still learning...

[USAF2W2]

So, I've been doing non-stop reading lately on X-over designs, calculations, etc., but still not crystal clear on the whole spectrum. I'm starting to understand (well, i think. lol) how to calculate x-over points thanks to Ray Alden's "Speaker Building 201" book and NUMEROUS hours on the web, however, I still am not sure what is best for a center channel that I want to build. I want to build a 2 way system, but then my instincts tell me to go with a 3 way. So just by opinion and experiences, what is "best" (for lack of better terms) and why? I am currently eye-balling http://www.parts-express.com/pe/showdetl.cfm?Partnumber=297-437 (two of those guys) and http://www.parts-express.com/pe/pshowdetl.cfm?&Partnumber=275-085 for a 2 way center channel with a x-over point of 3,500hz. I hope i'm not out in "left field" somewhere! :p Just going out on a limb here, so any opinions, learning and criticism would be awesome please. thanks in advance

[arlis_1957@yahoo.com]

for horizontal alignment, it seems a 3-way Wm/tW works best. also a mt 2-way if you keep it vertical. there are, of course, exceptions to the rules. if the tweeter used will cross low enough, a more common mtm can work just fine. think aviatrix. i have also seen wwmtmww used with a projector and acousticly transparent screen. a lot of it has to do with combfiltering vs crossover frequency vs ctc spacing. the 3-way aleaviates those issues.

[greywarden]

Ok, this is what I did:

1) Look at woofers and tweeters over and over again online, thinking they would work together.

2) Order said woofer and tweeter combo

3) Join the PE Forum

4) "Holy crap, that's how you actually design a speaker box and crossover"

5) Freak out becuase I've got $100 worth of parts laying around (A1C at the time :p)

6) Try to work out a crossover with "Crossover Calculator" online.

7) Buy $160 of crossover parts that probably won't work, and will definitely sound like crap.

8) Get Deployed, so I can't work on it for the next 8 months...

SO, what I'm trying to get at, is you should read through some of the documented builds, easier $100 ones, like the Overnight Sensations, Nano Neos, maybe even the $40 GRS speaker Johnny worked out, or the other $40 Don Radick (??) worked out over the past month or so - Instead of wasting what ended up being about $400-$500 in the end, sitting at home in the basement.

Safe to say, when I get home, I'll be ordering parts for every single cheap DIY design there is and building them working my way up to the big boys, like the Statements and some of Troels' or Tony Gee's designs ($$$$$)

[johnnyrichards]

The 3.5k crossover point will not be the best option for a 2-way center channel. It has to do with nulls appearing as you move off axis in the horizontal (assuming a horizontal MTM).

The lower you can cross, the better.

Another option is to do a 2.5 way center channel, but those are inherently "full baffle step corrected" designs, and may result in a very boomy sound in most center channel installs (in entertainment center, on top of TV etc where there is still some boundary reinforcement).

As Arlis noted, the Wt/mW arrangement or a MTM with ultra low crossover point is your best compromise.

There is nothing wrong with your driver choices by the way, but I would rethink the planar tweeter and look for something that will cross lower. On a budget, the Vifa BC25SG15-04 can do 1850 Hz (I crossed them at that in the Driscoll) which would be much better for horizontal lobing. 1000 Hz can be done with the Dayton RS28F.

Ultimately, your best bet is to use a basic 2-way as your center channel if your install location allows for it. It will avoid every single issue plaguing center channel design.

[Chris Roemer]

So, I've been doing non-stop reading lately on X-over designs, calculations, etc., but still not crystal clear on the whole spectrum. I'm starting to understand (well, i think. lol) how to calculate x-over points thanks to Ray Alden's "Speaker Building 201" book and NUMEROUS hours on the web, however, I still am not sure what is best for a center channel that I want to build. I want to build a 2 way system, but then my instincts tell me to go with a 3 way. So just by opinion and experiences, what is "best" (for lack of better terms) and why? I am currently eye-balling http://www.parts-express.com/pe/showdetl.cfm?Partnumber=297-437 (two of those guys) and http://www.parts-express.com/pe/pshowdetl.cfm?&Partnumber=275-085 for a 2 way center channel with a x-over point of 3,500hz. I hope i'm not out in "left field" somewhere! :p Just going out on a limb here, so any opinions, learning and criticism would be awesome please. thanks in advance

For a CC MTM to work (w/out major lobing issues), take 13,500 (sos - speed of sound in ips) and divide it by the center-to-center spacing between your mids. Cross no higher than THAT freq.

You'll find that the lower the tweeter can cross, the better. Also, the closer you can get the mids, the better. Two options are to use smaller mids, and/or bring the mids together and put the tweeter above them (like a triangle).

chris

[USAF2W2]

Well, as usual it looks like I still have a lot to learn then. lol. Oh, and Greywarden, sounds funny but its the truth...I've done the same thing with many projects (and kind of with this new found hobby I've found in speaker building.) So what did/do you do in the USAF? for the rest of the replies that were posted, thank you again, but I do still have a question regarding crossing over the tweeter at the lowest possible freq.....why is that such a must?

[Chris Roemer]

Well, as usual it looks like I still have a lot to learn then. lol. Oh, and Greywarden, sounds funny but its the truth...I've done the same thing with many projects (and kind of with this new found hobby I've found in speaker building.) So what did/do you do in the USAF? for the rest of the replies that were posted, thank you again, but I do still have a question regarding crossing over the tweeter at the lowest possible freq.....why is that such a must?

Any time you have 2 drivers seperated horizontally by a distance that's greater than 1 wavelength (at whatever freq. you're playing), you have a good chance of hearing "comb-filtering" effects as you cross in front of them (OR, have a bunch of people spread out horizontally, like on a couch).

Imagine looking straight down at your speaker with it sitting in a shallow pool of water. Drop a pebble in the water just in front of your speaker. You'll see "ripples" radiate out which move VERY similarly to sound waves.

Now drop 2 pebbles in at the same time, maybe 4" to the left of center, and the other, 4" to the right of center. As the 2 "wave fronts" radiate out from the 2 sources (pebbles), there will be MANY places (where they overlap) where both waves reach the same point at the same time. The volume at these locations will be louder (not sure if it's +3dB or +6dB) than the "average" volume level.

There will also be MANY points where the "crest" of the left hand ripple (for instance) will reach the same point at the same time that the "trough" of the right hand ripple will. At these points, the two different waves will effectively cancel each other out (creating "nulls").

The phenomenon is referred to as "comb filtering" because as you move laterallly in front of this speaker, you'll hear all the alternating peaks and dips in response. If you plotted these, it would sort of look like a comb.

Like so . . . /\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\.

Chris

[greywarden]

Well, as usual it looks like I still have a lot to learn then. lol. Oh, and Greywarden, sounds funny but its the truth...I've done the same thing with many projects (and kind of with this new found hobby I've found in speaker building.) So what did/do you do in the USAF? for the rest of the replies that were posted, thank you again, but I do still have a question regarding crossing over the tweeter at the lowest possible freq.....why is that such a must?

Vehicle operator, doin time in Kuwait right now.

[arlis_1957@yahoo.com]

nice explaination chris.

[arlis_1957@yahoo.com]

chris, why does the wmtw not have this problem?

[djg]

Paul Carmody has the Swope HT system on his site if you like the HiVi look.

www.sites.google.com/site/undefinition/swope-ht

[johnnyrichards]

chris, why does the wmtw not have this problem?

It does, as does every driver arrangement, we just tend to cross over the drivers at lower frequencies. I am not well versed on the physics, but the lower your crossover point, the closer your system gets to "point source".

In the case of Wt/mW alignments, the mid might be crossed over at 400 or 500 Hz, maybe even lower, said wavelength is longer than the distance between the two woofers. This is especially helpful for horizontal center channels.

There might be problems on the vertical axis, which is not as big of a problem on center channels. Vertical driver arrangements generally have better horizontal dispersion, and horizontal driver arrangements have better vertical dispersion. So combining a small t/m section to take advantage of the horizontal dispersion properties is always a good center channel design decision.

[Wolf]

Or my MAX:
http://s18.photobucket.com/albums/b117/wolf_teeth_speaker/MAX/

Or my Mn-14:
http://s18.photobucket.com/albums/b117/wolf_teeth_speaker/Mn-14/

Or my Dragon Foals:
http://s18.photobucket.com/albums/b117/wolf_teeth_speaker/Dragon%20Foal/

Or my Sophomores:
http://s18.photobucket.com/albums/b117/wolf_teeth_speaker/Sophomores/

Heh,
Wolf

[johnastockman]

So, I've been doing non-stop reading lately on X-over designs, calculations, etc., but still not crystal clear on the whole spectrum. I'm starting to understand (well, i think. lol) how to calculate x-over points thanks to Ray Alden's "Speaker Building 201" book and NUMEROUS hours on the web, however, I still am not sure what is best for a center channel that I want to build. I want to build a 2 way system, but then my instincts tell me to go with a 3 way. So just by opinion and experiences, what is "best" (for lack of better terms) and why? I am currently eye-balling http://www.parts-express.com/pe/showdetl.cfm?Partnumber=297-437 (two of those guys) and http://www.parts-express.com/pe/pshowdetl.cfm?&Partnumber=275-085 for a 2 way center channel with a x-over point of 3,500hz. I hope i'm not out in "left field" somewhere! :p Just going out on a limb here, so any opinions, learning and criticism would be awesome please. thanks in advance


The process greywarden described reflects many folks who want to DIY their own speakers...including me about 11 years ago when I first found the Tech Talk forum. Pay attention to #'s 6 & 7...without any XO designing experience, you're feeling around in the dark and if you try to use any form of textbook calculations (those on-line XO calculators) or pre-made XO's, the end result will be akin to barking your shin on an unseen coffee table while "feeling around in the dark." Meaning your ears will not be happy. Any of those on-line calculators use simple, totally inadequate "textbook" algebraic equations. You can't use a fixed-number in a calculation for a number/spec that constantly varies. When a particular "calculator" asks for just the driver impedance and an arbitrary XO frequency, you'll know immediately that the end result will not be very nice to listen to. Many of any given drivers parameters and specs are variable and you must take that into account when coming up with an acceptable XO design. Fixed-number formulas won't work. Ray Alden's book is one of the best DIY-speaker-building tomes available. But you'll need more info and experience to design your own XO from scratch. Until you get comfortable with some of the freeware XO design programs, you can still learn a lot by building an existing design. In that process, you can start to learn Mr. Bagby's excellent design program, PCD. PCD uses actual measured data, which does account for the many variable parameters you'll be dealing with...if you want to continue on and maybe start designing your own speakers from scratch. djg and Chris Roemer are offering very good advice. There's many fine center channel designs. Paul's Swope is a terrific way to get your feet wet. While at Paul's site, make sure you read everything under "Resources":

http://sites.google.com/site/undefinition/diy

Gives you a pretty good idea about how to get started and the common mistakes/headaches that beginners run into and some of the ?'s that come up. If I'd had those resources in the beginning and all the great advice/suggestions that are available through Tech Talk, I would have saved a lot of $$$, time and frustration trying to get all those textbook calculations (both on-line and with my own scientific calculator) to work. None of my efforts produced anything close to an acceptable level of sound quality. I came here about ready to give up, but was told by the kind gents that were here back then that I could get a good result. First was to stop using textbook formulas and pre-made XO's. Second was to just try a proven design to give me a reference point. I did, with Wayne J.'s Dayton Budget Project. That simple 2-way rendered my own DIY efforts repurposed for other existing DIY designs. If you're set on doing your own from scratch, you'll need to get Jeff Bagby's PCD and start the learning curve to get comfortable and proficient with it. Lots of guys here use it, so the learning process is fairly easy, compared to other XO design programs. Paul even has a write-up on how to design from scratch without a measuring set-up. Chris is very savvy with this stuff, so if you follow his suggestions, you'll have a great-sounding center speaker that will rival or exceed most store-bought stuff for a fraction of what commercial name-brand speaker companies charge.

John A.

[USAF2W2]

Any time you have 2 drivers seperated horizontally by a distance that's greater than 1 wavelength (at whatever freq. you're playing), you have a good chance of hearing "comb-filtering" effects as you cross in front of them (OR, have a bunch of people spread out horizontally, like on a couch).

Imagine looking straight down at your speaker with it sitting in a shallow pool of water. Drop a pebble in the water just in front of your speaker. You'll see "ripples" radiate out which move VERY similarly to sound waves.

Now drop 2 pebbles in at the same time, maybe 4" to the left of center, and the other, 4" to the right of center. As the 2 "wave fronts" radiate out from the 2 sources (pebbles), there will be MANY places (where they overlap) where both waves reach the same point at the same time. The volume at these locations will be louder (not sure if it's +3dB or +6dB) than the "average" volume level.

There will also be MANY points where the "crest" of the left hand ripple (for instance) will reach the same point at the same time that the "trough" of the right hand ripple will. At these points, the two different waves will effectively cancel each other out (creating "nulls").

The phenomenon is referred to as "comb filtering" because as you move laterallly in front of this speaker, you'll hear all the alternating peaks and dips in response. If you plotted these, it would sort of look like a comb.

Like so . . . /\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\.

Chris


Chris, thanks for the awesome explanation. That actually brought a lot into perspective and made perfect sense.

Also, Djg, thanks for refreshing my memory on Paul's SWOPE HT. I actually think after reading all your guys' informing responses that I'm just going to build that center channel (thought about it before). Plus, I don't know about you guys, but I think it looks sexy! :p

Still trying to understand the science of X-overs and how everything ties into each other, but unfortunately it's been a rough road. I think my next step will be to start testing different components or whatever and play with that to get some "hands on" understanding or something like johnastockman also suggested. Anyone have any preferences or favorites when it comes to a software that's easy to understand and learn for speaker / x-over testing?

thanks again for the replies and knowledge.

[USAF2W2]

actually, scratch that last sentence in my last post. I'm just gonna keep playing with what I have after reading Paul's http://sites.google.com/site/undefinition/simulated-measurements . I also have Bassbox 6 Pro and I really like the program since its easy to use and understand. Anyone have any experience with X-over Pro? Looks like a great program, but not sure if I wanna spend the $$

[Chris Roemer]

actually, scratch that last sentence in my last post. I'm just gonna keep playing with what I have after reading Paul's http://sites.google.com/site/undefinition/simulated-measurements . I also have Bassbox 6 Pro and I really like the program since its easy to use and understand. Anyone have any experience with X-over Pro? Looks like a great program, but not sure if I wanna spend the $$

I'm fairly dang certain that X-over Pro does NOT use frd/zma files (regardless of what actual X-over Pro users THINK). It CAN generate impedance profiles from T/S parms, but ignoring actual driver response makes no sense at all.

To get ANY reasonable crossover design worked out (short of trial and error, and error, and error - the way it USED to be done), you MUST use a modeling tool that uses both the driver's frequency response AND its impedance profile.

$$$s would be much better spent on acquiring Microsoft Excel.

Chris

[johnastockman]

actually, scratch that last sentence in my last post. I'm just gonna keep playing with what I have after reading Paul's http://sites.google.com/site/undefinition/simulated-measurements . I also have Bassbox 6 Pro and I really like the program since its easy to use and understand. Anyone have any experience with X-over Pro? Looks like a great program, but not sure if I wanna spend the $$


Bassbox will work, as you're not dealing with the variables as with a driver's performance. I've used WinISD:

http://www.linearteam.dk/default.aspx?pageid=winisd

And Unibox:

http://www.pvconsultants.com/audio/boxmodel/unibox.htm

There's two versions of WinISD, the regular and the "Pro". Start with the basic version and if you want more, the Pro version is more comprehensive. Don't waste your money on XO Pro. While it looks like it should work, it's about as effective as using textbook formulas. I tried on my own with XO Pro, but the end results weren't even acceptable. I asked another DIY'er to try it, thinking I must've made some mistakes, but after 3 tries, still totally unacceptable. A very nice gent, Shawn A. offered to help with Jeff Bagby's PCD. Same drivers as was tried with XO Pro, a Seas 25TAFC/D tweeter and a Def-Tech buy-out 5.25" woofer. The XO topology was very different than what the Harris program came up with...and the SQ was night-and-day difference. No more harsh highs, shouty mids. From the bass response up through the high-end, just a much smoother and balanced sound. Plus, the imaging was by far more suprerior, with the speakers now being quite transparent. One of my girlfriends liked them so much, she now has them in her art studio. Lots of compliments from customers and visitors. I'd build some more for those folks, but the drivers are NLA.

That's the difference in using measured data as opposed to T/S parameters (XO Pro) for XO calculations. Jeff's program needs Excel and you don't need a measuring rig as outlined in one of Paul's excellent write-ups.

http://audio.claub.net/software/jbabgy/jbagby.html

A lot of the guys here use it, so if you have any questions along the way, these guys can help sort it out. I've built a number of OPP's, other persons designs, and I can attest to the quality of the XO designs Jeff's program generates. And it's free. The learning curve is much shorter than with other XO design freeware, like Speaker Workshop. If you want to get to measuring your own drivers, the guys here can point you to the right gear needed. Good to have you on board and if you have any ?'s, no matter if you think they're too dumb or redundant, ask away. I can't help with XO designing, but through those years of frustration trying to get textbook/pre-made/XO Pro type XO's to work, my cabinet-building skills have improved tremendously. I've pretty much made all the mistakes a beginner or noob can make, so if you want, I can help in those areas. Betcha can't build just one (pair)!

John A.