• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Appropriate crossover frequency with down firing driver in three-way speakers

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,836
Likes
4,785
Location
Sweden - Слава Україні
Smart, I think, to use the floor as bass support BUT since these bass drivers will be part of a three-way solution, there should be a suitable upper limit for how high up in frequency they should operate. What do you think this limit is?

Different slope on the crossover, up to 24 dB, bass-mid can be weighed into the equation.

Thank you for your comments on this.:)
 

MarnixM

Member
Joined
Sep 28, 2020
Messages
95
Likes
102
Location
Netherlands
I crossover mine at 160Hz with 24dB linkwitz-riley. Active X-over between bas and mid-high panel (= Magnepan LRS+).
 
OP
DanielT

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,836
Likes
4,785
Location
Sweden - Слава Україні
Intuitively, it seems strange to have, for example, male voices voiced out with down firing drivers instead of the drivers directed at me in the listening position. Hence my question about this with the crossover point.:)

Edit:
I don't know...150,200 or 300 Hz?
Note that the down firing bass drivers will then be in two speakers so it is not a question of a subwoofer crossover point.
 

fpitas

Master Contributor
Forum Donor
Joined
Jul 7, 2022
Messages
9,885
Likes
14,213
Location
Northern Virginia, USA
Were it me, and based on nothing but intuition, I'd cross at 100Hz or lower. I bet Hornresp would give you some real quantitative answers.
 
OP
DanielT

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,836
Likes
4,785
Location
Sweden - Слава Україні
Okay. I have four Peerless SLS10 and two SB15NBAC30 that I intended to use as mids and two tweeters, SB26ADC-C000-4. I can build just about anything. For example, four separate boxes with SLS10 placed in each corner of the listening room together with mids and tweeters in own speaker boxes. Time delay between bass and mid/tweeter can be solved then, if needed.Or any other solution.:)

What I'm most interested in is a wide soundstage. That at the expense of pin point and accuracy.
So what can you do with four SLS10?

 
OP
DanielT

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,836
Likes
4,785
Location
Sweden - Слава Україні
I know that lower frequencies are omni but this with direct directed sound compared to floor reflected sound up to around 300 Hz how it works, how it is experienced?

Edit:
Or just smack on a classic. Two speakers. BBMT, that's it.:)
All drivers are directed forward.

Incidentally, I do not consider the SLS10 to be a pure sub driver. Also, if anyone is thinking about these classic drivers, buy at least four. They have low power so it is needed, believe me.
 
Last edited:

tmuikku

Senior Member
Joined
May 27, 2022
Messages
302
Likes
339
It works with wavelength. 100Hz is 3.4m long, or about 11 feet. Direct sound and a reflection have different path length to ear, thus they would interfere constructively when they are roughly in phase at some particular frequency (wavelength) or destructively when they are out of phase. What is in phase then? perhaps if the two have only about 1/4wavelength path length difference, which means 90deg phase difference, which means they would interfere constructively.

At 1/2 wl path length difference phase of the later sound would be 180deg, so completely out of phase and destructive interference. At 1wl path length differemce they'd be in phase, 1.5wl out of phase etc, a comb filter.

So, at 100Hz a singular reflection would sum constructively with direct sound if it arrives to ear less than 3.4m/4 = 85cm after direct sound, 2.5ms of spund travel. At 200Hz this would be roughly 42cm, 400Hz about 21cm, and so on.

Simplified, sound reflects like light from a mirror so you can use simple trigonometry to calculate path lengths. Notice how 100Hz is about size of a room, ecpect reflections from all room boundaries, from all objects for wavelengths that are comparable in size, and so on, so it gets rather complex quick. How it sounds? it sounds like a mess inside a room, just like yoi hear it. Take your set outside and have a listen, notice a difference? :)

In general, sound interacts with physical objects when they are close in size. Low frequencies are omni because wavelength is (much) longer than size of the speaker, thus the speaker is invisible acoustically, and sound radiates to all directions. When wavelength is smaller than the speaker, the box has strong effect acoustically, and there is some shadow behind, edge diffraction and so on, sound interacts with the box.

The 1/4wl rule is kind of easy to remember and approximate with. Reflection is less than that behind direct sound, and its good for all frequencies below. Frequencies above this problems start, comb filter.

Another, your room dominates low frequency sound making it bad, as they are about same size. And your speaker box being about size of midrange sound makes sure the whole midrange is also full of issues. Difference between good and worse system is how you deal with such things, and system is not just the speakers but how they are setup in a room and how they work together so sound at your ear was maximally enjoyable. If you like music, it's likely enjoyable what ever the quality is.

It's all about wavelength. Have fun! :)
 
Last edited:

mcdn

Addicted to Fun and Learning
Forum Donor
Joined
Mar 7, 2020
Messages
579
Likes
806
A similar concept is used with Lyngdorf W210 woofers. The W210 is designed to go in a corner - the ports face forward and each side has a 10" woofer. It's tuned very low so goes quite deep. The principle is to excite many room modes and get a lot of room gain, then use room correction to cut down the peaks to give a deep even response, as well as time align the woofers and main speakers

The Lyngdorf manuals suggest crossing over at 400Hz! I use 180Hz with great success, going higher doesn't seem to gain me much, and we have to give those nice Sonus Faber's something to do! The other benefit is it means you can have the main speakers close to the back wall with reduced SBIR problems.

IMG_0886.jpeg
 
OP
DanielT

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,836
Likes
4,785
Location
Sweden - Слава Україні
It works with wavelength. 100Hz is 3.4m long, or about 11 feet. Direct sound and a reflection have different path length to ear, thus they would interfere constructively when they are roughly in phase at some particular frequency (wavelength) or destructively when they are out of phase. What is in phase then? perhaps if the two have only about 1/4wavelength path length difference, which means 90deg phase difference, which means they would interfere constructively.

At 1/2 wl path length difference phase of the later sound would be 180deg, so completely out of phase and destructive interference. At 1wl path length differemce they'd be in phase, 1.5wl out of phase etc, a comb filter.

So, at 100Hz a singular reflection would sum constructively with direct sound if it arrives to ear less than 3.4m/4 = 85cm after direct sound, 2.5ms of spund travel. At 200Hz this would be roughly 42cm, 400Hz about 21cm, and so on.

Simplified, sound reflects like light from a mirror so you can use simple trigonometry to calculate path lengths. Notice how 100Hz is about size of a room, ecpect reflections from all room boundaries, from all objects for wavelengths that are comparable in size, and so on, so it gets rather complex quick. How it sounds? it sounds like a mess inside a room, just like yoi hear it. Take your set outside and have a listen, notice a difference? :)

In general, sound interacts with physical objects when they are close in size. Low frequencies are omni because wavelength is (much) longer than size of the speaker, thus the speaker is invisible acoustically, and sound radiates to all directions. When wavelength is smaller than the speaker, the box has strong effect acoustically, and there is some shadow behind, edge diffraction and so on, sound interacts with the box.

The 1/4wl rule is kind of easy to remember and approximate with. Reflection is less than that behind direct sound, and its good for all frequencies below. Frequencies above this problems start, comb filter.

Another, your room dominates low frequency sound making it bad, as they are about same size. And your speaker box being about size of midrange sound makes sure the whole midrange is also full of issues. Difference between good and worse system is how you deal with such things, and system is not just the speakers but how they are setup in a room and how they work together so sound at your ear was maximally enjoyable. If you like music, it's likely enjoyable what ever the quality is.

It's all about wavelength. Have fun! :)
That may be so, but Kimmo, the creator of software VituixCAD, says this about wavelengths and c-c distance between speaker elements in relation to lobing effects:

Minimum c-c is 1.0 x wave length and maximum about 1.4 x wave length at XO frequency assuming that design is conventional uni-directional box (not open baffle) with phase matching (acoustical 4th order) slopes. Good and quite flexible initial/design value for c-c is 1.2 x wave length at XO, giving smooth combination of power and early reflections i.e. balanced sound without significant power dip at XO due to bump in DI and dip in vertical early reflections. In other words, this concept aims lobe nulls to directions which are the least significant for power response and vertical early reflections - and listener sitting in sweet spot of course.


About Kimmo:

From Kimmo:
 

Attachments

  • Screenshot_2023-10-06_091333.jpg
    Screenshot_2023-10-06_091333.jpg
    647.8 KB · Views: 31

tmuikku

Senior Member
Joined
May 27, 2022
Messages
302
Likes
339
Hi,
yeah if you start applying the 1/4 rule on things you'd notice it is not possible to have drivers close enough and start looking second best option, which might depend on multiple things and cause rippling effect that might change your plans completely. This is very common pattern in anything regarding loudspeaker design, and I'd bet engineering in general, choosing compromises. How to choose compromises so that at the path you eventually take leads to success instead of failure? First principles, make your goal straight, then arrive some very basic priority list of things or any other tools to help you with decision making to be able to pick compromises that lead you towards the goal.

Like, if your goal is best possible sound quality, then prepare to accept cost could be high, looks might be ugly, size might be bigger than you thought, and so on. If you start compromising your goal by buying some trendy drivers and then wondering how to make a small speaker as it looks kinda cool, everyone has a two way speakers right? :D Sure, you've just trumped your goal, it is very likely you do not get best possible sound quality but something else.

You see, what I wrote about wavelength was a small piece of something I've learned, that wavelength is key to many many questions that come and should help you start solving issues and questions yourself.

You can pick any advice from anyone, problem is there is all kinds of advice to all kinds of situations and much of it is pure marketing purposedly leading you to wrong path. How to pick the correct one for your situation and context?:) just keep your own goals and set of priorities straight and apply any advice you find appropriate in your situation.

Key thing is to have fun, it's a hobby and the time you put in should be fun! So go after what ever feels fun on the moment as that keeps you motivated. Even if it was wrong thing to do to reach your goal, a mistake, it was still very very important learning experience and you've now hopefully learned a ton as you were motivated and interested and likely want to continue toward your goal to reach success. You couldn't have done any better at the time, so, have fun!:)

ps. kimmosto is very experienced person and generously shares what he knows, also doesn't mind what everyone else thinks and is very straight about things. Usually there is context where he writes, and might write something else on another context, like everyone else might do or not. Somebody might give advice that would work for one situation but not for another, and its your responsibility and duty to be able to pick out relevant information for you, to understand the context and if it is relevant to you or not and what to learn from it. If the context is not relevant to you it could be exactly wrong advice. Most imporant thing is try to understand the context and this is often very hard online. It would be better to have a buddy or mentor near by who you could work with in real life and do it together knowing each others contexts and goals without extra work and confusion.
 
Last edited:

ZolaIII

Major Contributor
Joined
Jul 28, 2019
Messages
4,197
Likes
2,477
Well definitely not below 115 Hz, in given case I would go with 250 Hz one.
Why don't you try to find better matching mid woofers? little bigger (about 6.5~7") with about same measured SPL per W and same impedance of course. I probably wouldn't go with alu cone for mids but that's me.
 

tmuikku

Senior Member
Joined
May 27, 2022
Messages
302
Likes
339
Smart, I think, to use the floor as bass support BUT since these bass drivers will be part of a three-way solution, there should be a suitable upper limit for how high up in frequency they should operate. What do you think this limit is?

Different slope on the crossover, up to 24 dB, bass-mid can be weighed into the equation.
Thank you for your comments on this.:)

Quick example, direction of driver is irrelevant below frequencies whose wavelength is perhaps circumference of your construct, or twice, you could check with sim. But it might be relevant in long term usage, if sag develops or something, or if you feel it actually makes a difference in perception although it shouldn't in theory so basically you just have to try it out.

Thinking the stuff bit more, for example floor reflection is not any kind of problem for bass but it is problem for midrange. A floor bounce is typically about 40cm late from direct sound which makes huge dip somewhere between 400-1000Hz or so, depending on height of the driver and listening distance and height. So you should be concerned about where your midrange driver is instead. Don't forget ceiling, if you put driver closer to floor it is now further from ceiling and you might do some listening tests and figure out what is meaningful and what is not. You could notice that depending directivity of your tweeter it is very important the tweeter must be at ear height and integration to the mid is crucial, so there is not too many options to put the midrange driver, below or above the tweeter, or both. Now where the bass driver should be to properly integrate with the mid?

See :) In order to be able to make very very good sounding system you would have to know very accurately what matters, does floor bounce matter and to what? If it does, what can you do about it without causing even more problems?:) Minimizing important audible issues is the key, which means sacrificing non important ones like position of bass speaker in regards to floor. Mistake not, it is very important where the bass speaker is in regards to front wall for example, the wall behind the speakers, which depends again on multiple things and most importantly it's a round trip so surprisingly long path length difference to direct sound.

Crossover slope and frequency can be utilized to try and blend some issues, but more likely you'd need adjust them not because of room related issues but the speaker box related issues, like distance between drivers, modes/sound of the construct, edge diffraction, and so on.

Again, have fun!:)
 
Last edited:

tmuikku

Senior Member
Joined
May 27, 2022
Messages
302
Likes
339
Incidentally, I do not consider the SLS10 to be a pure sub driver. Also, if anyone is thinking about these classic drivers, buy at least four. They have low power so it is needed, believe me.

I'm preaching too much :) so answering this: yeah in general it is a good plan to make a three way speaker and have distributed bass sources. What is your SPL requirements? If you have four bass drivers that are not particularly subwoofers, then I might try to use two of them per the main speaker in bipole configuration in order to reduce box sound. This would happen if you put the drivers opposite sides of a box and mechanically coupled together to reduce vibration. Or any other tricks to reduce "box sound". Then make separate subwoofer system at some point. If you do not play very loud then make two way speaker and distribute the woofers around the room, call it three way. You could also make three way mains, run them fullrange and use the extra two woofers for additional subs.

Although, having big stereo sound and no need for clarity is easy goal to meet so you could use what you have in such way that feels most logical, interesting and fun to to you. Big speakers are fun even though they weren't perfect, so I'd be inclined using the dual woofer per three way box, and then consider separate subwoofer system in addition. Make them look good and so on.
 
Last edited:
OP
DanielT

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,836
Likes
4,785
Location
Sweden - Слава Україні
I'm preaching too much :) so answering this: yeah in general it is a good plan to make a three way speaker and have distributed bass sources. What is your SPL requirements? If you have four bass drivers that are not particularly subwoofers, then I might try to use two of them per the main speaker in bipole configuration in order to reduce box sound. This would happen if you put the drivers opposite sides of a box and mechanically coupled together to reduce vibration. Or any other tricks to reduce "box sound". Then make separate subwoofer system at some point. If you do not play very loud then make two way speaker and distribute the woofers around the room, call it three way. You could also make three way mains, run them fullrange and use the extra two woofers for additional subs.

Although, having big stereo sound and no need for clarity is easy goal to meet so you could use what you have in such way that feels most logical, interesting and fun to to you. Big speakers are fun even though they weren't perfect, so I'd be inclined using the dual woofer per three way box, and then consider separate subwoofer system in addition. Make them look good and so on.
My SPL requirements are pretty much minimal because I live in an apartment. Nowadays, I rarely play louder than normal listening volume. A note regarding down firering bass. That might then seem stupid considering the neighbor below me, but she is a student who, luckily, likes to party and play music at high volume, so she has to count on a little bass umpf from me sometimes. It doesn't bother her one bit. :)
Damn, sometimes you're lucky with which neighbors you have he he.;)

Yes, there is a thought. I heard these at a fair and I thought they sounded really good. I wanted some of these. No more advanced thoughts than that really, mostly based on perceived impressions.

I have spoken to Samuel about his construction and seen his design in detail. He's done a ton of measurements on them and they look really good. FR that is. This is what they look like:
IMG_4299 (2) (1).jpg


Samuels drivers in them:
Tweeter:SB26ADC-C000-4
Mid:SB SB15NBAC30
Bass: Satori WO24P-4
He uses a 24dB filter, which I was also thinking of doing. Via miniDSP and an active crossover.
Nice drivers and good solutions that I was inspired by and thought of copying. Why then the Peerless SLS10 and not the Satori WO24P-4 that Samuel has in his? Well Satori WO24P-4 is certainly great but it costs a lot. I got the SLS10 for a good price and I think it will work well for me.
SLS10 seems to work well up to 400-500 Hz according to tests. SLS10, 88dB sensitivity into 8 Ohms. If I then connect two in 4 Ohm, I increase the sensitivity by maybe 6 dB, which is good. Then they don't need so much power. Also with four SLS10's placed appropriately it should reduce the problems, or challenges of :

I have a subwoofer with LP-HP filter that I can set to 50Hz. SLS10 can then work from 50Hz up to..X Hz..

I also got inspiration from McFly. The same mid and tweeter as Samuels DIY:

IMG_0365.JPG


Woofer: SB15NBAC30
Tweeter: SB26ADC-C000-4


Well definitely not below 115 Hz, in given case I would go with 250 Hz one.
Why don't you try to find better matching mid woofers? little bigger (about 6.5~7") with about same measured SPL per W and same impedance of course. I probably wouldn't go with alu cone for mids but that's me.
The SB Acoustics SB15NBAC30 has an x-max of 5 mm and seems to go down nicely to around 200Hz before distortion increases below that. Incidentally, the SB15NBAC30 generally has really low distortion. The advantage of a 5 inch driver is that nice off axes a bit up in frequency. Hopefully, it makes it easier to integrate with a tweeter around for example 2.5 kHz than with the larger size of the mid driver. Which can be solved with WG if a larger size mid driver is used.
You know, as usual, compromises.:)

SB15NBAC30:
FR:
sb15nbac30-4_offaxis_0 (1).png

Distortion plot:
sb15nbac30-4_0.3m_2v_hpf2-50_hd.png


Edit:
If I have time, I will also build a smaller two-way speaker with these drivers that I happen to have. Maybe with a crossover point around 4-5 kHz. Together with SLS10, which can then work up to 400-500 Hz. What do I know, I might like that sound better than with said SB drivers.;):)

Faital Pro 4FE32
https://faitalpro.com/en/products/LF_Loudspeakers/product_details/index.php?id=401005100

 
Last edited:

ppataki

Major Contributor
Joined
Aug 7, 2019
Messages
1,253
Likes
1,421
Location
Budapest
Intuitively, it seems strange to have, for example, male voices voiced out with down firing drivers instead of the drivers directed at me in the listening position. Hence my question about this with the crossover point.:)
I agree, I even find it strange to have male voice coming out of a sub facing front....
In my experience if you use 24dB/octave slopes at around 80-100Hz male voice shall no longer sound intelligible
If you use 48dB/octave that will be 100% fine or potentially even an overkill
Best is to actually play around with these settings and listen how it sounds (assuming that you have an all digital xover that you can change in real-time)
 

Salt

Addicted to Fun and Learning
Forum Donor
Joined
Aug 7, 2023
Messages
618
Likes
342
Location
DE
With four sub drivers one might consider impulse correcting opposite site constellation (left-right) instead of downfire.
Does not change the effect to neighbours, but probably may reduce distorsion.
Lowpass up to 120 Hz should do right with that.
 

Wolf

Addicted to Fun and Learning
Joined
Nov 21, 2020
Messages
573
Likes
611
Location
Indiana
If you have woofers to the side, bottom, or top, and they are not on the same panel as the mid(bass), you will find the lower you xover, the more they will be easy to blend. The ROT for side mounted woofers tends to be below 200Hz, and lower than 150Hz is better still. It has to do with the localization of the side woofers, and 150 is the same idea with separate woofers. However, going one octave below there keeps the second harmonic at bay and under the threshold. This makes 75Hz more optimal if able.
 

kemmler3D

Major Contributor
Forum Donor
Joined
Aug 25, 2022
Messages
3,358
Likes
6,887
Location
San Francisco
I would intuitively say no higher than 200hz for a down-firing woofer, and the lower the better if you can manage it.

That said, the LS60 crosses the side-firing woofers at around 350 IIRC and they sound great, so...?
 

Duke

Major Contributor
Audio Company
Forum Donor
Joined
Apr 22, 2016
Messages
1,585
Likes
3,911
Location
Princeton, Texas
Smart, I think, to use the floor as bass support BUT since these bass drivers will be part of a three-way solution, there should be a suitable upper limit for how high up in frequency they should operate. What do you think this limit is?

Calculate the floor-bounce-dip frequency for your midrange driver. Cross over to the down-firing woofer a little bit ABOVE this frequency, so that you effectively remove the floor bounce dip from your room interaction.
 
OP
DanielT

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,836
Likes
4,785
Location
Sweden - Слава Україні
Good tips, advice and suggestions from those of you who write in the thread. Thanks for them.:)

Now I have to think and I will probably have some questions so I will come back later today. BUT I remember one thing. There are speakers where bass, bass/mid drivers and even tweeters are not aimed directly at the listener. I'm thinking of omni speakers. Some people seem to like that sound. Others think the sound will be too "smeared" and not exactly pin point so to say (which is the purpose of omnis?). Some examples:
retropia++3-21-2014-23-56-51.jpg

11 (1).jpg

Here an example of a two-way construction with the bass driver directed upwards. I've had a couple of those (not the ones in the picture below).
Unfortunately, I can't say it was a positive experience BUT I don't know if it's due to the construction itself. The old Philips 9710 driver in them had lost all its former glory for example:
Sonab-OA5-Typ-2-48635.jpeg


 
Last edited:
Top Bottom