ever design/build a speaker?

gmorris

Has anyone here ever designed or built a speaker?

I want to start a fun project to build my own speakers. Anyone who wants to help is more than welcome. Here is my idea:

start with a design similar to the RTi70's, a dual 6-1/2" woofer with a 1" tweeter in a nice big tower style cabinet. I was thinking of making it a 2.5 way design, meaning the bottom woofer will be in a seperate ported portion of the cabinet. The top woofer will be in it's own small sealed portion, and function as a midrange, with the tweeter up top, in a third sealed cabinet. I would like to set it up for Bi-wire, possibly with the bottom woofer a one circuit, and the top woofer & tweet combo the other circuit.

I would like to buy all drivers & pre-made crossovers from Parts Express.

As a Pro/E designer (3D CAD system) I will design the cabinet & have it cut on my companies CNC machines.

As I have no idea how to do this, I was going to buy that book "Loudspeaker Design Cookbook". I think its up to version 6 or something.

Tell me what you all think.

tryrrthg

I've gotten that book from the library before and it's pretty confusing for the most part. I didn't spend that much time with it but confused me enough...

My advice to you is, if you're going to make a speaker I'd start out with a 2 way. it's hard enough to get the 2 way crossover right let alone 2.5, or 3 way. Keep it simple! I'd also start out making a pair of bookshelf speakers. they'll cost you less to make the cabinets in case you screw up...

good luck and have fun. you might want to stop by here and ask them about some things. these guys know their stuff.

DIY HT forum

madmax

One of the best places to start is by checking out the www.partsexpress.com website and going to their forum. Many of those people know far more than we ever will!
madmax

TroyD

Russ might be a source of info for you. I'm sure he will chime in and help out as much as he can. He has built a few speakers. In fact, he's built some for me, that I've yet to get my hands on.

BDT

gmorris

Ah, yes, very good ideas from all, thank you.

As I started paging through my copy of the Parts Express catalog (I still like printed catalogs better than the internet, can I get an Amen on that?) I realized that it was too dang hard to figure out a 2.5 or 3 way system, so I'll go with the 2 way design.

I would still like to go with a bi-wire design though. That shouldn't be too difficult.

Thanks all & keep the ideas coming.

Greg

burdette

I have designed and built speakers and my sub.


"Here's my idea"...

Honestly, I read that and thought... he might as well be saying... "I've decided it might be fun to be a doctor, I thought I'd start out with a heart transplant." You're talking about bass reflex, acoustic suspension, and a complicated crossover all in not only ONE design, but your first attempt.

"I would like to buy all drivers & pre-made crossovers from Parts Express."

Pre-made crossovers aren't going to do it.


"As a Pro/E designer (3D CAD system) I will design the cabinet & have it cut on my companies CNC machines."

That'll be nice.

You need to start out WAY more simple that your current plan if you want to stay sane.

The LSDC by Dickason is a great book, but it isn't a great first book. I'd start out reading everything you can find on the Internet. I probably have some links I can post....

burdette

This might get you started. Don't just follow these links, but follow the links included on these pages. You need to read.. and read.. and then read some more. Ask questions, over and over, until you get it. I highly recommend buying and building a kit first just so you've done a project that *should be* successful.

Speaker and general:

http://www.loudspeakers101.com/

http://www.speakerbuilder.net/web_files/

http://www.partsexpress.com/resources/spkrbld.html

http://ldsg.snippets.org/index.php3

http://www.geocities.com/f4ier/speaker.htm

http://www.lalena.com/audio/

http://www.ishtek.com/spkrs.htm

http://members.rogers.com/viperoni/Audio/NewTitDVCarchive/index.html

Subs but also good info for woofers in general:

http://www.diysubwoofers.org/

http://gamma.nic.fi/~ajaaskel/

Crossovers:

http://www.trueaudio.com/st_xov_1.htm

http://www.lalena.com/audio/calculator/xover/

Starter kits:

http://www.speakerbuilder.net/web_files/Projects/DBP/dbp.htm

http://www.speakerbuilder.net/web_files/Projects/D2/d2main.htm

http://www.partsexpress.com/pe/showdetl.cfm?&User_ID=14450506&St=2926&St2=48920158&St3=75516739&DS_ID=3&Product_ID=16226&DID=7

gidrah

Start out small. There are a bunch of freeware and shareware programs out there that can help figure out box and port dimensions providing you know the T/S parameters of each driver. Some even have a database of common drivers.

If you want to bi-amp I suggest an outboard electronic crossover. These can be found on E-bay. Don't just seach "home electronics". Include the pro/musician side. You might also look at automotive XOs. You'd need a power supply but a trip to Radio Shack or internet search and you should be okay. This would also give you a bit more versatility of mating components and future endeavors.

I have no data to back this up, but the 70's hit low enough (without boom) and have a weight, that it makes me think they are of transmission line design. It's not just a big empty box, but smaller box extending down a length (via folds) of pipe to the port.

What kind of investment are you willing to make? What will you be driving them with? Is it for 2-ch or is it semi-dependant on 5.1?

I've built one pair of speakers. I copied somebody else. They were very inexpensive, quick, and sound great. But they were full-range drivers in PVC.

Good luck! Let us know and show pics.

gmorris

I have taken everyones advice into account, and thank you all for that. Only one thing I'm going to have to go my own way on here. Everyone says "Start Small". NO WAY! This is America!

I have decided to go with a 3 way design. I'm forgetting the bi-wire idea for now. I'm thinking of something like this: 10" subwoofer in ported lower portion; 6-1/2" woofer in sealed middle portion; tweeter in seal upper portion; and my design idea that could prove to be a real hum dinger, a fourth sealed area where the crossover circuit will be located. This crossover cabinet will have a removable access panel so I can upgrade the crossover later.

As far as investment, here is a quick breakdown. I called the local lumber yard, and a 4'x8' sheet of 3/4" thick MDF is about $9. I'll need 2 sheets. From the Parts Express catalog: Peerless 6-1/2" woofer - $35.50 (2 required); Peerless 10" subwoofer - $54.00 (2 req.). I haven't selected a tweeter yet, but they go pretty cheap. A 100 watt, 3 way crossover goes for $13.90 (2 req.). All other crap like polyfill stuffing, screws, caulk, wire, binding posts, I'm not tallying as they are minimal.

This will be just 2 channel, as these drivers are not video shielded. The shielded options are too expensive and I just don't like the options available.

Thanks again to everybody for pipping in.

Later,
Greg

jcaut

I built my first set of speakers nearly 16 years ago, and a few pairs since then. I'm still not "good", but it's a hobby that can be very addictive.

I was going to be another one to chime in saying "start small, two-way, proven design, blah, blah......", 'cause I wish I would have gone that route instead of the huge 15", 3-way monsters that I built first. About the best thing one could say about them was that they were loud. Which wasn't a bad thing, since I used them in a small mobile DJ business at the time....

But, anyway, if you've made up your mind already I won't try to discourage you. I will say that using an off-the-shelf crossover won't allow you to get the most out of the drivers you choose. Perhaps you could start with one and then tweak it later, as you mentioned. Also, if you're going to build a 3-way using that Peerless subwoofer as the woofer and the 6 1/2" driver as a midrange..... That 833429 won't make a very good midrange driver unless you plan on crossing it <1.5Khz. For the same $35.50, the 830377 shielded 5" would make a better midrange. Or even the Vifa 299-492 for $9.57, etc.

Here are a few links and some free programs to check out:
(edit: burdette may have already listed some of these)

FRD Consortium

Speaker Workshop -- Impressively capable but hard to learn speaker design and measurement software

LDSG


WinISD box modeling program

Jason

gmorris

Thank you Jason, that was just what I was looking for. Actual usable info, not just telling what not to do. I thought about starting small, but what fun is that? Even if this blows up in my face, so what? I'm out what, $250 dollars? I still will have 6 quality speakers that I can try another design with.

Everyone keeps saying that the off the shelf crossovers are not good. Why not? I don't really know much about crossovers, and I certainly don't know anything about designing them. So my only choice seems to be an off the shelf unit. So, somebody give your opinion of why they are no good & what you recommend.

jcaut

The basic problem with off-the-shelf crossovers lies in the fact that the functioning of the filters that make up the crossover are dependent on the impedence of the drivers that they're connected to, and the impendence curve of a given driver, in a given box is rather unique.

For instance, say you've got an ordinary 8 Ohm woofer that you want to cross over to a mid or tweeter at 2500 Hz. So you buy a simple, first order crossover network designed for an 8 Ohm driver, and connect this woofer to the lug marked for 2500 Hz. What that most likely does is place a .50 mH inductor in series with the woofer--- the textbook, first-order, value for an 8 Ohm, low-pass filter at 2500 Hz. And the attenuation would indeed be -6db/octave above 2500 Hz, as long as the impedence of the driver remains a steady 8 Ohms.

But speakers don't behave like a simple resistance. The impedence curve of a typical speaker shows a strong peak at the resonance frequency, a relatively flat area for maybe a couple of octaves, and then a pronounced rise as frequencies increase---due to voice-coil inductance. The nominal (rated) impedence of a driver is derived from the relatively flat (or gently rising) area of the impedence curve after the peak at the resonant frequency. There is considerable variation here: The impedence of most "8 Ohm" drivers will drop to around 6 Ohms at some point. But you see where this is leading? By 2500 Hz, the impedence of your "8 Ohm woofer may be 10.5 Ohms. Or 6 Ohms. Since the function of your filter varies with impedence, and impedence varies with frequency, there's little chance that your off-the-shelf network will produce the desired roll-off at the frequency marked.

Another point to consider is that the driver's own acoustic roll-off rate is not replaced by the roll-off from the crossover: It is simply modified by the filter. A steep, natural roll-off or peak in response that is near your crossover frequency would be likely to produce an unwanted peak or dip in the acoustic response. A custom-designed crossover might take be able to handle that, but an off-the-shelf network will not.

To use an off the shelf network successfully, you need to choose drivers that have a frequency response that extends and overlaps SMOOTHLY for a considerable distance either side of the crossover frequency. (In that respect, you may be better off building a 3-way) A Zobel (impedence compensation) network could be used (especially on the woofer) to flatten the impedence curve around the crossover point. Choose drivers with comparable sensitivity-- mids and tweeters will usually be a little higher than the woofer--- and plan on using variable L-pads to adjust the volume of the midrange and tweeter to match the woofer. Tweaking a crossover can get very complicated and expensive. Measurements help but they're no substitute for a good ear. That said, most "experts" agree that using the minimum number of crossover components that are necessary to produce the desired slopes will result in better sound.

Good luck!
Jason

burdette

Originally posted by gmorris
Everyone keeps saying that the off the shelf crossovers are not good. Why not? I don't really know much about crossovers, and I certainly don't know anything about designing them. So my only choice seems to be an off the shelf unit. So, somebody give your opinion of why they are no good & what you recommend.

People are telling you what they are because I think they crazily assume that perhaps you want this thing to sound good, at least a little bit. No offense intended, but you are WAY underestimating the process of speaker design. You haven't a clue what you're doing but you think you can just throw all this **** together and it'll sound good. Really, you can follow the plan you have.. just don't expect much. I didn't give you all those links to waste your time. I happen to have a EE, I happen to have been building speakers since I was 16, I've read everything I can find to read and own my own books... I thought you just might.. might.. want this thing to sound better than a Orion shelf system.

Question... why are you going to use a sealed tweeter compartment?

A crossover is an *integral* part of the entire system. It isn't just some after-thought. It can easily make or break the entire project. For most serious DIYers with the means, crossover design and tweeking are THE meat and potatoes of the design and success (or failure) of the project. Even if you choose a tweeter and M/W each rated at a nominal 8ohms, and you choose an off-the-shelf crossover intended for 8 ohm drivers, you're still highly unlikely to ... aw hell.. why bother. Knock yourself out. I was just kiddin' anyway.. just don't tell, or Polk will never sell another pair of speakers.

gmorris

burdette, my friend, I didn't mean to upset you so.

I readily admit to have no idea what I'm doing here. I've said that from the beginning. The first post I made said I wanted to start a "Fun Project".

The way I see it know, for around $250, I can build the cabinet & buy the drivers. Now, I'm starting to see why off the shelf crossovers are no good, thanks to Jason. I also just plain like to build **** out of wood. It's fun. So, I've spent a few hours here and designed a kick **** looking speaker cabinet with some inovations I've never seen in a speaker before. Thats all. So I'll have to figure out how to get some better crossovers.

You started out very helpful then started to peter off. Sorry my project is so over my head, but to me it would not be satisfying to build a pre-selected kit and simply screw all the components together.

Yes, I obviously want it to sound good. That is why I got on the forum in the first place. And thanks very much for listing all those sites to check out. That is what I was trying to gain by posting. I just haven't had a chance to look at them yet.

tryrrthg

if you like to build **** out of wood, why not build your own subwoofer instead? your results will be MUCH better and you'll enjoy the speaker much more than your current idea...

gmorris

tryrrthg,

not a bad idea. i haven't yet bought my subwoofer. i was planning on buying the Polk PSW 404.

tryrrthg

a well built DIY sub will absolutely CRUSH the 404.

There are many good designs already planned out for you at Adire audio based on the Tempest and Shiva subwoofer drivers:

Shiva Driver

Tempest

Shiva Designs

Tempest Designs

I believer burdette has built a sonosub with the Stryke AV12 driver that will CRUSH the 404. the Stryke drivers are supposed to be even better than the Shiva or Tempest drivers from Adire audio.

Accoustic Visions sells a lot of subwoofer related items.

read about DIY subwoofer designhere

here are a bunch of examples of sonosubs (the cheapest DIY sub route) Sonotube Subwoofers

jcaut

I would second (or third) the DIY sub idea. Use a free box-design program, build a good strong cabinet (or Sonotube) and put the appropriate high quality woofer in it. Plate amps are easy and not much $$. You can build a giant-killer DIY sub for much less than buying a finished one.

burdette

Originally posted by gmorris
burdette, my friend, I didn't mean to upset you so.
So I'll have to figure out how to get some better crossovers.

All right.... gotta get this figured out. This is three times in a couple of weeks that someone thought I had my panties in a bundle. I do believe that my stupendously funny but dry sarcastic wit doesn't make the transition in all instances to the written word. Sorry gmorris... didn't mean to come off as all in a serious tizzy. I just don't rattle as easy as I must appear to rattle.

It really is not a problem. I honestly want to help you have a great DIY experience the first time out.. and I honestly thought you were on the wrong track to make that happen.


I think the sub idea is a GREAT idea for a first project. The upside is that you can play and play with different alignments (box size, port, etc) and see the differences using software. If you have access to the wood cutting facility .. you may want to go ahead and build a wooden box rather than a tube. But, if you have the ability to cut circles with great precision, I think you'd really enjoy building the tube sub. I would have probably built a box sub if I'd had the tools to cut with the precision necessary.

The primary downside of a sub is that it won't give you experience with crossover design, as you'll most likely buy an amp with an electronic crossover. A passive subwoofer crossover, with the chokes and caps required, could easily end up costing.. and weighing.. more than the driver... and you'd have no power.

The reason I recommended you read all that information is so that you'd learn all the different perspectives, opinions, philosophies, etc. Crossover design is .. honestly.. a REALLY complicated thing. I'm a EE, and I *know* I'd use all the forums I could find and post question after question if I were trying to design a crossover for anything but a basic 2-way.


Again.. no problem with anything, not on my end. There are a lot of great resources out there to help with whatever you decide to do.. and there is now a LOT of DIY sub experience right here on this forum. Regardless of what you decide to do, let me know if you want any help. More than happy... now.. if I can only get these panties unfurled....

burdette

Double post.. sorry.

madmax

Did you know:
*If you place the tweeter wrong it will sound terrible.
*Sometimes you wire the tweeter in reverse polarity.
*A system can sound good but have huge impedance problems.
*There are several configurations each having its own benifits.
*Bass rolls off differently in different rooms.
*Size of the front panel affects soundstage and dispersion.
*The cheapest speakers in the right enclosure sounds better than the most expensive ones in the wrong enclosure.
*You have to design your crossover around the drivers or pick the drivers around the crossovers
*Passive radiators are cool and not too hard to calculate weights after someone shows you how.
*Enough area in the port is just as important as getting it the right length.

I think these are some of the things burdette is probably thinking of. You can think about all the designs you want but don't waste your time cutting wood and making boxes until you read A LOT on the subject. Each one of the above items plus a hundred more will kill your project if you start too soon. Then again, if you did as I did, you would keep reading and never get anything (good or bad) accomplished. That isn't good either. Pick a cheap (but good) woofer and tweeter, read all the specs and design using the internet resources. Do a simple design with a sealed box and a simple crossover using just a capacitor. As soon as you get it done and get your jollies from listening to it for a couple of hours go the next step and figure out how to add an inductor for a steeper xover. (leave the wires running out of the box and play with the xover outside the box). Remember, there are thousand dollar systems designed this way. Keep it fun!

madmax

jcaut

PE has a buyout, 5.25" woofer that's supposed to be a Definitive Tech driver, currently selling for $5.97. Part number 299-706. This is a really nice little driver for $6: Cast magnesium frame, rubber surround, etc. Not shielded, though, and 4 Ohm nominal.

I think I've got about 16 of 'em in my closet, in my big box 'o speaker parts.

How's this for an inexpensive first project: One of these in a .25 cu.ft. vented box tuned to 55 Hz. Predicted F3 of about 60Hz. Combine with an inexpensive dome tweeter like Tang Band 28-847SA for $15. Simple first order crossover at about 2500Hz. Should produce a decent little bookshelf speaker for under $100 in parts.

gmorris

any now, my forum friends, for something really different.

I have in my possesion 2 JL Audio 12W-1 12" subwoofers. (auto subs) They used to be in an old car of mine. I am lucky in that they are both 8 ohms. So, here is my idea. (everyone take a deep breath, go get a snack)

Get on the JL Audio website, find all the Theile/Small parameters (spelling?), then go about designing a cabinet for the subs. I just looked up in the Parts Express catalog the subwoofer plate amp. $128 for a 180 watt RMS amp with features out the arse!! IF it is possible to design a cabinet for this sub that will be suitable for home audio, I could buy 2 amps, build 2 cabinets, and have a sweet sub system.

Now for your part, forum buddies. Do you think it will be possible to design a cabinet for these subs? Again, they are 8 ohms, so compatibility with the plate amp should not pose a problem. And I'm pretty sure 180 watts should be enough to power them.

Now, about the tower design. I think I'll put that on hold until I can read all the info that everyone has provided links to. Thanks again for all that. I agree that the sub project will be alot easier & alot less likely to end up a steaming ****.


Late,
Greg

jcaut

Most of the JL Audio subs seem to work best in sealed boxes. If you download that WinISD program, they're probably already in the database. Looks like they might like a rather large box.

Also, I haven't checked the price, but I might go for a 250 watt plate amp instead of the 180. Really just depends on what your output goals are.

BTW, Dr. Spec would be an excellent person to chat with regarding plans for a sub.

madmax

Download the freeware program called WinISD off of this website. (not the "pro" one). It will give you everything you need. Box size, port if you want one, different configurations etc.


http://www.linearteam.dk/


Start with "new" then pick your driver. The parameters are already in it. You can probably take it from there. It calls for an area of about 5 cubic feet. If you mount the drivers face to face you can cut the area in half. It all depends on your preference. You can set the QTC of the box as well. The program uses a default of .71 which is good. Go with .63 and the sub goes lower but can be over extended easier or with .75 and the sub is louder but doesn't go as low. You can change the size of the box and the port frequency and look at the graph to see what the frequency response looks like.

That should keep you busy for awhile!
madmax

gmorris

madmax,

what is meant by QTC? I am going to try that software out that you refer to. But, you provide 3 values for QTC & explanations as to what changing the value will do. What is that?


Also, to everyone. What does everyone feel the best choice will be for home use, ported or sealed? I understand that depends largely upon driver parameters, but from general experience, what do you all think?

madmax

QTC is somewhat a loudness to how low of a frequency you want to go to ratio. (I know thats wrong but it kinda explains it). The easy answer is that if you want really low bass go closer to a .63 number and if you want a really loud 40 hz and above thump the go towards .75. The smaller the enclosure the higher the number. If you have a VERY small room you can get away with a higher number because the room will have its own gain. If your room is really big then you may want to port the speaker because there will be very little room gain. Don't decide what you want yet. First of all figure out what your design parameters are. (room size, what kind of sound you like, how loud, how low, for what purpose). You are now in control. You can design it to do anything you want!
madmax

tryrrthg

From my understanding QTC is how "tight" you want your bass. .5 is the tightest. I read in a lot of places that if you want a sub for HT and music that you should shoot for a QTC of .7 or so. A QTC of .7 will give you the extra "slam" or hump in the midbass, but also cause the sub to roll off sooner than a lower QTC sub. A sub closer to .5 is more of a music thing but you'll also get the best extenstion from a low QTC sealed sub.

I'd also recomend selling your JL subs and getting a sub made for home use. Most car subs won't do as well in the home...

just my .02...

burdette

Not all drivers are equally suited to bass reflex or acoustic suspension designs, but you can force the issue.

Qtc is relevant only to an acoustic suspension design.

Qtc of 0.707 is the flatest response with minimal rolloff and *the* lowest F3 for the driver. Moving off 0.707 either higher or lower will increase the F3. You cannot extend to a lower F3 than you get at Qtc = 0.707, and you do not get extra "slam" or a hump in the response at that point.

Designing to a higher Qtc will allow you to use a lower internal volume, but increase the F3 and give an artifical boost above that range. Most inexpensive subs are designed this way, not only to save on box size but to give that artifical boost in the upper bass region, which is why most people often find a "true" sub lacking. Boomy subs have a Qtc that is too high. Many people like higher Qtc subs for HT because it artifically increases *apparent* response, but it is an artificial boost from flat. Some people prefer a lower Qtc sub for music because it does away with the boom.

A properly designed, quality sub will perform both jobs well, but you can design in whatever 'effects' you'd like.

gmorris

WOW!! Burdette, Madmax, & Tryrrthg you guys really know your ****. (not that I doubted you, I just had no idea what you were talking about)

I downloaded the WinISD program that Madmax told me about last night. I've been playing with it tonight since I got home & it is sweet. I now understand all that stuff you've been telling me about Qtc. It all makes sense when you can see it graphed out.

So here is where I'm at:

I've designed a box with 5-1/4 cubic feet of internal volume using 3/4" MDF. This yields acceptable, albeit a little large, external box dimesions. I can live with this large box as this is my first project & I am utilizing drivers I currently have.

This volume yields a Qtc of .73.
Also, I'm not sure what F3 is, but I think (from the explanation in the software help) that F3 is the frequency where the trace on the graph intersects the -3 dB line. (is this right?) Any way, with my design, the trace on the graph hits -3 dB at about 34.5 Hz.

So, guys, how does this sound? Aside from a rather large box, what criticisms do you have? I think from everyones explanation of what Qtc is, I would prefer it a little lower, but I can't seem to do that without a box the size of a VW microbus.

jcaut

WinISD's a pretty cool little free program for box design. One of the neatest features is the ability to click on the box (on the "BOX" tab) and move the mouse around to "draw" the curve you like. You can lock either the box volume or the tuning frequency (for vented) by clicking one or the other to turn them from blue to black. You may have already figured that out. There's also a signal generator that's fun to play with, if not particularly useful.

Regarding F3: As you figured out, that's the frequency where the response is down 3db below the reference level. It's generally considered the lower limit for meaningful bass output, as the output of a sealed box (with a Qtc ~.7) drops off at about 12db per octave below F3. Vented box alignments generally drop off much quicker below F3, usually around -24db per octave.

If you can live with the large box, I think your sub idea sounds like a good one. You can increase the apparent volume of a box by adding Dacron pillow stuffing or fiberglass insulation. I honestly don't know if the WinISD program's calculations for a sealed box assume any stuffing will be used or not. You should use some, regardless, to help absorb standing waves inside the box. If you have MS Excel, and want a little more involved design tool you can download the Unibox spreadsheet from the link I provided above to FRD Consortium. It's a little harder to figure out, though. If you want to post the woofer specs (or tell me the exact model you've got), I'll run them on there and send you the information.

A sub will generally play a little lower in a room (unless it's a really big room) than the predicted response shows. A sealed box sub with an F3 of 34Hz will give you some pretty impressive bass, provided you've got the amp power and the cone excursion to handle it.

Jason