I'm in the final stages of building my first pair of 2 way speakers and already thinking what next. Ive built plenty of subwoofers. When I started this 2 way project it was originally going to be a 3 way but I must have heard at least a half a dozen people tell me not to go 3 way since it is a way more complex build. I reluctantly decided to go 2 way and Very glad I did. I learned a lot on this 2 way build but what I didnt learn is why are 3 way builds so much harder? Isnt it just an extra driver and crossover. Is it that much more difficult to match up 3 drivers? Anybody care to educate me on the subject? I'd appreciate anybody willing to take a few minutes to explain it to me. This forum has been a huge help for me already so thank you for that
Hi,
If you use DSP you'll add few more cables and use few more mouse clicks🙂
Kidding aside it increases complexity, you now have to measure one more driver for example and getting good measurement data with high ebough resolution on low mids could be more work. Or if you have a passive xo filter, tweaking it needs more parts. But, these are kind of trivial tasks in a way not black magic, just something that needs to be done, more work and more possibility for humanly errors.
On the other hand, if you dig deeper into how speakers work and start to eliminate issues you'll find out why a 3-way speaker in the first place, and that actually makes some problems go away. It's got more separation of concerns and in this way ideally better, possible to weed out more issues because each driver has narrower bandwidth.
Perhaps the general notion is due to more work, it's easier to get ~same performance from 2-way speaker than from 3-way speaker. But, if you need to go beyond what a 2-way can offer that's the only way to do it, add more ways.
Better yet, don't even think about amount of ways, thats kinda irrelevant in itself, but think about your goal for the project and solve issues to reach that goal. Amount of ways is just one variable in the design space, and you'll either achieve your goal or not.
Example, you'd like small speakers for nice looks but would like nice bass because nice bass is very important audio quality. Study it and you'll find out some bass issues are due to room modes which are affected by physical location of sound sources and the observer within the room. Having separate bass box(es) and mid-top you can optimize physical location of bass with the room to make good bass, and mid-top for good stereo image. This "problem" is in all small rooms, some just don't mind about it, but if you do you'll need the third way. Or make it two way system with fullrange driver and a sub, if the fullrange driver meets design goals for the mid-top.
Have fun!
If you use DSP you'll add few more cables and use few more mouse clicks🙂
Kidding aside it increases complexity, you now have to measure one more driver for example and getting good measurement data with high ebough resolution on low mids could be more work. Or if you have a passive xo filter, tweaking it needs more parts. But, these are kind of trivial tasks in a way not black magic, just something that needs to be done, more work and more possibility for humanly errors.
On the other hand, if you dig deeper into how speakers work and start to eliminate issues you'll find out why a 3-way speaker in the first place, and that actually makes some problems go away. It's got more separation of concerns and in this way ideally better, possible to weed out more issues because each driver has narrower bandwidth.
Perhaps the general notion is due to more work, it's easier to get ~same performance from 2-way speaker than from 3-way speaker. But, if you need to go beyond what a 2-way can offer that's the only way to do it, add more ways.
Better yet, don't even think about amount of ways, thats kinda irrelevant in itself, but think about your goal for the project and solve issues to reach that goal. Amount of ways is just one variable in the design space, and you'll either achieve your goal or not.
Example, you'd like small speakers for nice looks but would like nice bass because nice bass is very important audio quality. Study it and you'll find out some bass issues are due to room modes which are affected by physical location of sound sources and the observer within the room. Having separate bass box(es) and mid-top you can optimize physical location of bass with the room to make good bass, and mid-top for good stereo image. This "problem" is in all small rooms, some just don't mind about it, but if you do you'll need the third way. Or make it two way system with fullrange driver and a sub, if the fullrange driver meets design goals for the mid-top.
Have fun!
Last edited:
A 3-way is just more complexity. That is all. One does not necessarily have to do a 2-way before tackling a 3-way, but it is a good idea to go this way if just to ease the learning path. A critical part of the learning journey is to get first-hand experience of the limits of each approach. You never really know until you experience it for yourself. (For instance, I am sold on 3-ways but, at this stage, my ideal is a 3-way plus subwoofer configuration).
Edit: @tmuikku said it so much better than I, and with all the detail! 🙂
Edit: @tmuikku said it so much better than I, and with all the detail! 🙂
it's total BS - there is nothing inherently more complex about 3-ways than 2-ways. absolute nonsense.
my first speaker was a 3-way and i was in high school and internet was barely even a thing back then and i had no idea what i was doing and it still came out great and everybody was jealous of it.
i wouldn't even bother with anything less than 3-way + subwoofer personally. it would be flushing money down the drain.
i suggest you skip 3-way and go straight to 4-way + sub.
i'll help you.
what's the worst thing that can happen ? it's not like skydiving or riding a motorcycle. you risk literally nothing. it is an absolutely absurd fear.
EDIT: when i say "i'll help you" i mean only with the part that makes 3 or 4 way different from 2-way, which is mostly a philosophical issue of how to select drivers that compliment each other. i won't help you with anything like measurements etc. as you would need that also for 2 ways.
my first speaker was a 3-way and i was in high school and internet was barely even a thing back then and i had no idea what i was doing and it still came out great and everybody was jealous of it.
i wouldn't even bother with anything less than 3-way + subwoofer personally. it would be flushing money down the drain.
i suggest you skip 3-way and go straight to 4-way + sub.
i'll help you.
what's the worst thing that can happen ? it's not like skydiving or riding a motorcycle. you risk literally nothing. it is an absolutely absurd fear.
EDIT: when i say "i'll help you" i mean only with the part that makes 3 or 4 way different from 2-way, which is mostly a philosophical issue of how to select drivers that compliment each other. i won't help you with anything like measurements etc. as you would need that also for 2 ways.
Last edited:
Some people are more and some less talented. I advise to do whatever your gut is telling you and put that to test.
Build whatever X-way speakers you think you need. Being curious about everything, will cost you not just money,
but most importantly life energy.
Absolutely, what ever feels worthy of the time spent. Personally I consider this very much fun tinkering around, and it also rewards with very good sound in the end, but one must have motivation to plow through it, and it might take few years.
What should be the joint together in a 3-way system?
1 inch tweeter + 4 inch mid + 12 inch sub
Is that a good way?
1 inch tweeter + 4 inch mid + 12 inch sub
Is that a good way?
perhaps, or two 21" bass with 15" mid and 1.4" compression driver. Depends on many things what you expect from a system at an particular application you have in mind. Could be 8" coax with sub, or anything really, depends on what you tailor it for. Directivity and SPL capability, possibility to reduce issues by making them outside passband and so on.
Last edited:
if the 12" is a sub then it's not a 3-way but Sub+Satellite System.
a good XO frequency from 12" to 4" is about 300 hz while a good XO frequency to sub is about 80 hz.
JBL had a 3-way in their LSR ( Linear Spatial Reference ) system that had 1" 5" and 12" with 2.3 khz and 250 hz crossover points. this should be a good boring run of the mill 3-way. just don't call it a sub - call it a woofer.
personally i would rather do a 5-way like this:
bullet supertweeter: above 7 khz
dome midrange: 1 khz - 7 khz
8" cone midrange: 250 hz - 1 khz
2 X 15" woofer: 60 hz - 250 hz
8 X 21" subwoofers: below 60 hz
LOL
but seriously ...
this is the type of system people run on AVS ... where it's all about bass and the more subs the betterer !
the benefit of a system like this is you don't actually go deaf from bass ( surprisingly ) as much as you do from other frequencies so those guys aren't as dumb as they seem ... or at least they are accidentally right.
the problem is if you're going to spend that much money on subs you may as well finish the rest of the system as a ribbon array.
In the end all this is semantics, details, which matter only for the engineer/builder to be able to reach some target performance within constrains like budget and aesthetics. Personally I care only about sound and all the rest I'd skip if I could. I'd like the sound be exceptional and speakers invisible, cost low, so it's quite a task and involves all the details of course and cost becomes very high, aesthetics poor and so on, but as long as sound quality gets better and better it's fine 😀 So, it's just fun process, everyone should have fun with the hobby, what ever it is to them!
Without actual expertise I would say, regarding a passive 3-way:
- 3 instead of 2 drivers to be positioned on the baffle and be accounted for diffraction, directivity ....
- twice the crossover points to get right or wrong (SPL, response, slope, directivity, phase, lobing, impedance ....)
- eventual (probably very low) overlapping of three driver outputs
- twice the crossover parts (still without impedance linearization).
- usually much more expensive crossover parts (bigger inductors and caps for Bass-Mid-x-over)
- most bass and mid drivers need impedance linearization of resonance hump(s). usually expensive, big parts necessary
- with 1.5-fold driver count and probably 3-fold crossover parts count it gets exponentially more difficult to make decisions and tweak the system after measuring and listening.
- and finally with all the work and money invested the expectations also increase. that might be the biggest difficulty.
with an active (DSP?) system many of those difficulties are gone or diminish.
Last edited:
Do you measure raw drivers? Do you design your own crossovers? Or follow other designs?
In my opinion the most important part is to know what kind of sound you are looking for. After that, you choose the type and size of the units which provide the requirements and the big trick is to merge them together in one sound. That's more complex then only a cross over but involves radiation pattern, form and size of the speaker, room reflections ect. ect. to determent the end result, and lets not forget your ears and brain ability for interpretation in coming sound. 😉
Dealing with baffle step in a three way is more complex when you don't have an anechoic chamber to measure in. Especially if you cross the low woofers to the mid range anywhere in the baffle step area (which is likely). So yeah, it's more complex, but it's also more fun and probably more rewarding.
Why ?
Do you see a 4 way like a 3+sub ? Or more like bass + lower mid + upper mid + tweet (and so will also 'need' a sub ? )
Why? Isn't it obvious?
Every speaker is a set of compromises.
Single fullrange is terrible compromise. Select big one with bass, you get no highs, beaming and breakups. Select small one with decent mids and highs, but you get no lows, or with terrible distortion. Such is physical reality. No need to get offended. Moderators!
2way is still terrible compromise, while some may be satisfactory. Select big woofer, and you get poor mids with early breakups. And you need very powerfull tweeter going very low. You get mismatch in directivity, just look at heat map directivity plots, see all those butterflies, instead of smooth transitions. You can do waw (woofer assisted wideband) but you get all the worts of fullrange.
Even 2way bookshelf is most common speaker in the world, its far from perfect.
In 3way you have compromise to make how big the mid you select. Select big mid, and you have to cross to tweeter low, making things hard for tweeter. Often big dome tweeter with terrible beaming. Did you notice how dome tweeters, even those in high end speakers beam above 10kHz?
Ideal mid for me is 4-5" with minimum breakups, which can be crossed quite high. But 4"mid will not give you midbass slam.
So between proper, say 15", woofer you better intoduce lower midbass. Less compromise than 3way.
So select big woofer, smaller midbass, even smaller mid, and perhaps nice ribbon on top. Smaller the range each speaker covers, more linear it behaves.
When i work on crossover for multiway speakers, i only deal with one at a time. First i measure mid and tweeter separately. Select crossover point to eliminate breakups of mid. Select steepness utilizing natural rollofs. Design as simple as possible for flat response and phase. Ones done with this ccrossover point, it becomes fixed. Then i work on lower midbass and top (mid plus tweeter just finished). So i am only concentrating on one crossover point, essentialy optimizing as 2way. And so on. You can work out 5way crossover without much issues.
Everything is compromise, but if you are happy with one tiny fullrange, or small 2way bookshelf, its ok.
I am only building 4way and up.
Cheers!
Every speaker is a set of compromises.
Single fullrange is terrible compromise. Select big one with bass, you get no highs, beaming and breakups. Select small one with decent mids and highs, but you get no lows, or with terrible distortion. Such is physical reality. No need to get offended. Moderators!
2way is still terrible compromise, while some may be satisfactory. Select big woofer, and you get poor mids with early breakups. And you need very powerfull tweeter going very low. You get mismatch in directivity, just look at heat map directivity plots, see all those butterflies, instead of smooth transitions. You can do waw (woofer assisted wideband) but you get all the worts of fullrange.
Even 2way bookshelf is most common speaker in the world, its far from perfect.
In 3way you have compromise to make how big the mid you select. Select big mid, and you have to cross to tweeter low, making things hard for tweeter. Often big dome tweeter with terrible beaming. Did you notice how dome tweeters, even those in high end speakers beam above 10kHz?
Ideal mid for me is 4-5" with minimum breakups, which can be crossed quite high. But 4"mid will not give you midbass slam.
So between proper, say 15", woofer you better intoduce lower midbass. Less compromise than 3way.
So select big woofer, smaller midbass, even smaller mid, and perhaps nice ribbon on top. Smaller the range each speaker covers, more linear it behaves.
When i work on crossover for multiway speakers, i only deal with one at a time. First i measure mid and tweeter separately. Select crossover point to eliminate breakups of mid. Select steepness utilizing natural rollofs. Design as simple as possible for flat response and phase. Ones done with this ccrossover point, it becomes fixed. Then i work on lower midbass and top (mid plus tweeter just finished). So i am only concentrating on one crossover point, essentialy optimizing as 2way. And so on. You can work out 5way crossover without much issues.
Everything is compromise, but if you are happy with one tiny fullrange, or small 2way bookshelf, its ok.
I am only building 4way and up.
Cheers!
that is like saying that tying shoelaces on both feet is more complex than on one foot.
no it's the same level of complexity. you either can tie shoelaces or you can't.
complexity implies a different level of interaction or abstraction.
simply having more of the same thing doesn't make it "complex"
a 10,000 page library written by a 2 year old that simply repeats "baaa waa waa" for 10,000 pages isn't complex because of the number of pages.
to give you another example a V12 engine is no more complex than a V6 - in fact it is SIMPLER. because V6 needs balancer shafts and V12 does not.
i see the world complex thrown around a lot in car reviews and it never is in context of anything actually complex. i mean maybe relative to a horse buggy.
you know what is complex ? the relationship between US and Israel is complex. literally nobody could explain what it is in a way that is both accurate and concise. NOT EVEN ME. THAT is complex.
REPETITION IS NOT COMPLEXITY.
Last edited: