the 3 drivers that make it up and start designing? below I write what comes to mind: the first of all considerations is that of the best quality / price ratio without spending too much (thinking of a possible replacement in the future) and on this I think many of you agree.

Second consideration: how to split cut frequencies? to divide the frequencies according to the chosen drivers or to choose the drivers based on the chosen frequencies? But the real question is this, which technology to choose for each single frequency band? As for the low frequencies I believe that the choice is precluded to the magnetodynamic drivers but for the rest of the medium and high frequencies, which ones to choose?.

I would direct the reasoning to do this: which technology best expresses the medium frequencies? And which one for the high ones? Another important thing: the 3 drivers must be well blended, there must be a fusion between them and a synergy. They have to work well together without one of them prevailing over the others. Crossower: as simple as possible, slope? Active or passive ? you say. Requirements for placement in the environment and format of the boxes: the dimensions have their importance and since many people have small to medium rectangular or square rooms including mine, I think that the best choice is a tower leaning against the wall.

Sensitivity and pairing with the amplifier: on this I let you say. as for the choice of technologies, personally I would do it like this: magnetodynamic loaded to TL for the low ones, CD to compression for the middle ones, for the high ones I am still uncertain, maybe a tape? For now, it is not specific which drivers I would choose because I do not have clear ideas, I let you say which ones are better. I am aware of not being an expert so if I have written any nonsense please correct me as well.
 
Build a passive 2-way, but with plans for later bass addition, via active xo.

Choose your tweeter first, and spend the most on it. It dictates 80% of 'the sound' because the top end is where ALL the harmonics are. There are indeed very interesting 'planar' or AMT-type tweeters around today.
 
  • Like
Reactions: 1 user
Find a driver that covers at least 300Hz to 3kz without issues. Measure it. Choose lp filter so you stay clear of cone breakup. Find a tweeter that can be crossed where it needs to and that you like. As has been noted the tweeter WILL influence perception of mids and bass.
Find a woofer that can be used as high as necessary.
At least to start out go active.
 
  • Like
Reactions: 1 user
Thinking of a 3 way, if one day you felt like doing it, how would you do it?

3 way how would you do it?

Is this a SEX thread???
Are you a SEX pervert, or just an Audio Pervert?
Audio Pervert! ........ well.... ok then....

Give us some solid info and goals....room + seating sketch, favorite music, budget, input sources -analog?, amplifiers -analog?, construction skills.... "The woodshop capabilities often determines the speaker design."

1) you can copy off of the Smart Kids ($32K Avalon beveled style cabinet(1"+6"+12"), $22K Sony AR1 (dual 8" woofers), $100k Magico style large radius cabinet for under $500/speaker + cabinet)
2) you can master DSP to "follow the $100k super speakers" ( for $2000/speaker + cabinet)
e.g....
You are Italian, but also a citizen of the World.
Buy some B&C and Faital, but seek value!

20,000 - 500Hz B&C DCX464 coaxial compression driver (Italy)
---SEOS24 waveguide (Poland)
500 - 40Hz Faital 15FX560 (Italy)
20 - 40Hz Dual Peavey 18" woofers, side-side mounted(Tiawan)
 

Attachments

  • Avalon with Sealed 12in.jpg
    Avalon with Sealed 12in.jpg
    216.5 KB · Views: 489
  • Sony AR1 dual_8.jpg
    Sony AR1 dual_8.jpg
    174.9 KB · Views: 486
  • Magico TMWWW Rounds.jpg
    Magico TMWWW Rounds.jpg
    210.9 KB · Views: 465
  • Future BIG.jpg
    Future BIG.jpg
    185.8 KB · Views: 469
Old school

I like the old-school approach of having crossover frequencies at approximately 500-700Hz & 5k-7kHz
(instead of the 200-300Hz & 2k-3kHz that are more common nowadays).

In other words:
Mid-woofer + Mid-range + (super)Tweeter
(instead of Woofer + mid-woofer + tweeter).

Marco
 
12”-18” crossed at 130hz to a 4” crossed at 2.1khz to a dome tweeter...can’t beat it, naked midrange. Active 48/dbLw, tri amped powered by dsp amp for voicing and room correction.

I’m trying to figure out how to use a compression driver with same philosophy. A 12”-15” crossed at 500-700hz in a two way is the best I can figure. There’s like 3 compression drivers up for the task of either running all the way to 20khz or maybe crossing to a dome at 10khz for a 3 way. There’s also one potential AMT to cover 700hz-20khz, and at this point it’s a maybe.

I try to design for near and far field. With nearfield only one driver can be pointed directly at ear level. With the first 3 way, the mid is the center listening axis, the ka on the woofer is something around .55 so off axis is great and the tweeter being a dome is the best you can do dispersion wise
With the horn system, I can’t go low enough to provide the ka I want, barely getting to 2ish on the woofer, a sacrifice, so it would only make sense to offset that sacrifice to run a two way but like I said, it’s a great task for the top driver and beaming in top register is inevitable it will be at the center listening axis (the horn) which is slightly better than having the beaming coming from the dome tweeter, where as a 1 inch would seemingly beam above 13550 though I do not understand the dispersion characteristics of a dome completely. If you ran a AMT tweeter the center listening axis would have to be the tweeter because of how directional it is, and then you’d try to design around that, crossing *** low as possible to create low ka values for the mid and woofer.

Then there’s the fast speaker if you can find the perfect coaxial driver for the top and cross it as low as possible to a woofer.

I also could be crazy ��
 
Last edited:
The best 3-way is a 4-way. I wish more people would realize this. It's all but ignored at diy websites. I know it's more money, but we're already saving a bundle on store bought.
We used to have a few real guru's here. The law's of physics where often discussed. A 6" midrange can't be paired with a 1" tweeter, but everyone does it. Let's try to be different.

-tweeter
-2" - 4" midrange
-6" - 10" midwoofer
-15"+ woofer

or

-tweeter
-2" - 4" midrange
-multiple 6" - 10" midwoofer (ported, transmission line etc.)

Someones going to respond; it's more difficult to design a 4-way than a 2-way. I would counter; if you chose your drivers carefully, no out of band peaks/dips/breakups will need to be heavily addressed, making the crossover easier? (that's a stretch, it's still hard).
 
...Thinking of a 3 way, if one day you felt like doing it, how would you do it?
What would be the things to think about before choosing the 3 drivers that make it up and start designing? below I write what comes to mind:

  • ...considerations is that of the best quality / price ratio without spending too much (thinking of a possible replacement in the future)...
  • ...how to...to divide the frequencies according to the chosen drivers or to choose the drivers based on the chosen frequencies?...which technology to choose for each single frequency band? As for the low frequencies I believe that the choice is precluded to the magnetodynamic drivers
  • ...for the rest of the medium and high frequencies, which ones to choose? ...to do this: which technology best expresses the medium frequencies? And which one for the high ones?
  • ...the 3 drivers must be well blended, there must be a fusion between them and a synergy. They have to work well together without one of them prevailing over the others.
  • Crossower: as simple as possible, slope? Active or passive ? you say.
  • Requirements for placement in the environment and format of the boxes: the dimensions have their importance and since many people have small to medium rectangular or square rooms including mine, I think that the best choice is a tower leaning against the wall.
  • Sensitivity and pairing with the amplifier: on this I let you say.
  • ...as for the choice of technologies, personally I would do it like this: magnetodynamic loaded to TL for the low ones, CD to compression for the middle ones, for the high ones I am still uncertain, maybe a tape? For now, it is not specific which drivers I would choose because I do not have clear ideas, I let you say which ones are better.

One observation that you might have made already by reading the responses is that there are many entry points to the cycle of designing and implementing any loudspeaker.

If you start with a set of performance requirements in some sort of hierarchical order, then you add your particular design constraints to use some sort of technology while using those hierarchical performance requirements, you will increase the chances of a successful outcome. I strongly recommend writing down your requirements, and then assigning precedence to the importance of each requirement relative to the others. This is where most loudspeaker designers (DIY or commercial) tend to make the most mistakes: they fail to realize that they haven't properly balanced the requirements and they end up with something that doesn't survive the test of time.

I think that when you look at successful loudspeaker designs and the kind of person that buys each of them and their specific proclivities for the sound they seek and total ownership experience, i.e., the degree of "exclusivity" and image attached with whatever they build--sort of like the Pygmalion effect--which seems to be a big part of the incoming requirements, you will begin to see why the answers that you get from others in this thread may not be as useful as you had intended. When you write down all that stuff mentioned just above and arrange it in a hierarchical list (irrespective of the specific technologies or arbitrary design constraints chosen), I think that you'll go a long way toward answering your own questions.

Don't forget to include your room size/acoustics in the performance requirements of your 3-way loudspeaker, and any specific requirements to the type of sound reproduced based on specific music genres that you might intend to listen to more than others. In my experience, when you do add this operational view to your design problem, the design usually changes.
_____________________________________________
For me, I would not consider direct radiating acoustic drivers. The reasons for this come from my overall requirements for full range directivity control and the space that I intend to put the resulting loudspeakers into. There are psychoacoustics reasons (i.e., how the human hearing system works) that are the source requirements for this design decision.

Chris
 
Last edited:
Disabled Account
Joined 2017
I like the midrange to be as wide as possible; usually this means staying in the 3 inch to 5 inch size. Ideally; what I am doing most of these days is a fullrange or near fullrange with super-tweeter and woofer (adding a sub-woofer completes the package). So, for me, I try to choose a fullrange that can do 100 Hz to 10 KHz ideally, usable from say 200 Hz to maybe 8 KHz; 300 Hz to 6 KHz is good and I rarely ever choose a mid or full range that can't do at least 400 Hz to 4 Khz. Of course; there are many different views and "schools of thought" on how to best approach a new design. two 6 in or 7 in or 8 in woofers in parallel I think overall have a better quality sound than say one 10 in or one 12 in woofer just because of a quicker transient response if nothing else. Do you have a price range in mind yet? What room size? What types of music? Many different questions. I'll stop here; read up on the many successful designs out there already. One more thing; I have done designs based on starting with a known woofer I wanted very much to try and built around that. I have also done the same with a known mid or fullrange. I have also done the same with a known tweeter. This is a never ending hobby for me; once bitten by the DIY bug; it's like a disease or something that never seems to ever completely go away. I often think about what I might try next...have fun!
 
I like the old-school approach of having crossover frequencies at approximately 500-700Hz & 5k-7kHz
(instead of the 200-300Hz & 2k-3kHz that are more common nowadays).


+1, this keeps crossovers and the phase muck they create outside the range of frequencies where we are most sensitive to it. Of course the application the system is designed for will impact this, for home listening at a single position off axis response can be sacrificed with widely spaced crossovers but if broad listening coverage is required then all drivers need to be crossed well below where beaming begins.
 
Disabled Account
Joined 2017
+1, this keeps crossovers and the phase muck they create outside the range of frequencies where we are most sensitive to it. Of course the application the system is designed for will impact this, for home listening at a single position off axis response can be sacrificed with widely spaced crossovers but if broad listening coverage is required then all drivers need to be crossed well below where beaming begins.

Agree; I think that beaming is the least important parameter. Of course, most of my critical listening is just me or maybe one other person. The Fletcher-Munson curve shows us our most sensitive range and I also try to avoid any crossovers there. Also; if one were to look where the top fundamental frequencies (highest notes) of some musical instruments are, you would see the piano for instance has its range past 4100 Hz! This is why I rarely cross-over a 3-way mid to tweeter below 4 KHz; 5 or 6 KHz are preferred. The REAL magic to me is getting the tweeter and mid to blend "seamlessly" so that the very important overtones, harmonics, etc. of these higher notes all arrive at my ears with the original harmonic structure intact. People like me that are (or were at one time) musicians get really picky about these things; this is what really makes a great loudspeaker stand out from a good or mediocre design! Call it phase purity if you will. I will add that I try to use 1st order X/O's whenever I can get away with it for the exact same reasons I just gave above.
 
how to split cut frequencies?

This and many other questions depend on what you want to achieve. Assuming you're designing a speaker with a passive x-over, the first thing to ask yourself is how big the speaker is allowed to be, how efficient it needs to be, and how deep the bass needs to go.

These points will guide the choice of the woofer. With a larger woofer, the woofer/mid x-over will end up at a lower frequency than with a smaller woofer. For example, a 15" woofer will work well up to maybe 200-400 Hz, whereas an 8" may be okay up to 800 Hz or so.

Once the woofer and it's x-over cut off is set, choose a midrange that fits with this x-over. A tiny 2" midrange might struggle to go deep and loud, so choosing a large woofer will also require a relatively larg midrange (or midwoofer).

When choosing the tweeter, I'd try to keep an eye on matching the dispersion of the midrange.
 
Talking old school if you are placing them close to a backwall I'd go for a wide and shallow cab rather than deep and narrow. No need to bother with baffle step 'correction'.


Also start off active and stay that way.
This (cab dimension choice is what I chose for my surrounds).

A 15 or 18 for the MB, an 8 for the MR and a TPL150H for the HF. Where the actual xovers are would depend on the specific drivers, but they would be active and steep.
 
Disabled Account
Joined 2017
How about these drivers as a "starter" set? Inexpensive but all around good reviews. I'm just looking at different sites thinking about what combinations of drivers I might like to try in the future. (at various price points)

Your Madisound Cart

In case the link won't open; the drivers are:

Fostex FF85WK

Seas 22TAF/G

SB20PFC30-8

All low cost but very good performance.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.