18W, 12M and R3004 active 3-Way

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
#rocky - I hear you. I hope my experiment with my old cabinets can give an answer to this question. I should be able to go from the right size to my undersized target. Then i will know if i should either stick with one woofer and make the speaker smaller and sealed or keep the ported design or if i can use two woofers in the current design as a sealed box.
 
Hi Doppler,

First benefit is dividing one large volume into 2 equal halves will increase cabinet rigidity and decrease panel resonance.

The more subtle benefit is spreading of driver resonance.

2 drivers in one box will result in one F3 and one Fb. So whatever peaks / troughs in the response are fixed.
As no 2 drivers are 100% identical (most specs are plus minus 5% to 10%) when you place the 2 drivers in 2 separate boxes this results in two different Fb and F3 so the peaks and troughs will be at different points.... They fill in the gaps but both stay within the same response limitations.

Wilson Audio use "bass resonance balancing" in their $6 figure speakers.... They use a 12 inch and a 15 inch driver in separate chambers to cover the bass.
 
Sealed box with DSP

with undersized i mean a driver that is fitted in a smaller box than the TS parametres suggest. Please note I am open to knowledge about this - I Just dont get how this method Will work better than the traditionel on.

My audio journey started about 35 years ago with Hi Fi Answers "Compact Monitors", 3 cubic feet of classic sealed box 3 way monitor.

After a decade of "Loudspeaker Design Cookbook" I tried to "upgrade"
to a ported Wilmslow Audio kit, "The Vortex" , a 10 inch Volt bass mid / passive crossover Morel dome tweeter.... Disaster!

Ever since then I have been researching and listening, my conclusion is sealed box is the only way to accurately reproduce sound.

A few years ago everything clicked when I discovered the attached....
I have read the article over a dozen times and researched as much as I can on the HAS and how our ear brain detects, decodes and interprets sound.

Using DSP (FIR) to Eq the frequency response, implement crossovers and time align drivers / subwoofers enables one to build compact non resonant sealed enclosures for the right drivers...


By this I mean, low Qts, high Bl / low Mms, fabric surround, large voice coiled Pro drivers.

The end result will be a smaller, better sounding, easy to install (achieve design performance in a wide variety of rooms) and more reliable system.
 

Attachments

  • John Watkinson July 2014-Conventional loudspeaker failings.pdf
    584.9 KB · Views: 88
Interesting reading once again alex.
So in theory there is no lower limit to a speaker cabinets volume as long as:
-The drivers motor system is strong enough to defeat the air spring.
-The driver cone is stiff/strong enough to not break up.
-And all the other stuff I am to dumb to know about.
I know this is only theory and thus the lowest possible volume is decided by the drivers abillity to do the above.
 
Hydrogen Alex: Thank you for elaborating. Very interesting read. I think we are actually on the same page :)

Two questions arises from my mind:
1) I dont know if it is a myth but - Does some pro audio drivers in your experience have a tendency to "die/lack" dynamic sound when spl is low?

2) Do you have good examples of drivers in in 12" category that also wont cost a small fortune.

I have thought about using Pro Audio drivers, but have not tried it yet. My hesitation is that they will not be able to play down to 30 Hz and also the dynamic issue. I use miniDSP with FIR and Jriver and have no plans (like ever) to go back to passive filters.

That said - It is mostly a matter of proper implementation. It takes loads of experience to design a good sounding speaker. Right know I really like the sound of Samuel Harsch filters. Quite easy to get close to a textbook like stepresponse with minimal phase filters.

DannerD3H: In my opinion I think you should consider a bigger driver for bass. Personally for a fullrange setup that can play loud and deep I think 2*8" is minimum or one 10". If you live near me, you are welcome to listen to my setup. Just send me a PM if you are interested.

Another option is to use a passive radiator. One of the best speakers I have heard in the "small" speaker category are these: S400 SMOKED OAK (PRE-ORDER DISCOUNT) — Buchardt Audio

They can play LOUD and deep. It is all SB acoustic drivers. Diffraction issues are taken care/eliminated with the custom made waveguide. See the detailed description on their page for a description of the use of a passive radiator. The passive radiator may or may not be custom made. Good think is you can actually buy it.
 
Hi Guys,

Glad the info is of interest and I will dig out some sims for good 12 inch Pro drivers that are well suited to sealed box / DSP loading.

I know what you mean about big pro drivers lacking life at low SPL.... Many are designed to handle several Kilo watts so not a good match for SET brigade!
One key driver parameter I look for which ensures superb dynamics throughout the SPL range is the Mms:Bl ratio.

Low Mms / high Bl. The lower ratio the more dynamic potential, but go too low and the low end drops off too quickly.
As a rough guide here are a few actual example with AES power handling and suggested ratio range:

150 watts AES, 8 inch - 24g Mms Bl 14. ratio = 1.7 (1.4 to 2 is a fair range.)Beyma Great all round
450 watts AES 10 inch - 38g Mms Bl 18.6. ratio = 1.4 (1.4 to 2.2 is a fair range.)
600 watts AES 12 inch - 65g Mms Bl 22.5 ratio = 2.8 (2.2 to 2.8 is a fair range.)
600 watts 15 inch - 85g Mms Bl 22. ratio = 3.3 (3 to 4 is a fair range.)
1,000 watts AES 18 inch - 219 Mms 31.5 Bl = 6.9 (5.5 to 7.5 is a fair range.) Awesome below 60Hz for home cinema.

Rocky, I will check my files tomorrow, but I am fairly sure the above 12 inch sim is for an RCF driver. Beyma, RCF, PHL, Precision Devices and 18 Sound all make suitable drivers, as do most good Pro brands.
 
Last edited:
Danner,

18W or 12MU for mid - the choice is up to you based on your priorities. It would be a toss up for me, there are pros and cons both ways. If soundstage/imaging is a high priority, although a little dated, this Loudspeaker Imaging Theorum by Jeff Bagby might be a good read.

Cardboard? If you use Troels' baffle as your model you don't really need to do any further modelling with cardboard.

I'm not sure what you mean when you say "the cap is shaped like a long tube at the rear."

About exceeding woofer excursion - there are 3 driver limits to be concerned with, xmax, max mechanical and max power rating. When xmax is exceeded, sound quality simply decreases. The 18W still has a lot more cone excursion left (4.5mm) before reaching its mechanical limits, but mechanical excursion isn't actually linear in that there is more resistance with more excursion so it's not like the driver is going to just suddenly come apart at its mechanical limits. You'll probably hear the problem of the driver hitting its physical limits before it tears itself apart. I think more likely to happen is for the voice coil to exceed its max power limits and get fried instead if you keep feeding it more and more power, although the 18W's long term max power isn't listed but you can be easily comfortable with at least 60W per driver which will give you about 100dB at 1m with an F3 of 40Hz. But at that point you have far exceeded xmax and should not be really enjoying what you hear so I don't know why you be pushing it that hard in the 1st place. My preference however for all the reasons I've already stated, would be to use 2 woofers per side.

Also note again that the closed box acts as a resistive spring against cone movement and that this force increases as the Q of the system goes higher (ie. as the box gets smaller). That actually means that a closed box offers a driver a little bit more protection against over-excursion than a ported one and more so if the box is undersized. That is part of the issue with going with a smaller box for you - will the increased resistance to cone excursion create a problem when you try to boost the sub 50Hz frequencies? Have a look at cone excursion in the 30-40Hz range in WinISD as you change box size at high power to help figure out what is going on. I don't have the experience with this situation so I can't say if it'll work, but Q=.8 isn't as extreme as Q>1 so I'm inclined to think it'll be fine. And if the higher Q air brake is a little too high, a small aperiodic vent to relieve the pressure might easily solve the problem.

But a decision to test it out is a very smart way to go.
 
#Rocky - When i eventually move to a house then the true full range speakers may be an option for me. But that won't happen fo another 5-8 years anyway.. Until then 8in and above is simply to big for me when used as a main driver. However, I have done some sims and i see what you mean.. They do go MUCH lower.

I am pretty sure I would like to try the sealed kind of speakers for this build.



#jReave - True, no need for cardboard sims when copying his baffle!

I was just thinking that the way the mid cabinet narrows in at the rear and then goes straight to the back to meet the aperiodic vet. Wont that passage to the went make it act like a long port? It may not be an issue though.

Hmmm I have had a bit of a hard time understanding how you got to your Watt numbers. When i used WinISD i got that when using two relevators i would only need around 4W to go over the linear excursion limit. I think i realise why now. When i use the filters/Eq function in WinISD to boost the bass, then the cone excursion follow but the wattages does not really. However, i have to remember that those 4 watts is for the parts that i did not boost. If i remove all filters then i get the same wat numbers as you to max out the driver.
So thanks for the explenation i think i get it now!




Okay. I have been trying out some things yesterday. First i tried to use two of SB's 8 in subwoofers instead of the revelators. I see that the bass response gets WAAAAY deeper, but for my current situation theese drivers are simply to big even though they require less volume than the 18W they still require a much bigger baffle. Why are they so much bigger? arent the 18W supposed to be 7½in? xD
I also looked at the Satori drivers sugested by a few of you guys. Also in my own sims they seem to outperform the revelators when it comes to deep bass in a compact sealed setup like this.

My conclusion is:
Given my limitation on size for the front speakers then the right thing would be to start from scratch completely regarding driver choices as the 18W arent made for speakers like this. I am now considdering to scale my cabinets back down to around 22L. Then either use one revelator and cut it higher than 40 Hz to get a decen SPL. Or sell the 18W and instead use two satori drivers.
In both cases i should build a subwoofer that i can hide under the couch for the last couble of Hz (I was thinking of 1 or 2 of these: SB Acoustics :: 10" SW26DBAC76-8 )

What do you guys think of this choice?
Will the two satoris in 22L perform better, worse or equal to two 18W in 28L? this is basically the answer i need to take the choice.
The Satori will end up with a Q of 0,7 and the 18W wil end up at 0,8.


(Alternatively, in a 3-way setup. Is the sound qualitiy affected if the 8in woofers are placed on either side of the cabinet instead of on the front?)
 
Last edited:
Hi Doppler,

First benefit is dividing one large volume into 2 equal halves will increase cabinet rigidity and decrease panel resonance.

The more subtle benefit is spreading of driver resonance.

2 drivers in one box will result in one F3 and one Fb. So whatever peaks / troughs in the response are fixed.
As no 2 drivers are 100% identical (most specs are plus minus 5% to 10%) when you place the 2 drivers in 2 separate boxes this results in two different Fb and F3 so the peaks and troughs will be at different points.... They fill in the gaps but both stay within the same response limitations.

Wilson Audio use "bass resonance balancing" in their $6 figure speakers.... They use a 12 inch and a 15 inch driver in separate chambers to cover the bass.
I would think that a perforated panel would be structurally equivalent to an otherwise identical solid panel.

I too thought of sample variance but came to the opposite conclusion.

Say one driver is +10% and one is -10% on Vas.

With separate chambers, the transient response of one will be relatively poor, assuming the box was sized for the spec. The other will have improved transient response but this may be masked by the first driver.

With a shared chamber, the errors will better cancel, with an improved overall transient response, was my feeling. These are just thought experiments – I have no interest in actually testing the theory with measurement.
 
Last edited:
Hydrogen Alex: For sure interesting stuff! Typically the Pro Audio driver dont have low Fs. Does this matter much ? - Ie is it possible to boost so that that they actually play epic good lows even if the Fs is quite high.

DannerD3H: For sure many options to consider. To be honest, you get a long way with a simple 2 way - Especially at low volume - But it will never be able to make the same sound as a bigger speaker. Relieving the driver playing low Hz ie under 150-200 in my experience does only good for the sound.

BUT a heavy woofer or subwoofer in my opinion does not sound good crossed high. At the moment i cross over at 160 Hz. With two nice drivers for bass you would be able to cross much high. But does is sound better? - Depends on taste and testing.

Also: Small midranges have in my experience a more forward sound than a bigger driver. Hard to describe. Some like it, some dont.

Dont know if you have factorized room gain. Troels Gravesen has a link on his page of a Excel sheet where you type TS parameters box size/type and distante from walls and floor. This give you a good idea of what you can expect.
 
I was just thinking that the way the mid cabinet narrows in at the rear and then goes straight to the back to meet the aperiodic vet. Wont that passage to the vent make it act like a long port? It may not be an issue though.

It won't be an issue - the port won't act like a port if you stuff it. It'll be aperiodic: instead of 2 impedance peaks on either side of the tuning frequency, you'll only get 1 smaller peak like a sealed box. Also, an inverted horn is an excellent shape to attenuate the rear wave. Again, think B&W. Whether it's best sealed, lightly stuffed or heavily stuffed is up to you - let your ears decide.

... two of SB's 8 in subwoofers ... Why are they so much bigger? arent the 18W supposed to be 7½in? xD

In the case of SB's 8" drivers, it looks like they are referring to the cone size. The frame size is 23cm, so closer to 9". I see that Madisound refers to the 18W as a 7" driver which is more or less correct for the outside diameter - 182mm. It's an apples and oranges kind of thing - there's a lack of an industry standard in terms of sizing.

I also looked at the Satori drivers sugested by a few of you guys. Also in my own sims they seem to outperform the revelators when it comes to deep bass in a compact sealed setup like this.

Check your sims again - I would actually call the Revelator a slightly better performer in the LF given the same Box Q, it just needs twice the volume to do it in. Max SPL at 40Hz is about the same for both drivers, however the Satori needs a little more power to do it. Not that a small difference in the LF response matters much in your case because you are going to be aggressively manipulating it anyways.

My conclusion is:
Given my limitation on size for the front speakers then the right thing would be to start from scratch completely regarding driver choices as the 18W arent made for speakers like this. I am now considdering to scale my cabinets back down to around 22L. Then either use one revelator and cut it higher than 40 Hz to get a decen SPL. Or sell the 18W and instead use two satori drivers.
In both cases i should build a subwoofer that i can hide under the couch for the last couble of Hz (I was thinking of 1 or 2 of these: SB Acoustics :: 10" SW26DBAC76-8 )

I am assuming that all of the above are still using the 12MU as the mid. If so, 1 x 18W sealed and crossed over to a sub or 2 anywhere above 60Hz will now easily give you over 100dB at 1m without exceeding xmax. No need really for 2 woofers in this case especially if you cross just a little higher than that. The same will be true for the Satori - there is no need for 2 of them if you are using a sub. As to which woofer is better, it may be a case of just flipping a coin, but since you already have the 18W's, I'd stick with those in this scenario.

What do you guys think of this choice?
Will the two satoris in 22L perform better, worse or equal to two 18W in 28L? this is basically the answer i need to take the choice.
The Satori will end up with a Q of 0,7 and the 18W wil end up at 0,8.

I don't think you can make that choice until you hear what the 18W can do in a smaller sealed box with the bass boost. Personally, when you are crossing the upper mid at about 700Hz and leaving a whole lot of the woofer to still handle the lower mids, I would still want to minimize the heavy slugging the woofer has to do by going 4-way with a single woofer crossed higher and letting the sub handle the big excursions. I would have mentioned this before but you stated that 1 of your design criteria was to go without a sub.

Alternatively, in a 3-way setup. Is the sound qualitiy affected if the 8in woofers are placed on either side of the cabinet instead of on the front?

For a 3-way with a side 8" woofer, the xo point needs to be lower, somewhere below 200Hz usually. It's not a bad implementation but you would have to use a larger mid that can cross that low, so it wouldn't work with the 12MU. It could work with the 12MU plus one 18W on the front though, but your size constraint probably makes that impossible. Plus a side woofer on curved sides is a more complicated construction. Not sure if you could mount all the amps necessary on 1 cabinet either. One 18W as your mid with a larger side woofer would work but that doesn't fit your design criteria either.
 
#Doppler - I think your thought experiment makes sense! only issue is what if both drivers are 10% in the same direction?

#Rocky - Yes there are to many options.. I think i have to limit my self a bit here.. Hence what follows:


#jReave - Once again thankyou for the elaborate explenation!

Good to hear about the chamber shape, and yes of course the B&W is a good example again..

Yeah incredibly annoying that they do not agree on a standard..

Yes the 18W is sliightly better at the same Q but i was refering to the same size cabinets thus a higher Q for the 18W.

Again really good to get some clarifications on all my tangled thoughts..
Yes in the beginning i stated that i did not wan't a subwoofer. That was because for the next couple of years a traditional subwoofer would be annoying in my current housing. I was originally thinking of rebuilding my sub to fit under the couch, but my couch is fairly low. I only have 15cm of air under it. While looking at SB's homepage i came acros those flat woofers and they actually fit my couch perfectly!

Based on your responses, my sims and the flat SB woofers and my love for the 18W this is currently my 3 possibilities in prioritised order:

1:
A 3 way floor stander of 20L (woofer) with my tweeter, 12MU aperiodic and one 18W sealed. This speaker will be supported by a seperate subwoofer made of two of SB's flat 10in woofers which i can place under my couch.

2:
A 3 Way floor stander with my tweeter, 18W as mid woofer with 17-18L and one of SB's 8 In subwoofers as a down or side firing unit.

3:
A 3 Way floor stander of 22L with tweeter, 12 MU, and two satori units sealed.




PS: How close to the floor can a downfirring subwoofer be without drastically affecting sound?
If i build a sub for under my couch will it then be better to let it fire upwards?
 
#1 looks like the best option to start off with, given that you already have the woofers. I would still possibly test out the woofer with boost in the higher Q cabinet with the possibility of waiting to do the subs until your new place and in the mean time, not pushing the single woofer crazy hard.

So in other words maybe taking it in steps instead of making all your decisions at once. That's easy for some people. Harder for others.

Man those are some slim subs. I don't know a lot about that type of application but it looks possible when looking at something like this build: The Boogieman. If you currently have a sub, why don't you try an experiment and set it up downfiring just an inch above the floor. See if you hear any difference.

But are you sure you couldn't find another spot or 2 elsewhere? Also with multiple subs, there are advantages to placing them in different spots in the room to better overcome room modes.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.