"From the formula above, it is clear..." (1 Viewer)

[Chuck Bogie]

Nope. It sure ain't.

You know, I think I've gotten more out of carefully rereading and cross-referencing a few pages of the Parts Express catalog than I've gotten out of several hours darn near sweating over the Loudspeaker "Cookbook."

Let's see if I have things more or less right - A crossover, say a Linkwitz dealie that'll do a 12, will have two sets of inductors ("L" Why "L"?) and caps (at least I hope that "C" means cap...), and the different values will allow different frequencies to pass (or not pass) through... I think that folks also use resistors to lower the volume to select speakers, so that one may use a more efficient tweeter with a less efficient woofer? Or am I mistooken?

Looking at the formulas that the guy has in there to determine the settings for a two-way 4th order, I'm wondering about more than a few things - Such as "What's a Henry?" "What's a Farad?" "Why doesn't the speaker catalog have this information, so I'll at least know which speakers to buy?"

Hey, I can do the math - but it'd help to have at least some of the numbers to plug into the equations...

Now... I'm wondering... Say I'm gonna go with a 2nd order LR crossover that's prefabbed (the blue one that PE sells) that rolls at 2,000 or 2,500 hz, why won't this work with the 6.5 Dayton woofer and 1.8" Dayton tweeter? Seems to me that, according to the table in the VD book, that 2,000 hz would be low enough for the woofer, and also according to the book and the catalog, 2000 is high enough for the tweeter. So, PLEASE, what's wrong with this picture?

What would be "wrong" with the speaker, assuming I have things correctly mounted in a properly sized box? Not asking this just for grins - I'd like to know how I'd be screwing up, without first screwing up, you know? Would the speaker sound like I accidentally sealed a cat inside, what?

And I'm _still_ wondering (the dang thing doesn't have an index... or at least I haven't found it yet...) how to determine box size if you're doing something like using two woofs, or a woof and an unsealed mid/midwoof...

 

[Anthony_Gomez]

A henry is a unit of inducance and a farad is a measure of capacitance. ..just as a second is a measurement of time and a foot is a measurement of distance.

The thing wrong with "generic" XO's is that they assume that the drivers is exactly 8ohms (or 4 or 6) at the XO point, that they phase match, that they have the same acoustic centers, have an "ideal" c-c distance, that individual frequency over lap is infinate, that there are no anomalies (like a rising responce) and that they are on a baffle that doesn't muck up anything.
...NOTHING could be farther from the truth in the real world.


An XO is the brains to speaker. you can't just drop any brain into any body and expect it to work ...well...for some, that is how it always has been

I have said this before ...MANY TIMES...and I will say it again...and anyone with even some XO designing experience will say the same.

you can have the worlds best drivers, enclosure, XO parts, electronics and an inproperly designed XO (insert generic XO) and you still end up with a crappy sounding speaker

A properly designed XO is FAR from textbook calculations. It acounts MANY MANY differnt things.

 

[Mark Hayenga]

Textbook crossovers make it so that the output of the electrical network (called its transfer function) matches a pre-defined curve (Butterworth, Linkwitz-Riley, etc). The problem is, we really don't care about the electrical transfer function of the crossover. We want the acoustic transfer function of the speaker to match those pre-defined curves. The acoustic output is simply a cascade of the crossover network's transfer functions and the drive units' transfer functions.

Textbook crossovers will give acceptable results if your speaker drivers have perfectly flat impedance and perfectly flat frequency response. This is because you're cascading the desired frequency response (ie the ideal output of the textbook crossover) with a transfer function that will not affect it (the perfectly flat freq response of the driver). The problem is, there is no perfect driver, and most are very far from it. Textbook crossovers thus give undesireable results in almost every case.

In real-world speakers, non-ideal electrical transfer functions are combined with non-ideal driver's frequency response and impedance parameters to achieve an approximation of those ideal curves for the complete system.

 

[Dave Milne]

Chuck,
Tony and Mark's comments are essentially correct regarding generic XOs. However, I hate to see someone abandon DIY speakers before they get started just because "real" XO design is "really" complex.

If you're just looking to throw some drivers and a crossover into a box to save some money, then you probably are wasting your time. But from the nature of your questions and comments

how to determine box size if you're doing something like using two woofs, or a woof and an unsealed mid/midwoof...

for two "woofs" just double the enclosure volume and the port area (if vented). Don't try sharing enclosure airspace between "a woof and an unsealed mid/midwoof". You need to build a sub-chamber for the mid/midwoof to isolate it (make sure you account for that volume in your main woofer enclosure calculations)

 

[Anthony_Gomez]

You could always play with "speaker workshop".

If you download it, then I can send you some files to play with.

 

[Henry_W]

A 'Henry' is an attractive, well mannered, house broken, southern boy...

Henry

 

[Seth_L]

That's not what your SO says Henry.

Seth

 

[Henry_W]

Well the SO is skewed - she doesn't yet know if the last thirty years is her fault or mine

Now to Chuck's question(s). Chuck, from one newby to another, Dave Milne said it well here

>>pick some parts... and try it out. It won't be optimum at first, but still probably listenable... and it's fun to see and hear engineering effort turned into reality.

 

[Aaron_Smith]

Dave and Henry are right on- if you truly believe in Doing It Yourself, it's the journey that matters- not the destination. You must remember that most people who do their own crossover designs have a pretty highly trained ear, able to pick up subtleties that most people wouldn't notice. I say make the speaker, build the standard Dayton crossover outside the speaker in a project box, and play them. Odds are they'll sound pretty good compared to the normal crap you'd get at Best Buy. Then learn a little more, build a new crossover, and BEHOLD! a big difference in sound! So THAT'S what baffle step compensation does. Then another change. WOW! Impedance correction makes a big difference.
If all you want out of DIY is a good sounding pair of speakers, in the long run it would be cheaper to buy them- seriously. If you enjoy the work involved and the learning process then go for it and start building- the mistakes you make will be your best teachers.

 

[Mark_E_Smith]

Pickup David Weems book on designing speakers, its a great place to start with out SCARING you out of trying to have some fun. I was there 2 years ago, and wouldn't do it any other way. There is good, great and perfect. You should be able to build some good sounding speakers esp considering the cash outlay. The next step is understanding and being able to measure your creation and tweek the combination to achieve a great sounding speaker. Then if you are still in the game as I, you will acquire the necessary knowledge and equipment to go to the next level, near perfection. I now have a CLIO measurement system, Harris Tech Bass Box and X-Over Pro for design and I just got Calsod for modeling speaker designs on a computer. There is no way I could have understood these programs when I started, in fact CALSOD is really complex and it may be awhile be for I am confident in using it. I just now have a good understanding of CLIO. But my first speaker, a center channel MTM sounds better that ANY commercial cennetr speaker I have heard! Its somewhere between good and great. I am in the process of designing 2 new speaker systems with a pair of subs on the back burner. After you read Weems book and build a speaker or 2 pickup Vance Dickason Loudspeaker Design Cookbook. I don't recommend it for the first book because it is very complex and is confusing for the beginner. Have fun but read a book first.

 

[Hank Frankenberg]

Chuck, you're getting the best thread of advice for a beginner that I've seen on any forum.

Admins: I recommend archiving this thread and putting a link up top near the Sonotube Projects link.

I usually recommend the Radio Shack books on loudspeaker design. There used to be two: beginner and the one I now have, "Advanced Loudspeaker Design" (I think). They are like the Weems, not too intimidating.

Henry, with the modern mid-woof driver performance available, I submit that there is no need for a three-way speaker. Some modern mid-woofs have pretty flat curves over a broad enough frequency range to give you fine music range coverage. Three-ways are complex (lots of experimenting involved) and put a LOT of parts between the music and the drivers. Well designed and built two-ways (or MTM's) will sound clean and when augmented with a subwoofer, you'll cover the organ pedal notes and the movie special effects.

Now on with the technical explanations:
A Linkwitz *dealie* requires the use of film-and-foil caps and a Linkwitz *thingamajig* requires metallized film caps. Capacitance is indicated by "C". Inductance is indicated by "L" - why not "I"? - because that stands for *current* - well, why not "C" for current? No - "C" is already taken by capacitance. Makes perfect sense.
Henry and Farad were an old vaudville comedy team. They had a routine called: "Who's On First-Order"

 

[Mark Hayenga]

Hank, you are getting a punch in the arm next time I see you And I though my jokes were bad...

 

[Henry_W]

Hello Hank (Henry?) -

Yeah-on the MTM (which is somewhat like I built my center)you are probably right. Since I am a 'horny' I really think what I am doing is nearer to a 2.5 that I see folks talk about. I used vintage tweeter and mid horn in my center (to match my front and side speakers) flanked by some very surprising mid woofs.

By using horns for my mids and highs (the horn sounds that really make my day) I then need to determine if I horn load the woofer. Do you think that my ultimate (after the Xoverless tweet/woofer I am now on, then the 2 way horn/woofer that is next) is more or less complicated? My thoughts are to build a separate (but attached) box for the woofer and get one that has a very natural rolloff from the sub and before the midhorn.