Building 3-way speakers

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
You are very far away of admitting or accepting what others tell/told you (advising and counseling from experienced diyAudio members, not me).:cool:
Right now I have all the drivers that mdocod has recommend for me to buy in my cart, so how is that not accepting? I'm working on a tight budget I can't afford some of the stuff the others have recommend. I'm also going to build my own crossovers which was recommend by everyone so what are you talking about?
 
You don't need the impedance of the drivers to match at all. The celestion driver you have in mind there would work just fine. Just have to account for the impedance differences and power consumption differences when designing the x-over. It's seriously no big deal. If you took all of your concern about ohms and watts and shifted that concern towards proper integration with a well design x-over then you would be much better off.
 
You don't need the impedance of the drivers to match at all. The celestion driver you have in mind there would work just fine. Just have to account for the impedance differences and power consumption differences when designing the x-over. It's seriously no big deal. If you took all of your concern about ohms and watts and shifted that concern towards proper integration with a well design x-over then you would be much better off.
Gotcha ok these are the drivers I am looking at getting for the project then:

Vifa BC25SG15-04 1" Shielded Dome Tweeter 264-1026
http://www.parts-express.com/pe/pshowdetl.cfm?&Partnumber=295-334&scqty=4
Celestion TF0615MR 6" Professional Midrange Speaker 50W 294-2054 (Just because it's sealed back so I don't have to make an enclosure.)

My dad does have some mids and tweeters currently but they are just radio shack replacements a piezo tweeter and sealed back so it would probably be a good idea to just get these new ones right?
 
Last edited:
I was going with the goldwood because its less wattage than the ones you posted and sealed because the ones you posted were not sealed don't I have to make some sort of separate enclosure for them? He only has 100wrms @ 8 ohms / 150wrms @ 4 ohms to work with so having the woofers which are 160wrms and a mid that's 100wrms seemed like way too much...
Right now I have all the drivers that mdocod has recommend for me to buy in my cart, so how is that not accepting? I'm working on a tight budget I can't afford some of the stuff the others have recommend. I'm also going to build my own crossovers which was recommend by everyone so what are you talking about?
My bad...
 
I'm having a bit of a problem with the celestion driver... The specifications claim that it is an 8 ohm driver, but the impedance chart provided by the manufacturer is far from dipping into the 8 ohm territory. In fact, it seems to chart more like a 16ohm driver. This discrepancy makes it hard to move forward since they claim that the SPL chart is based on "1W" input. In other words, we have no frame of reference to discern whether they used 2.83V like a normal 8 ohm driver would be tested at, or something else.

Another interesting discrepancy, is the 10" woofers, however, I have already solved the problem.... The chart for these drivers suggests a sensitivity of about 93dB, however, they are actually only ~88dB sensitive based on T/S perimeters. I was curious about the discrepancy and contacted PE tech guys to ask about it earlier today. They acknowledged that the driver was probably charted incorrectly, 5dB over what it should read. We can fix this with an adjustment to the SPL trace so it's not a big deal.

I would not want to move forward with a simulation on the celestion driver until I had confirmation of the impedance and SPL charts being accurate.

Eric
 
I'm having a bit of a problem with the celestion driver... The specifications claim that it is an 8 ohm driver, but the impedance chart provided by the manufacturer is far from dipping into the 8 ohm territory. In fact, it seems to chart more like a 16ohm driver. This discrepancy makes it hard to move forward since they claim that the SPL chart is based on "1W" input. In other words, we have no frame of reference to discern whether they used 2.83V like a normal 8 ohm driver would be tested at, or something else.

Another interesting discrepancy, is the 10" woofers, however, I have already solved the problem.... The chart for these drivers suggests a sensitivity of about 93dB, however, they are actually only ~88dB sensitive based on T/S perimeters. I was curious about the discrepancy and contacted PE tech guys to ask about it earlier today. They acknowledged that the driver was probably charted incorrectly, 5dB over what it should read. We can fix this with an adjustment to the SPL trace so it's not a big deal.

I would not want to move forward with a simulation on the celestion driver until I had confirmation of the impedance and SPL charts being accurate.

Eric
Wow you are going all out to help me thank you for all the help. Ok use the http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-020http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-020 driver then (might as well stick with dayton) if we have to build small little sealed boxes on the back of the front that is fine I'm sure I have enough mdf to do that. No circular or table saw tho so might have to use a jig saw.
 
Last edited:
Hi,

This thread is far too much "discussion" and bugger all common sense.

Appallingly bad mixture / choice of drivers and utter garbage talked about
the c/o details and sensitivity matching of the drivers. Its the blind being
led by those who agree with them the most, a certain recipe for disaster.

It has been a lost cause for pages, and a fine example of how not to do it.

Sorry for being so blunt, I just can't believe where it has ended up.

rgds, sreten.

The Tarkus design will completely slaughter what it being considered,
which IMHO is not even worth building as it will be so badly wrong.
Don't mess around with difficult drivers if you don't know what your doing.
 
Last edited:
Ex-Moderator R.I.P.
Joined 2005
Sorry for being so blunt, I just can't believe where it has ended up.

rgds, sreten.

long time since I decided to focus on my own things ;)

but Im still glad not everyone does that :D

anyone is free to make whatever choice of drivers he like, or design
and perfectly fine fore other to comment, positivly or negatively
but also true that only he himself will be able make it work
noone can do that fore him
good to try and help out, sure thing
but in the end, the builder is the only one there, with the speaker, and the choice he made

well, to be honest, its also a long time since I stopped recommending designs I have never even heard
and I dont trust a design just because a certain name have been put on it
but thats a different matter

I think its wrong to say 'give up, you cant do this or that'
even if it might be true
well, might be ok to just say it
but repeating it becomes boring

as much as I think its equally wrong to promisse a good result
too many variables can still change the result
doesnt really take much to tilt a speaker
even if making all the 'right choices'

basicly, with sound, there is no possible way to ensure or promisse a good result
 
Ex-Moderator R.I.P.
Joined 2005
but maybe I look at it differently

to me its more like a priority list

and the builder decides the order of priorities
if the goal is building a speaker with his dad, or vice versa, then thats top priority

fore me, the priority list rules
the rest just follows that, good and bad

I really think an individual priority list would solve a lot of issues
and I would really like to see one, some day
 
basicly, with sound, there is no possible way to ensure or promisse a good result

Hi,

True. But there are lies, damn lies and statistics.

Statistically its possible the uniformed can come up with something good,
but not likely, especially when you can see glaring faults. I'm not promising
anything or ensuring anything other than what I say is to me common sense.

It is a learning process, and because a person can't see why something
will be wrong, doesn't change the fact that it is. Whilst what you say is
true to an extent, in my book it is the most common excuse to do things
"this" way because a person can't see what is wrong with the approach.

In my book its not a good attitude if building something for somebody
else. Would you drive a car built by someone who can't talk steering
geometry, stability and safety ? But says its got 4 wheels and goes
round corners ? The chances of it being good to drive ? And safe ?

Your right I've never heard the Tarkus, but that doesn't mean I cannot
appraise the design methodology and the apparent measured results.
Statistically what I'm saying is true, that I can promise / ensure.

From what I can see the chances of the OP ending up with a decent
3-way speaker from where this thread has got to is very near zero.

It is an opinion you can take or leave, I'm not forcing it on anyone.

rgds, Sreten.

From undefinitions FAQ's I linked to ages ago : The Repository->DIY->Proven Designs


Q: I don't get it. Every time I ask questions on DIY forums, people start suggesting that I build a proven design.

A: Well, have you ever built a speaker before?

Q: No, but why should that matter? Why can't I come up with my own design? I'm not an idiot.

A: Believe me, by suggesting you build a proven design, they are in no way insulting you, or calling you incompetent. In fact, when DIYers suggest this, it's because they actually want you to succeed.

Q: But where's the originality and creativity in just building someone else's design? The world needs people who can come up with new ideas.

A: I agree, and you're certainly entitled to "think outside the box." But what people are trying to help you with is the fact that designing a speaker is a far more complex task than it seems.

Q: It's just some drivers and a crossover which you can put in a box you like the looks of.

A: That's like saying that playing playing a piano concerto is just pressing a bunch of keys.

Q: Ouch. OK fine. So tell me why you think speaker design is "so complicated."

A: All right. I'm going to get a bit technical, so hang in there with me. The first thing most often considered in a design is the box and bass alignment. The two most common bass alignments for woofers are vented and sealed, but transmission lines, horns, infinite baffles, and dipoles can sometimes be an option as well. However, every woofer is usually best suited for only one of those applications. To figure out which one, we examine its electrical and mechanical strength (Qes and Qms), and come up with a figure called its Qts, which, depending on its value will tell the designer which bass alignment will work the best. Next, we need to know the equivalent air volume (Vas) of the driver. Using this number in conjunction with the Qts and the drivers' resonant frequency (Fs), we can start designing appropriate cabinets and alignments to get the proper bass response from a driver. Getting this step wrong means that the speaker's bass will either be way too weak, or overbearing and boomy... neither of which is fun to listen to.

Q: Can't you just get the best bass from a big woofer, and just sort of add stuffing or fiberglass until it sounds right?

A: It's a common misconception that a "huge" woofer is gonna make super-deep, visceral bass. Actually, the point of having a woofer with a lot of surface area is to increase its Sound Pressure Level. You'd be surprised; an 8" or 10" woofer can easily extend down to the bottom range of human hearing (20 Hz or lower), when it's in the properly tuned enclosure. Now please, pay attention. We still haven't gotten past step #1.

So, now that we have a general idea of our box volume, we can start considering how we want the cabinet to look and where we might want to arrange the drivers on the baffle. A lot of beginners (myself included, once upon a time) think you can just fenestrate drivers anywhere on the baffle, so long as it looks cool. Visions of rows of tweeters and midranges are sketched up--and often shot down by the DIY community.

Q: Yeah, probably because they can't handle any new ideas.

A: Don't kid yourself. We weren't born yesterday. A quick glance at home speakers from the 70s by companies like Marantz, Cerwin Vega, and Sansui shows that these sort of designs were once popular. But the reason manufacturers don't do this anymore is because through the magic of computer simulation, we've learned that "comb filtering" was a serious problem with most of these designs, and we have since optimized driver layouts so that speakers sound as uniform as possible from as wide a listening axis as possible.

Q: What does "comb filtering" sound like? Some sort of synthesizer effect or a guitar's wah-wah pedal?

A: No. What happens is that, depending on where you are listening to a speaker, relative to its location, you will experience "dead spots" where the treble and mids become quiet and difficult to distinguish. These dead spots, if you plotted them on a map of the room, would look neatly arranged, like the tines on a comb.

So, to avoid as much comb filtering as possible, driver layouts are often simulated on some sort of software first. The designer may still try a few different physical baffle arrangements, but the simulation saves him a lot of wood and backache from cutting out all the possible combos and trying them out.

Q: Great. So now we're down to the crossover. Just buy one that matches the crossover points of the drivers, right?

A: Well, you can, but the end speaker still probably won't sound all that impressive.

Q: Why not? I'll just use really good drivers.

A: I wish I could say that building a speaker is like building a computer--that you just match up bus speeds and buy the latest high-quality gear you can and you've got a rip-roaring machine. Speakers are a strange analog animal, though, and getting high-quality performance from them is a delicate balancing act, where components are calculated, simulated, listened to, tweaked, measured, evaluated, and revised again and again until you have a speaker that is truly "transient" and transparent. And you'll know when you hear a speaker that gets this right. You'll hear depth and detail in music that you've never heard before, but the speakers will be non-fatiguing to listen to, and will just sort of effortlessly reproduce the music without using any sort of trickery on their own part. It is that sound that makes people pay thousands and tens of thousands of dollars for HiFi speakers. And yes, it is definitely possible for a DIY speaker to achieve that sound, but not with an off-the-shelf crossover.

Q: So what's so special about a "custom designed" crossover?

A: As I was saying, speakers present all sorts of strange analog quirks. A crossover designer almost becomes one of those "plate spinners" from the circus, as he tries to successfully manage several factors at once: filter order needed to attenuate the unwanted frequencies (aka: "stopband"), combining Acoustic Phase introduced by the filter and physical offset of the drivers, flat frequency response, gently-tapered power response, and keeping electrical Impedance and Impedance Phase safe for a given amplifier. Hopefully, when all those factors are in-check, the speaker will be sounding pretty good--but often there's still more tweaking to go until the speaker starts sounding "right" in terms of effortlessly reproducing a recording. Then, to really complicate things, changing the value of one component in a filter often cascades through the rest of the components in that filter. For example, a designer might increase the value of a capacitor, which may suddenly require the use a smaller-value inductor and perhaps a larger-value resistor... and so on down the line.

Q: Really? But in my car, I can just install better speakers and a better amp and I can hear an improvement.

A: You raise a good point, but the two worlds just aren't the same. In car audio, they say you can "throw parts at a problem," but HiFi speakers require a whole different kind of effort. As I said before, crossover design is a delicate balancing act.

Q: So making a speaker sound good is hard to do?

A: Yes, it is. The crossover in a HiFi speaker is like its brain. And as a result, a badly-designed crossover can make extremely expensive Scandinavian drivers sound like crap. Alternatively, a well-designed crossover can take "cheap" Chinese drivers and make them sound excellent.

Now, as with many things in life, there is an easy way and a hard way to learn the ropes. Honestly, most DIYers' first design is something they'd rather not talk about. We were all stubborn once: insisting on using XYZ drivers, designing any enclosure we thought was cool, and using just about anything for a crossover. And the results were baaaad. Pretty much all of us have been humbled by our first "inventions," and it is this experience that people are trying to help you avoid when they suggest you build a proven design.

Q: Well, I'm going to be smart about it. I want to do all my research, and find out everything there is to know about crossovers before I build my speaker.

A: Be my guest! I don't mean to dissuade you, but you are probably headed for an "in over your head" feeling. But there's nothing wrong with doing your first design from scratch, following the rules and trying to use good design principles; but no matter how hard we try, maiden voyages are often pretty rough. Many veteran DIYers stubbornly built their first design from the ground up--and then consecutively built better and better projects. As with most skills, this is an iterative "practice makes perfect" approach, and it will work.

What's really cool about DIY audio, though, is that there is another road to success--a shortcut, if you will. You can follow someone else's plans (which have already been approved by many listeners as "high quality"), and turn a pile of parts into something that puts out a jaw-droppingly good sound.

Q: So you're saying "take someone else's plans, and focus on the woodworking?"

A: Yes. Good cabinet construction is one of those things that seems easy until you get your hands dirty and do it. There's a lot of work involved! Even if you're an extremely skilled woodworker, you'll still be doing a few inventive tricks you never had to before. The woodworking process is project in its own right. Put your initial energy into making a beautiful cabinet... and take the burden off yourself of somehow making the cabinet sound good once the woodworking is done.

You will feel no less proud at the completion of your project if you've used someone else's crossover; in fact, chances are you won't be able to wipe the smile of satisfaction off your face every time you look at and listen to your new speakers. By doing someone else's design, you're not "selling out" or "being too conventional." You're really just spending your money wisely: if it's going to cost you a few hundred dollars in parts, wouldn't you like to be guaranteed that the end result will sound really freakin' good?

But don't let me stop you from taking a risk. If you feel as though you are on a mission to design your own speaker from the ground up, then you should do it. This will all make sense to you after you finish.
 
Last edited:
Ex-Moderator R.I.P.
Joined 2005
Your right I've never heard the Tarkus, but that doesn't mean I cannot
appraise the design methodology and the apparent measured results.
Statistically what I'm saying is true, that I can promise / ensure.

ah, well, its a start ofcourse
and maybe even a good one
or at least better than 'nothing'

fore a first time builder, or 'designer', I think a 2way with 5" woofer is the best choise
almost anything else is bound to give trouble

but I also once met a guy who built big 3way speakers with random drivers he just found
and random premade crossovers, he also just found, or ripped out of other random speakers
and built ramdomly sized boxes, etc etc

he was very happy, and proud of his 'work'
well, I couldnt argue with that
and to be honest, I had heard worse, and better

the better one Im still working on, 15 years later :D
 
With your budget go for a two way not three, that means you can get far better quality drivers for the money. Unless you want sub frequencies with no sub 3 ways are useless, more money on s**t passive crossovers, more driver cost and more nasty crossover points to tackle.
 
Ex-Moderator R.I.P.
Joined 2005
but I also once met a guy who built big 3way speakers with random drivers he just found

he was very happy, and proud of his 'work'

the thing is, he was probably more happy with his speakers than I ever will be with mine
and my 'perfect' speakers would probably only result in a shrug from him
still puzzles me :D

one morale to conclude could be, never try to design hifi speakers
its impossible anyway :eek:

but to build disco speakers, that is possible :D

not even sure yet if its a joke or not :headshot:
 
next project...subs

carefully planned building-blocks can be a good choice, yes
especially when a budget is 'a bit limited'

but also means you have to live with some 'limitations', fore a while
but I guess 'limitations' are there anyway, no matter what

man, we are not making it easy, are we ?

Every speaker is a compromise, it's only a comprise because we don't have unlimited money (most of us!) but then again there will always be comprises whether it be size, distortion....
 
Hi,

Once you've got used to decent loudspeakers there is no way back.
This applies to everyone, hifi buff or not, and especially women.

Convincing yourself your half-baked creation is really good is an
entirely different matter, the same thing applies to flashy looking
but poor quality bought loudspeakers.

A few years ago I told a friend his speakers were letting his music
system down. He said they can't be, look at the quality of those
"cast frame" drivers. I took the grilles off which included the
flashy looking "cast frames" to reveal dirt cheap drivers.

Only then would he consider what I was saying was true. I found
him a pair of very boring looking used speakers, fitted with decent
drivers and a proper crossover, after a period of getting used to
them he admitted they were simply miles better, for £20.

(And they were, Dynatrons fitted with Peerless 8" and the famous
paper tweeter, proper c/o versus 3 way music centre tat with
just caps on the cheap "midrange" and dirt cheap tweeter.
Aside from the better response, distortion was vastly better.)

Every home made lash up I've ever heard has been poor, not
that they don't work, but they could have been miles better.

I might be a relatively (to some) old fart on my soapbox, but
I thought it worth saying. Long threads like this one always
wander off into pure mediocrity and a lash up mentality.

$400 spent a proper design is far better than spending
somewhat less but still an appreciable amount on what
will be almost certainly a lot worse. YMMV though ....

rgds, sreten.

I did contribute near the start of this thread. 14 / 15
pages in and car c/o's and poor driver combinations are
being discussed, its not a good sign to the experienced.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.