Seas-based Center channel...need help

[ssabripo]

I've narrowed down my WMTW Center channel design to the following components:

2 x Seas Excel W18E-001 7" mid woofers
1 x Seas Excel W12CY 001 4.5" midrange
1 x Seas Millenium T25CF 002 tweeter

I need help with the Crossover....I started the x-over design based on the Odin Mk3 (see attachement). But I'm having some trouble trying to squeeze the midrange without messing up the dynamic range of the network.

any help is appreciated....

 

[ssabripo]

I get no love....

 

[sdclc126]

I'm not a crossover expert, but perhaps you will get some responses if you define "squeeze the midrange."

As a sidebar, this is somewhat off subject, but I think many people on this forum would opt for ditching the 4.5" midrange altogether and going WTW. If the woofs and tweet can cross over to eachother within their useful ranges, some of your crossover design issues may be addressed. A simpler design, less money, and performace should be indistinguishable.

I hope this is at least somewhat useful - sometimes more is not necessarily better. Nice high-end project by the way, with the SEAS stuff - worthy of mains. Which begs the question...

 

[ssabripo]

I started thinking about the x-over design....what do you think of doing this:

using the W18E-001 in parallel, crossed over between 50hz to 1250 hz
using the Excel W12CY 001 crossed between 600hz and 6khz
using the T25CF 002 tweeter highpassed at 5khz​

this obviously has some merge areas, but given the responses of the drivers, it may actually blend very well.



sdclc126, yes the question is definitely a good one .... I currently am running a pair of swan Diva 6.1s as my mains, but they will soon be a thing of the past.

if you haven't noticed, the DIY bug bit me HARD!!! I am completing the subwoofer, and will start cutting wood as soon as the driver gets here next week.
http://www.diyaudio.com/forums/showthread.php?postid=810697#post810697

Next, is this Center channel, and if it sounds as good as I think, then I will go Seas mains as well.....

 

[Paul W]

Great drivers, but you'll need a great crossover network to go with them.

You'll need measurement capability, a simulator, and experience to do them any justice at all. If you don't have measurement capability and SW, I suggest turning the crossover job over to a true pro. Crossovers for the mag drivers are hypercritical.

An externally hosted image should be here but it was not working when we last tested it.

 

[sdclc126]

I started thinking about the x-over design....what do you think of doing this:

using the W18E-001 in parallel, crossed over between 50hz to 1250 hz
using the Excel W12CY 001 crossed between 600hz and 6khz
using the T25CF 002 tweeter highpassed at 5khz​

this obviously has some merge areas, but given the responses of the drivers, it may actually blend very well.



This is the kind of stuff I don't know about - see Pauls' post above.

 

[ssabripo]

Great drivers, but you'll need a great crossover network to go with them.

You'll need measurement capability, a simulator, and experience to do them any justice at all. If you don't have measurement capability and SW, I suggest turning the crossover job over to a true pro. Crossovers for the mag drivers are hypercritical.

http://www.geocities.com/pnwright3/Wings2001.jpg[IMG] [/B][/QUOTE]


Paul...fair enough, and same advice I've gotten from many, so I guess these drivers come with a penalty for their great performance: gotta have really excellent networks, or you wont make them shine (not as forgiving).

on that note, since I have to admit that I dont have all the necessary tools (a mic, sound card, and impedance jig, is prolly not gonna cut it), who would you suggest that may be able to lend me a hand (for relatively affordable price of consult?)

 

[sdclc126]

"On that note, since I have to admit that I dont have all the necessary tools (a mic, sound card, and impedance jig, is prolly not gonna cut it), who would you suggest that may be able to lend me a hand (for relatively affordable price of consult?)"

John Krutke at http://www.zaphaudio.com/ might do it - I think he charges $30 an hour or so (I have no idea how many hours it would take!). I don't know if he's done any designs with the SEAS magnesium drivers, but he has with their aluminum units (http://www.zaphaudio.com/audio-speaker17.html), and they are supposed to be very close to the Excels (the 27TBFCG tweeter is supposed to be BETTER).

Which leads me to this suggestion: Since this is a center channel speaker, consider perhaps a less "ambitious" design and go with the aluminum drivers, in a WTW configuration. You could use Zaph's design, add another woofer, and adapt the crossover accordingly.

This will save time, money, and frustration in trying to come up with a suitable crossover, and still sound fantastic. Paying someone to design a crossover could double the cost of the project. You can put the money you save toward your new mains!

The other alternative is to find a kit, someone else's design that is similar to your idea, or hope that someone here on the forum can offer some more advice on your crossover.

 

[Paul W]

In addition to John, there are at least a couple of guys that frequent the Madisound board...Dennis M and Jeff B are also well respected.

Personally, I'd stick to the 3-way WM/TW. An MTM would be simpler, but a professional designer will be able to give you a 3-way with much better performance. With the level of investment you are making, do it well and you won't regret it later.

 

[ssabripo]

In addition to John, there are at least a couple of guys that frequent the Madisound board...Dennis M and Jeff B are also well respected.

Personally, I'd stick to the 3-way WM/TW. An MTM would be simpler, but a professional designer will be able to give you a 3-way with much better performance. With the level of investment you are making, do it well and you won't regret it later.

I completely agree Paul, and thus my desire to go WMTW with a pair of 7" W18's, a single 4.5" WCY, and the millenium tweeter...I think that combo would give a KILLER CC if designed correctly.

something like this:

An externally hosted image should be here but it was not working when we last tested it.

but not as high (I have some height constraints in my setup) and using the Millenium tweeter instead.

I've done some Linkwitz-Riley 4th order designs in the past with success, so I don't think I will be shooting in the dark.....but again, I gotta see if the investment in "paying someone" is worth it....plus, what is the fun in that?! that's no DIY

I can always end up buying an Aerial Acoustics CC5 or something if necessary...but that is my LAST resort, since I actually want to build something.

 

[fazman]

ssabripo,
I'll throw this one at you for your mains.
What about ripping the crossovers out of your Swans 6.1's and going totally active? Maybe a digital crossover from Behringer. Since you already have the speakers already and they contain some pretty descent drivers in them.
You could go a three way with seperate amps for each woofer mid and tweet.
Many people swear that active is the only way to go.
Would probably sound waaaaay better than they do now.
For the money that you would be investing in the Seas drivers you could could get some appropriate amps to do the job active.
Maybe solid state for the woofers and tubes for the mid and tweet.
For that matter you could do the same for your center channel as well.

 

[ssabripo]

Ok everyone...

I've started the work on the Crossovers....so, first let me give you guys an update on the design:

The Center Channel will be a WMTMW design, consisting of:
2 x Seas Excel W18E-001 7" mid woofers
2 x Seas Excel W12CY 001 4.5" midrange
1 x Seas Millenium T25CF 002 tweeter

After reading thru several DIY designs using these drivers, and in several arrangements from full towers to small surrounds (and everything in between, including the center channels) I will be doing the cutoffs at:

========LPF=========HPF==========
------ 600hz-------------------------3000hz---------

So, using the theoretical values from madisound for the drivers, I am getting this:

Highpass for Tweeter (@ 3000hz, Z=6 ohm)

C1: 4.69 uF
C2: 9.38 uF
L1: 0.2 mH
L2: 0.9 mH

Lowpass for Mid(@ 3000hz, Z= (8 ohm)x2 = 4 ohm)

C3: 21.11 uF
C4: 4.69 uF
L3: 0.4 mH
L4: 0.2 mH

Highpass for mid(@ 600hz, Z= (8 ohm)x2 = 4 ohm

An externally hosted image should be here but it was not working when we last tested it.

C1: 35.17 uF
C2: 70.33 uF
L1: .67 mH
L2: 3 mH

Lowpass for Woofer(@ 600hz, Z= (8 ohm)x2 = 4 ohm)

C3: 105.54 uF
C4: 23.46 uF
L3: 2 mH
L4: 1 mH


I also noticed there was a good amount of variance in the sensitivity of the drivers between the three ( tweeter= 88 dB , midrange = 85 dB, woofer = 86.5 dB), so I put some L-pads on the tweeter and woofer to bring them down to match the mid's 85 dB setting:

An externally hosted image should be here but it was not working when we last tested it.

R1: 1.75 ohm
R2: 14.54 ohm

An externally hosted image should be here but it was not working when we last tested it.

R1: 0.63 ohm
R2: 21.22 ohm
Question: I used the theoretical values of the two woofers in parallel (Z= 8ohm x2 -> 4 ohm), so I assume the L-pads will work fine for this arrangement, since we are dealing with impedances......right?


next step is to put this into LEAP and see how it behaves......

Also, if some of the values are obviously not available from off-the shelf components, so any suggestions would be appreciated (whether it is buying the available parts, and tampering with the X-over points, etc).

 

[sreten]

Hmmm..........

its enough to make you cry.........

so expensive drivers, so absolutely no idea what you are doing.......

/sreten.

 

[ssabripo]

Hmmm..........

its enough to make you cry.........

so expensive drivers, so absolutely no idea what you are doing.......

/sreten.

Ouch

well, its a learning process..... I thought this was the whole idea of the forum, to learn and share, not to put people down

 

[sreten]

Ouch
well, its a learning process..... I thought this was the whole idea of the forum, to learn and share, not to put people down

Hi,

I'm not trying to put people down.

I'm expressing a sentiment, there's so much wrong in your post
I wouldn't know where to begin, and with some of the best drivers
in the world a slight mistake can seriously affect the outcome.

Learn - I'd say learn to walk before you try running.
Go to zaphaudio.com and try to understand the design process
of one of his designs. If you can understand them all then you
would be a long way towards understanding the many things
wrong with your current approach.

Also checkout :

http://www.humblehomemadehifi.com/
http://speakerbuilder.net/web_files/default.htm

Sharing - I couldn't see the point of taking your post apart.
(but believe me nearly everything shows much misunderstanding)

/sreten.

 

[ssabripo]

Sreten,

thanks for the links. I actually did read a lot of the info on ZaphAudio, and got some good ideas as well from some of the kits and DIYs from partexpress and madisound.

I agree that a slight mistake with the Seas Excel line will drastically affect the outcome, as they are very precise yet unforgiving drivers.....that's why I'm gonna go with many MANY iterations before I put the X-over together.

But still, I don't see what is "completely wrong" about the initial stab at the X-over. This is not rocket science....at least I didnt think so . I took a look at the drivers and their ranges, looked at other implementations of them and their frequency cutoffs, and used something similar (thus the reason of using 600hz between the woofer and mid, and 3000hz between mid and Tweeter).

I would love to hear what I did wrong....if you don't know where to begin, start from the top:

1. What's wrong with my X-over points for LP and HP
2. What's wrong with the values for the Linkwitz-Riley 4th order?
.
.
.n) what would be a decent similar X-over to start with?

 

[sreten]

Ok :

with those crossover points you don't need two midranges.

The two woofers will be 89.5dB/W when used in parallel.

allowing for some baffle step means midband is 84 to 86dB/W.

The single mid is bang on target at 85dB/W for 5.5dB of baffle step.

The tweeter needs padding back to 85dB/W.

L-pads on bass units are a disaster.

The 4th order L/R should be acoustic, not electrical.

Drive units are not resistances, they have an Re and inductance Le.
They cannot be modelled as resistors. The crossovers inductors DCR
must also be included in modeling high oder networks.

/sreten.

 

[Landroval]

Have you built the enclosures already? To be able to do a good crossover design, you should measure the drivers separately in the enclosure, and base your design on the results.

What are you using (or going to use) as main speakers? It would be important to keep the center speaker as similar as possible to the mains.

Edit: If you want to get away from the designing process and just build a predesigned system take a look at Selah Audio Pro5CC and Sardonyx center-speakers. They're quite close to your plan and certainly will not dissapoint.
http://www.selahaudio.com/id71.html

When you're ready to design your own design I suggest you star with an easy 2-driver system with simple crossover. When you learn, you can move to more advanced systems.

 

[ssabripo]

Thank you sreten....

Man, I guess I was waaaaay off . Ok, so let me ask some simple questions:

Ok :

with those crossover points you don't need two midranges.

1. I have a height constraint of 10".... that was the reason for going with a W-M-T-M-W design, so that both mids (one on each side of the tweeter) would disperse equally.... What is wrong with that design?
As an alternate, I could do 1 midrange right below the tweeter, but would have to me right on top of each other in order to fit...would this be adequate?

the mid is 4.74" in diameter, the tweeter is4.35", so I'm already at over 9" ! Add the .75 MDF on each side and I'm already close to 11"..., and thus the reason to go side by side.

The two woofers will be 89.5dB/W when used in parallel.

Yes, you are right! I forgot to do that. I will recalculate accordingly, but since you suggested that no L-pads should be necessary on the woofers, then this point is mute.....right?

allowing for some baffle step means midband is 84 to 86dB/W.

The single mid is bang on target at 85dB/W for 5.5dB of baffle step.

that's true. But unless I put a single mid right below the tweeter, which will not only push the height to 11" at minimum (see above), but will not look good aesthetically either, then what alternatives do I have?

The tweeter needs padding back to 85dB/W.

L-pads on bass units are a disaster.

I will remove the L-pads from the bass units.
What is wrong with the L-Pads I put around the tweeter? Are they wrong?

The 4th order L/R should be acoustic, not electrical.

Drive units are not resistances, they have an Re and inductance Le.
They cannot be modelled as resistors. The crossovers inductors DCR
must also be included in modeling high oder networks.

/sreten.

Ok...so I should let LEAP or lspCAD determine those for me then, instead of manually cranking it out, I suppose

Also, the 4th order network for the mids will be very complicated to build and fit on the center channel.....would you suggest going with a simpler 3rd order?

thanks in advance

 

[pinkmouse]

I don't get where you get the 10" restriction from. What about the extra volume the bass units will take up? Could you do a sketch to clarify?

Side by side mids are a bad idea as they will lead to vertical lobing near the crossover frequency, giving a very small sweet spot.