My attempt at 4-way digital active

Hey Guys,

I've been around a few threads asking questions and stuff but wanted to start a thread specifically for what I'm trying to build.

What I want to end up with is a 4-way fully active setup, only for two channel music. Below is what I'm working with as of right now.

2 15" subs in sealed, 120 liter enclosures.

A set of Home Theater Direct Level Four tower speakers. I also have a set of matching bookshelf speakers sitting in the closet.

My system will be PC based and I already settled on JRiver Media Center for all DSP functionality.

I already purchased a MiniDSP U-DAC8 to take the 8 channels of digital from JRiver and turn them into analog feeds into my amplifiers. I have it up and running already and I'm using JRiver to cross over from my 2 big subs to my mains, at 90 Hz. I've also messed around a bit with REW to smooth the response of my subs pretty well, for my tastes. I'll post the response which was measured at my listening position.

I was planning to start with a well known 3-way DIY design and just try to replace the passive crossovers with active. But I changed my mind and now I want to start with a pair of mid bass modules, to address the severe lack of low frequency output from my current mains. Something to cover from about 60 Hz to 400 Hz or so.

The only modelling software I even know how to begin using is WinISD. I've been playing around with different 10" and 12" woofers and I'm really close to going with dual Dayton RS270 10" woofers on each side. But I'm hoping you guys will check my work here and see if I'm messing this up.

Below is what I'm coming up with in WinISD. I specified 2 drivers in a 4.5 cubic foot closed box. The SPL and cone excursion graphs are with 50 watts of input signal, 8 feet from the speaker at zero degrees from the direct line of the drivers. I added a second order Butterworth low pass at 400 Hz. There is no high pass in the model. If these graphs are trustworthy I'm thinking I might not need one.

Right away I'm not sure how much of this to trust since I don't see anywhere to specify whether I will wire the drivers in series or parallel. Parallel is what I want to do, for 4 Ohms.

But here's what I'm getting. If the numbers are right I get 107 DB SPL at 50 watts without exceeding xmax at any frequency. But the Qtc seems a little high at .74.

Transfer Function Magnitude
TransferFunctionMagnitude_zpsyart95wb.jpg



Transfer Function Phase
TransferFunctionPhase_zpsx6d9dkxb.jpg



Group Delay
GroupDelay_zpsmnvwuc1v.jpg



SPL
SPL_zpsgopjexje.jpg



Cone Excursion
ConeExcursion_zpsdp74wimx.jpg



Driver Impedance (Can this be right without considering how they're wired?)
Impedance_zpsy3fodrsb.jpg
 
Last edited:
I spent yesterday afternoon modeling my listening room with Sketchup. That way I can play around with where I might be able to place my subs later.

It is completely to scale. I even found the right brand and model of home theater recliners to download. The room isn't as neat and tidy as it looks in the model though. There's other stuff in there and my desk has computer speakers, headphone DAC and headphone amp, etc.

This overall build is about the sound while sitting in the two recliners most centered in the room.

-Chris

Here's the response of the two subs sitting where they are in this model. Green line is with no EQ and red line is after 30 minutes of frequency sweeps and adding PEQ.
SubSweep2_zpsbablyogm.jpg



BonusRoom1_zpso2zwtwv7.jpg



BonusRoom2_zpsvow7xdie.jpg



BonusRoom3_zpsltg8qhic.jpg
 

krivium

Member
Paid Member
2009-10-13 2:43 am
. But the Qtc seems a little high at .74.

Use an LT in Winisd (linkwitz transform, you'll find it with eq) if you want to change Qtc to whatever you target.

If these graphs are trustworthy I'm thinking I might not need one.

It will depend on what you are trying to achieve and behavior of you sub. Acoustical slope is not the only factor to take into account (this is what you see with your spl graph: intended response of speaker+box).

You want to cut at 400Hz, what do you plan up there?
 
Aside from the series or parallel wiring thing in WinISD something else throwing me off is the enclosure volume. Parts Express recommends 1.38 cubic foot for 1 driver, sealed. So maybe round that up to 3 cubic feet for two of the drivers in the same box.

But WinISD shows a Qtc of .833 at 3 cubes. Should I just ignore that and pay attention to the response curve, which still looks fine to me with the much smaller box? I can do 4.5 cubic feet but I certainly don't mind a smaller enclosure.

Thanks,
Chris
 

krivium

Member
Paid Member
2009-10-13 2:43 am
They'are 8 ohms drivers, run them parallel: you'll gain 6db for free (3 for doubling membrane surface, 3 from halving impedance). Your amp should be able to take care of them.

Variations from simulation programs and advice are normal. But don't worry, it'll give you a 'target' and as you run active you can adapt your final box to your needs (you can run an LT easily from an active setup given your amp have enough power).
In the end you will take measurement and you 'll shape using LT/eq to your desired response.
 
Use an LT in Winisd (linkwitz transform, you'll find it with eq) if you want to change Qtc to whatever you target.



It will depend on what you are trying to achieve and behavior of you sub. Acoustical slope is not the only factor to take into account (this is what you see with your spl graph: intended response of speaker+box).

You want to cut at 400Hz, what do you plan up there?

Thank you.

What would I need to measure and understand better about my subs? Does it mostly come down to group delay over the frequency range that overlaps around the crossover point? Or is there something more I'm really missing?

To start I will just have a bookshelf speaker sitting on top of each mid bass module, to handle 400Hz and up. I like the mid range of these HTD speakers and could easily live with that for months. Then I have a lot of research and reading to do. My first thought is something like the Scanspeak 12MU mid and a Scanspeak tweeter but I want to read up more on horns before making that decision.

Right now it's about getting mid bass handled though. It wouldn't bother me to throw away whatever mid bass enclosures I build and replace them down the road but I want to make a good choice on the woofers. The Dayton drivers look really good to me. I have a set of the 8" RS225 drivers in my car for mid bass and I like them.

-Chris
 

krivium

Member
Paid Member
2009-10-13 2:43 am
What would I need to measure and understand better about my subs? Does it mostly come down to group delay over the frequency range that overlaps around the crossover point? Or is there something more I'm really missing?

First thing is more about acoustic of your room: do some search about Schroeder frequency. ;)

When choosing a cross over you could decide to favor transient behavior, frequency overlap, phase coherency,... For each kind of filter you'll have a set of trade off... It's up to you to decide what does (or doesn't) matter from your taste. As you are running active it is easier to try many kind (just some clic and sometimes a bit of calcul) as long as you know approximately at which fr you want to cut.

Be aware about something as you say you won't throw off your drivers: depending of your physical layout drivers can exhibit different behavior frequency wise (from horizontal or vertical stacked for example, as a d'appolito arrangement,...). You won't throw your driver away but potentially the box...
 
They'are 8 ohms drivers, run them parallel: you'll gain 6db for free (3 for doubling membrane surface, 3 from halving impedance). Your amp should be able to take care of them.

Right, that's exactly what I intend to do. I just don't know why WinISD isn't asking me how I will wire them. I don't know if they just assume parallel and calculate based on that or what. When I switch from one driver to two, their SPL calculation goes up by about 3.5 DB. So I'm not sure what they're doing behind the scenes.

I'm OK with SPL predictions being off a few DB but I don't know if it matters which way they do it when it comes to phase calculations and such.

-Chris
 

krivium

Member
Paid Member
2009-10-13 2:43 am
Right, that's exactly what I intend to do. I just don't know why WinISD isn't asking me how I will wire them. I don't know if they just assume parallel and calculate based on that or what. When I switch from one driver to two, their SPL calculation goes up by about 3.5 DB. So I'm not sure what they're doing behind the scenes.

It seems they use double the membrane surface and don't care about Z. ;)

Anyway this program is all about box/driver simulation: what happens with your amp the prog doesn't care.

You could try to model using this link too:

http://www.petoindominique.fr/php/mysql_menu.php

It will give you other view about the same thing... including impedance which should be closer to reality. But it's in french... Not really difficult to understand, we talk about same things anyway. Use 'calcul d'une enceinte close' and fill the requested fields... should be ok and easy to understand. :)
 
Last edited:
Here's what I started playing around with last night for an enclosure. Not going to get too carried away with that until I'm fully comfortable with my driver choice though. Once I've got drivers and a target volume nailed down I'll finalize the overall dimensions and add internal bracing with holes for airflow.

-Chris


MidBassModule_zpsn0zrdpip.jpg
 
The weather is awesome here today so I think it's time to get off the computer for a while and take a motorcycle ride through the hills.
-Chris

Life offers a rich mix of dynamics that can rock and roll your emotions. JUST LIKE 97db/watt speakers :)

You room layout with central seating within irregular ceiling/walls fits the 90H x 40V controlled directivity of a waveguide like the SEOS-12. DSP driving individual sealed volume drivers allows great transient response with room-corrected flat-SPL.

You have read the waveguide threads....worth a second look during this final quest. These are "the real deal" in controlled directivity dynamic sound.

~$700 - $750/speaker for 97db/watt dynamics
2* Peavey LoRider 18" woofer = 2*$220 = $440
1* Eminence DeltaPro12 $130 or Faital 12FH530 = $245
1* SEOS-12 waveguide + Denovo 1" compression driver = $28 + $80 = $110
 

Attachments

  • BIG is Relative.jpg
    BIG is Relative.jpg
    110.2 KB · Views: 317
That was a nice 100 mile ride this afternoon.

I've read about the Schroeder frequency before and have a decent understanding of room modes and reflections. That's the main reason I'm looking at so much cone area for modules I only need to play down to where I'll cross them over to my subs.

I want plenty of headroom so I can cut with PEQ and maybe even boost a couple spots, to get the best overall response at my listening position. I learned a good bit about my room while tuning those two subs. When sitting in those recliners, there is practically no bass at all from those tower speakers. I think it's probably the sloping roof line going up to the 10' ceiling in the middle.

But it ended up being pretty easy to get a nice FR from the subs when measuring from there. It just takes some brute force to get any real volume right there. With those subs playing at a level that sounds good compared to the mains, you don't want to be right by a wall in the room. It's loud.

So I'm looking for brute force out of the mid bass modules as well. I don't mind if it ends up being overkill as that will just mean the woofers aren't having to work hard.

-Chris
 
Life offers a rich mix of dynamics that can rock and roll your emotions. JUST LIKE 97db/watt speakers :)

You room layout with central seating within irregular ceiling/walls fits the 90H x 40V controlled directivity of a waveguide like the SEOS-12. DSP driving individual sealed volume drivers allows great transient response with room-corrected flat-SPL.

You have read the waveguide threads....worth a second look during this final quest. These are "the real deal" in controlled directivity dynamic sound.

~$700 - $750/speaker for 97db/watt dynamics
2* Peavey LoRider 18" woofer = 2*$220 = $440
1* Eminence DeltaPro12 $130 or Faital 12FH530 = $245
1* SEOS-12 waveguide + Denovo 1" compression driver = $28 + $80 = $110

Honestly I fear the pro sound stuff as I just don't have a feel for how it does in a home stereo setup. I've heard plenty of live music at small venues that I thought sounded good so it's probably not the most rational fear but I still have it.

But I think my subs and mid bass modules can end up in the same place, just with a higher power requirement. I'm really interested in wave guide options for the mid range and treble though.

-Chris
 
I don't know why you specifically want 4 way when 2 or 3 might be more coherent (I now guess that one of the way's is for the sub so I now presuming you are really looking at a 3-way)

Anyways....next time you are riding, you can make some plans and buzz on over to Knoxville. We'll listen to some 2-way horns to see what you think about an all horn pro/cinema speakers in the home. (to be blunt it kicks azz)

If you like it....it was a successful bike ride. If you don't like it....it was a successful bike ride! Win / Win!
 
I don't know why you specifically want 4 way when 2 or 3 might be more coherent (I now guess that one of the way's is for the sub so I now presuming you are really looking at a 3-way)

Anyways....next time you are riding, you can make some plans and buzz on over to Knoxville. We'll listen to some 2-way horns to see what you think about an all horn pro/cinema speakers in the home. (to be blunt it kicks azz)

If you like it....it was a successful bike ride. If you don't like it....it was a successful bike ride! Win / Win!

Yep, I almost called it 3-way plus sub but I do have 8 channels to play with so I could do stereo subs in a real 4-way.

Wish I was closer to Knoxville as I would take you up on that offer in a second.

I don't think it's possible to ride a 1000 cc sport bike through the curvy back roads around here and not have a smile on your face.

-Chris
 

krivium

Member
Paid Member
2009-10-13 2:43 am
I've read about the Schroeder frequency before and have a decent understanding of room modes and reflections. That's the main reason I'm looking at so much cone area for modules I only need to play down to where I'll cross them over to my subs.

So you know that Shroeder frequency is room dependent. As it is variable from room to room it could be interesting to take that in your sub FC into account and use it to play with you rather than against you.

Stated differently you could choose to cut your sub at Schroeder frequency where your room became the main factor for sound reproduction. Is it going to be 60hz? I doubt from the sketches you provide: it'll probably be way higher.

Your comment about cone surface make sense (my point of view about bass 'quality' is that it's all about moving air and so all about moving area) but keep in mind that another strategy is to use distributed sub (or bass) to deal with acoustic limitation. That's why all the comments about the placements of your subs. ;)

If your looking about great sound and you are not too limited about your room (waf,usability,ect...) one thing to understand is that your room and speakers acts as a system and you must take into account the whole for a great result.

When sitting in those recliners, there is practically no bass at all from those tower speakers. I think it's probably the sloping roof line going up to the 10' ceiling in the middle.

Yeah, you are on the right track! This is probably a placement problem of your listening position. The roof may not be the culprit (in fact having a cathedral roof can be a desirable feature, a double one as you have a little bit less but it can be manageable don't worry about that now), you probably experience a mode problem. Have you ever listened to your tower outside (without wall to mess up the fr with reflections and modes)? It could give you an insight of what is happening in your place. Try experiment changing the placement of the tower/listening spot: it could give you important clues about your future project.

But it ended up being pretty easy to get a nice FR from the subs when measuring from there. It just takes some brute force to get any real volume right there. With those subs playing at a level that sounds good compared to the mains, you don't want to be right by a wall in the room. It's loud.

This IS an evidence your experiencing room mode! Usually bass content is louder on boundary of a room. ;)

So I'm looking for brute force out of the mid bass modules as well. I don't mind if it ends up being overkill as that will just mean the woofers aren't having to work hard.

That's ok BUT keep in mind that adding ways to a system usually make it more and more difficult to tune: others parameters play a role as directivity, interaction between ways at fc, time alignement,... That's the reason Coytee advice about coherency of lesser ways. ;)

About your fear of pro drivers this is not rational at all! In Pro driver what you should remember is Pro! :)
They are not worst than hifi drivers for most of theyr feature and they have definitive advantage regarding robustness and efficiency! And if you are looking for headroom this is definitive advantages.
They are different in the fact they are usually designed to play a little 'higher' than hifi counterpart (if you take 12 from hifi and pro and compare them you'll see that pro usually won't go lower than hifi) but that's all.

And if you fear aboout quality... do you know TAD drivers? :)
No? Ok visit that:

History

Then dream about that:

RM Monitor

You should not fear pro gear after... :rolleyes: :D

Probably even more after a ride to listen to Coytee's system! Nice invitation, you should go to hear what's it's all about if you never heard an horn hifi system... can be a (an audio) life changer.

Linesource: i like the text in the picture you linked!
Do you know more about the cabinet with the girl in the pictures? It looks like an Horbach/Keele layout! Or a (BIG!) Dunlavy cabinet... I would love to hear one (Horbach/Keele).
 
Last edited: