Baffle width difference in a 3 way

[willy65000]

Hi folks. I have a question about my plan to change the bass module on my active diy 3 way. A small two way cabinet with a midrange and tweeter currently sits on top of a bass module that has two Dayton RS225 8 woofers. The top of the bass module tapers to match the width of the 2 way. I would like to build new bass modules with 12 inch woofers so the upper and lower cabinets will have wildly different widths. I know there are drawbacks to this configuration but can they be dealt with using dsp or large round overs? the B&W Matrix isn't the look i am aiming for but an example of a similar configuration.

Thank You

 

[tmuikku]

Hi, what the drawbacks would be?

I think there is no drawbacks for sound, quite the opposite as long as you design it as a complete system. My bass boxes are even bigger, 60cm wide. Separate top allows wide bandwidth high headroom system. Bass can be hid into furniture if one wishes, while easy positioning (toe-in) the top.

 

[DatSpeakerTho]

I have mechano23 on top of 12" bass modules. 13" vs 7" baffles. Never posed a problem as far as I can tell.

 

[DanielT]

Hi folks. I have a question about my plan to change the bass module on my active diy 3 way. A small two way cabinet with a midrange and tweeter currently sits on top of a bass module that has two Dayton RS225 8 woofers. The top of the bass module tapers to match the width of the 2 way. I would like to build new bass modules with 12 inch woofers so the upper and lower cabinets will have wildly different widths. I know there are drawbacks to this configuration but can they be dealt with using dsp or large round overs? the B&W Matrix isn't the look i am aiming for but an example of a similar configuration.

Thank You

You have active speakers and, I suspect, the ability to set the crossover frequency according to your taste and preference between the top speakers and the bass modules? Can you choose the slope on the crossover as well?
If you have those possibilities, you can work back and forth, theoretically and practically. Place the top speakers physically like the KEF speakers as in your attached picture so they are time aligned (if you can't or don't want to set a delay via the electronics/in the digital world).

Bass modules - top speakers c-c distance 1/4 wave length with 4th order crossover filter should be fairly optimal. What the crossover point is then you have to calculate.

You might like 150 Hz, 300 Hz or if it works with your woofer for example 500 Hz crossover point. Listen and measure (if you have the opportunity to do so).

If you choose a lower crossover point, maybe downfiring the bass woofer? Nice with acoustic bass support from the floor. However, around 500 Hz I wouldn't have had downfiring woofers. A bit strange that singers sing straight down to the floor so to speak.

Keep the current bass boxes with their dual 8 inch woofers and add a really fat subwoofer. Convert the system to a 4 way solution. Could that be something?

Those were just some ideas I had now.

 

[Fledermaus]

Or, as long as you don't cross above say 200 Hz, you can use dual-opposed, side firing woofers for your bass module, which allows to adjust front-facing sub width to mid-high module's (on top of twice Sd and mechanical advantages of bipolar setting).
That being said, no inconvenience IME with cabinets of different width between bass and mid-high modules.

 

[Ramon Cota]

I have mechano23 on top of 12" bass modules. 13" vs 7" baffles. Never posed a problem as far as I can tell.

i want simillar . what is your crossover frequency and baffle step correction ?

 

[willy65000]

I have two SB34NRX75 6 woofers waiting for enclosures and will probably go with a front facing configuration and for sure sealed. Opposed side firing woofers would be great but more complex and expensive . Also,i have a Flex 8 so plenty of adjustability. The current crossover point between the satellites and modules is 280 but i suspect i will need to lower it. I should have mentioned that the system does have 3 subwoofers too so they don't need to play deep. I'm struggling a bit with enclosure size and do have the peramiters entered into Winisd i forgot how to look at excursion in Winisd .Enclosure size isn't much of a concern because this is s dedicated listening room so asthetics aren't a factor.

 

[DatSpeakerTho]

i want simillar . what is your crossover frequency and baffle step correction ?

There is no BSC to be considered at this stage.

My 12" is at the top of it's box, so about as close to the mechano woofer as it can get. I can use just about any xover I want between 80-400hz. I tend to keep things around 100hz. 85 sounds too low to me. If you take it higher in my room it starts to get kind of hard to knock out some room resonances I have and I like how things sound when the lower mids do not have a crossover in them.

Best part about these setups is being able to just swap out whatever top speaker I want and not have to worry about it's bass performance which is always tuned to the room and ready to go.

Two Cheap SD315a, mechano23, DSPi based dsp uniti, and twp amps. Really inexepensive way to get 20-20khz with low distortion.

Quick sweep at LP. I can make it better but I'm lazy.

 

[Ramon Cota]

ah i see , with this low crossover point you may not need bfc , but i will want to cross between 200-300 hz to better preserve that punchy base from large bass driver and at that frequency i will have to figure out compensation.

 

[DatSpeakerTho]

ah i see , with this low crossover point you may not need bfc , but i will want to cross between 200-300 hz to better preserve that punchy base from large bass driver and at that frequency i will have to figure out compensation.

BSC for what exactly?

 

[Ramon Cota]

according to my preliminary simulations , there is 4.5 db difference between 250 hz and 100 hz on a 420 mm baffle , and if your bass driver has a rising response this might be even bigger , Now imagine applying an active lr24 crossover to that curve at 250 hz , it would not result in a coherent flat bass response i think .

Now look at bs corrected curve for bass module only , assuming the mechano 23 measures anechoically flat ( we know they do ) i will get much better linearity .

 

[tmuikku]

That's why you'd always design crossover filters for actual measured data, the target is acoustic. Might take arbitrary electrical filters to achieve acoustic LR24.

 

[DanielT]

according to my preliminary simulations , there is 4.5 db difference between 250 hz and 100 hz on a 420 mm baffle , and if your bass driver has a rising response this might be even bigger , Now imagine applying an active lr24 crossover to that curve at 250 hz , it would not result in a coherent flat bass response i think .

Now look at bs corrected curve for bass module only , assuming the mechano 23 measures anechoically flat ( we know they do ) i will get much better linearity .

Build the bass boxes with the woofer mounted at the top. Test and listen, feel free to measure, set different crossover points.
Turn the bass box over and the woofer ends up near the floor (still facing forward towards the listener). Do the same then, listen test and so on.
Advantage of near the floor is the acoustic support it provides. There may then be no need for any baffle step compensation. It is also good because you do not have to reduce the efficiency of the speaker system. Plus it can have a positive impact on SBIR aspects.

Disadvantage may be that the sound is not perceived as integrated as with the woofer at the top near the top speaker. However, it has a lot to do with the crossover point and slope of the filter.

Plus using speaker calculation programs to work is also something you should do. I don't need to mention that to you of course, it says it mostly for the others who read this thread. Who are probably already into DIY and know the basics about the importance of it but still.
(by the way; theory interspersed with test boxes, different iterations, measurements, more theory, more iterations/test boxes, more tests, measurements and so on,... if you want to build SOTA full-range speakers)

 

[DatSpeakerTho]

according to my preliminary simulations , there is 4.5 db difference between 250 hz and 100 hz on a 420 mm baffle , and if your bass driver has a rising response this might be even bigger , Now imagine applying an active lr24 crossover to that curve at 250 hz , it would not result in a coherent flat bass response i think .

You're missing the drivers response sim'd on the baffle. Without that the diffraction sim isn't of much use in telling us what kind of corrections need to be made. Some drivers are tuned to give a neutral response when being placed on a baffle, the compensation being factored into the drivers response already. It would also appear that you are simulating an 8" driver on your baffle, not a 12" .

MY SD315A drivers do not need any compensation. Pro tip, ditch edge software, and xsim, and any other dated tool you might use, and just use Vituixcad, it's where the cool kids are at

I don't really actively factor in BSC for any speaker really, it's kind of just a consequence of making the speaker linear, but for bass modules I think it's easier to ignore it all together and just measure in room. If you linearize the drivers acoustic performance in room you can apply whatever xover filters you want and get predictable results. I could manually figure out baffle losses and apply a high shelf to fix it, but this gets to the end result faster.

RIP 50-100hz

Of course there a million ways to go about this, I just found this to be a quick and easy way to do it. I was also getting a little frustrated with my manually applied filters, could never knock down room resonances that well. REW EQ filters seem to be a better job. These large bass modules really tell you a lot about what your room is doing in the 200-300hz range and it's never pretty.

 

[Ramon Cota]

thanks for the enlightenment , i used to think that in room measurements of low frequencies are grossly inaccurate and thats why people do the tricks with merging gated measurements and nearfield woofers responses . i guess not , huh.
i know for sure i wont be able to lift my bass modules few meters off the ground to measure them :/

 

[levimax]

thanks for the enlightenment , i used to think that in room measurements of low frequencies are grossly inaccurate

In room LF measurements at the LP are definitely grossly inaccurate. The only way to get reliable LF measurements (or any speaker design measurements for that matter) is outside (or Klipple or Anechoic Chamber). For LF "room correction" measurements (not the same as speaker design measurements) MMM works well for magnitude.

 

[DatSpeakerTho]

thanks for the enlightenment , i used to think that in room measurements of low frequencies are grossly inaccurate and thats why people do the tricks with merging gated measurements and nearfield woofers responses . i guess not , huh.
i know for sure i wont be able to lift my bass modules few meters off the ground to measure them :/

LF in room measurements are inaccurate if you trying to get clean reflection free data, but that isn't the goal here. My argument is that it's much easier to just EQ the bass module flat while it's in room, than it is to try and measure the modules response anechoically. The way I see it is my bass modules are basically there to cover the regions the room impacts the most, so ~500hz down. You can always do the leg work and figure out what the modules response is anechoically and have a more neutral starting point, but you'll still likely need some hefty EQ so you might as well just do it all in one go.

TBH, if I could do it all again I'd just make subs for the corners and towers for everything else. The bass modules are nice if you want some oomph in the lower mids that you can't get without a big driver, but it also tends to mean your bass response is compromised as you have to pull the bass out from the walls. My bass response has far less nulls if I can place the subs in the corners. At some point I will probably change it all to the tower and corner subs.

 

[a4eaudio]

merging gated measurements and nearfield woofers responses . i guess not , huh.

Careful merging of gated farfield and nearfield measurements can get you very close of an anechoic chamber or Klippel. Best method for low frequencies is ground plane measurement, if done right.

 

[levimax]

I have some very large subs co-located with my mains which is similar to what you are doing and here are a couple of things I have learned:

1. I treat the 2 speakers as 1 full range speaker so I work to get the drivers as close together as possible (1/4 wavelength or less). This requires me to place the subs on their sides. 1/4 wavelength spacing allows the use of relatively high very steep linear phase crossovers (as long as you use a complimentary crossover) as the pre-ringing cancels. I have found that optimising the speakers as a point source with accurate time domain behavior which gives you a unified wave front, much bigger sweet spot, and better time domain behavior to sound better (slam and bass transients) than splitting multiple subs around the room. YMMV.

2. In my case I did have to make some allowance for baffle step as the baffles of the subs and mains are physically aligned and did change FR.

3. One surprize I had was that the mains were physically coupling to the subs and there was a cabinet resonance in the very large sub enclosure that was being excited. It was much higher than the subs would play by themselves so I never saw it in testing the subs. It showed up as a FR peak when testing speakers together that did not show up on the mains by themselves. I used sorbothane pads between the mains and subs and that solved it.

I think your concept is fine and gives you a lot of opportunity to experiment. I would still always measure each combination as unexpected interactions can happen. Have fun.

 

[willy65000]

I have finished the assembly of the bass module enclosures. Still lots of finish work to do but it's coming along. Does anyone have a suggestion on what i should line the enclosure walls with? Zaph uses this woofer in a similar size enclosure and uses 5/8" open cell foam and also uses acoustistuff. I can't seem to find this foam and am wondering if there's another option that would work for absorbing mid and high frequencies.