First 2way DIY speaker, where to start?

ELEFANTOMET

Member
2012-09-28 10:22 am
First some background.

I want to build my own 2way speakers, from scratch(except the drivers).

I want to get familiar with the whole process of the development. This means i'm not looking for a premade KIT to assemble.

My livingroom is ~21m2(3m X 7m) and the speakers are gonna be placed on 7m wall. My needs are speakers for normal music(primarily rock music), and movie watching. I'm not the biggest sound-enthusiast, but i care about quality and the primary goal for me is to get experience with DIY speakers.

I've been searching the internet for DIY projects and the process in DIY speakers. I understand that the cabinet design, crossover & speaker drivers have to "cooperate" to get the desired sound.

The Question
Where to start in this DIY process?

Im thinking the ~size of the speakers should be 20x40cm and i've been looking at theese drivers from scanspeak
ScanSpeak Revelator - FreQuence

My budget is around 5000 DKR ~ 900US$

I've got access to a workshop and all tools for cabinet assembling, and i got a soldering iron.

I dont have access to measuring tools, so i wanna design the speaker from simulations, and then when i get the $$$ to buy/access measuring tools, i can start the whole process of fine-tuning crossover from the measurements, if thats needed.

SO my big question is, how do i approach this?
Do i start with cabinet & drivers, and where can i find preferably free-ware software to simulate the appropriate cabinet for the drivers chosen? And what are the precise goals, when designing cabinet for speakerdrivers? And after that, i design the crossover?

Thanks for any help in advance!
 
It's good that you want to start with a 2-way because a 3-way will be a lot harder (as i'm about to find out)

Get some books, THIS is a good starter book.

In case you haven't seen it. Read the sticky at the top of the page.

If you like rock music, i'd personally stick to paper drivers. Other may disagree but they're wrong :)

When you have an idea of what drivers you want to use, start a thread here to make sure they are compatible with each other. i'd try to find some drivers with a smooth response (especially around the crossover point) to make crossover work easier.

Download a copy of WinISD and you may also find Jeff Bagbys software helpful.

Have fun!
 

tinitus

diyAudio Moderator R.I.P.
2005-11-24 1:47 am
I dont have access to measuring tools, so i wanna design the speaker from simulations, and then when i get the $$$ to buy/access measuring tools, i can start the whole process of fine-tuning crossover from the measurements, if thats needed.

SO my big question is, how do i approach this?
Do i start with cabinet & drivers, and where can i find preferably free-ware software to simulate the appropriate cabinet for the drivers chosen? And what are the precise goals, when designing cabinet for speakerdrivers? And after that, i design the crossover?

Thanks for any help in advance!

for drivers you have two options
either use good but cheap drivers
or use good drivers that can be 'reused'

another idea might be to use drivers used in a known good kit design
try to design your own with those drivers
you will then have a reference

and if you get tired of the whole thing :mad: you can always put together the original kit :D
 

ELEFANTOMET

Member
2012-09-28 10:22 am
Have you considered a kit? Lots of them out there.

All the box/xo work design work has been done.

btw, it would help to mention your location; availability & shipping is an issue.

I have considered a kit. But im more interested in designing my own from "scratch", than assembling a kit. Even if it could lead to failure and a worse performing loudspeaker.

If you mean my geographical location, im living in Denmark and i dont think availability shipping of components should be an issue.
 
Let me rain on your parade for just an instant. ;)

No, I wouldn't do that. It's awsome that you're getting into the hobby. :D

I would however toss you a few questions for you to consider.

What is the goal?
Do you want quick results or are you prepared to invest a sizeable amount of time?
Do you want easy och are you prepared to shed blood, sweat and tears?
How much are you prepared to spend? The details can eat up a lot more money than you think.

Depending on how you feel about this, there are several valid approaches you can choose.
Imho a clear goal is a very good help.
A well thought of approach is a great help and lessens the number of mistakes.

Good luck, and keep us posted. :)
 

ELEFANTOMET

Member
2012-09-28 10:22 am
for drivers you have two options
either use good but cheap drivers
or use good drivers that can be 'reused'

another idea might be to use drivers used in a known good kit design
try to design your own with those drivers
you will then have a reference

and if you get tired of the whole thing :mad: you can always put together the original kit :D

I'm not sure what you mean with drivers that can be reused. But i've been looking at several DIY projects, and thats one of the reasons i got led to scanspeak drivers.


Let me rain on your parade for just an instant. ;)

No, I wouldn't do that. It's awsome that you're getting into the hobby. :D

I would however toss you a few questions for you to consider.

What is the goal?
Do you want quick results or are you prepared to invest a sizeable amount of time?
Do you want easy och are you prepared to shed blood, sweat and tears?
How much are you prepared to spend? The details can eat up a lot more money than you think.

Depending on how you feel about this, there are several valid approaches you can choose.
Imho a clear goal is a very good help.
A well thought of approach is a great help and lessens the number of mistakes.

Good luck, and keep us posted. :)

I got the time, and got the will to shed blood, sweat and tears into the project. But again, i wanna do it the "quick way", since i'm not starting off cheap and building some low-fi speaker just to get the experience. I guess thats also an outcome of my economic budget being around ~5000-7000DKR = ~900-1200US$. At the same time i'm not just thinking this is a one-week project.




Back to the loudspeaker project.

Am i right if i should start picking drivers, that would work well together(with the Sticky in mind, regarding XO freq). And then choose cabinet design for better freq-characteristic. And finally design XO filter to get the final desired characteristic?

I have been looking at the scanspeak drivers. I know the 15W/8530K00 mid woofer have been used in a project by Troels Gravesen. So if im planning to use this, i want to find a tweeter to work well with this driver. Looking at tweeters from the Revelator series from scanspeak, and even other tweeters from scanspeak. I think the gap from the cone breakup of the midwoofer(for 15W/8530K00 its around 1500Hz) to two octave(2080Hz) above the tweeter resonanse freq is quite small( fs 520Hz R2904 Tweeter).

So what is the solution? Finding a woofer with a higher breakup freq? Or just choosing a XO freq between one octave above tweeter fs = 1040Hz, and 1500Hz? (there by disregarding this advice "A rule of thumb is to leave at least an octave (or two) between the crossover point and the cone breakup. ") ?
 

ScottG

Member
2003-02-04 12:23 am
US
SO my big question is, how do i approach this?
Do i start with cabinet & drivers, and where can i find preferably free-ware software to simulate the appropriate cabinet for the drivers chosen? And what are the precise goals, when designing cabinet for speakerdrivers? And after that, i design the crossover?

Thanks for any help in advance!


1st - forget about a speaker just yet.

2nd - STRONGLY consider doing a design with ONLY active dsp crossovers.. Ex.s: minidsp, behringer, or perhaps some software that requires a computer and multi-channel out sound card (..Soundeasy, Loudspeaker Frequency Allocator, foo_dsp_xover, Acourate, etc..).

3rd - see "2nd" above.

4th - look at proven designs from others for information on designs. Troels has a nice listing of designs. There are also some in PDF from Humblehomemade Hifi (in their "history" tab section - to "download" page), ZaphAudio, etc.. Also look for information on using the various software programs and hardware.

5th - there are 5 types of software/hardware items to consider: Box modeling, Baffle modeling, crossover modeling, freq./phase measurements, and T/S parameter measurements.

The easiest if not the most complete grouping would probably be:

Basta from Tolvan for the Box, Baffle, and crossover modeling;
HOLMImpulse for Freq./phase measurements;
Dayton Audio DATS for T/S parameters.

From Minidsp you'll need:
Umik-1 microphone (which needs no sound card);
At least their 2 in 4 out minidsp (which would only allow a 2-way). Consider their 4 in 10 out OR their 2 in 8 out kit (either of which would allow a 4-way);
The proper audio plug-in software for your application.

-that's about $350 US for all of the above if only using a 2-in 4-out minidsp.

You also need amplification for each driver - for a 2-way that would be 4 - channels of amplification (ie. 2 standard *stereo* amplifiers).

6th - use software and hardware extensively with items you have (or can purchase very cheaply.. perhaps used).

7th - NOW start thinking about a loudspeaker you might want to build. ;)
 
Last edited:
There are several different ways you can skin a cat.
Let's start with some simple issues.
I agree that a 2-way is by far the best choice for you, hell I'm gearing up for the build of a lifetime and it's still a 2-way.

Economics... what's your thoughts on the cabinet? Just enough, over the top or insanity at it's best?
A simple cabinet design for killing box resonanses could easily cost you 2000-3000Dkr. That's still going for a resonable cabinet. If you want to cheap out that's easy, just cut cost in half. If you want the "extra everything" expect the cash to flow...
Think finish as well, veneering is probably the cheapest and easiest rout for good looks?
It would allow you to use cheap Particle board (Low density fibre board/ Spånplatta).
Opting for paint and you are pretty much forced to go with mdf or plywood which is good materials as well but cost more. Expect the finish to be more costly as well with paint. (If you want a good paint job that is.)

The raw materials are freeking expensive in Scandinavia.
Heck, everything is expensive...
Where do you live btw? I live in Gothenburg.

Do you want floor standers or monitors on stands?
Where do you want your LF cut-off?
What sonic signature do you like?
Do you play LOUD or do you keep it within neighbour friendly levels? What's your expected max SPL to aim for?
What type of box are you curious about? Like I said above, there are more than one way to skin a cat and the result is not so much dependent on box type as on skill.
A good design will sound good regardless of cabinet type.

A lot of questions in this post and I haven't even touched the driver choice. Lol.
Imho you should start with the dream. Ask yourself what you want and envision that.
We can help you build that dream as soon as you're ready to share it with us. :)
 
To the helpful members across the globe.
I live in scandinavia as well and you can think of Sweden and Denmark as neighbours.
When you suggest equipment based on prices you are used to, try factoring in the actual exchange rate. $1 will be roughly 10Dkr when the taxes and duties are paid. Add shipping in the $30-50 range and you can see that "doing it right" might not be an option with a limited budget.

With 7000Dkr I would aim for good enough and leave perfect for later on.
A simple 2-way floorstander with good frequency respons is doable within the budget and passive filters will be much cheaper even if they aren't perfect.
When you're on a tight budget you can't afford perfect but very good is still possible. :)
 

Jay

Banned
2003-02-11 9:02 am
Jakarta
1st - forget about a speaker just yet.
7th - NOW start thinking about a loudspeaker you might want to build. ;)

Hehe. If you want an easy way that work, search for ARSXO (Audio Reality Series Crossover). It works better with smooth (roll-off) woofers and high power low Fs tweeters. But you will use your ears and some parts (coil unwinding and cap paralleling) :p

Seriously, when I build a speaker, I will search the net for every projects using the same driver. I will study other people works. Most probably I will prototype the best of them before deciding with my own. So, intentionally building a bad speaker for the sake of OWN design while better design is available, is a strange decision.

If, learning is the issue, there are many better ways to learn.

But usually the issue is that first builders are very optimistic and do not realize how bad the result gonna be :eek:
 
Lol, it just struck me...
Dare to think outside the box!

A ML-TQWT with a full-range driver would make it easy as pie.
Very musical, no crossover, manageble size and the focus is on the cabinet alone.
Software is cheap and time is your most valuable asset. :)

Your limit will be max SPL, possibly a slightly rounded sound on the top and not getting the lowest octave.
My own ML-TQWT hit 40Hz spot on and the only thing people have notised is a slightly laid back sound in the treable.
 
1st order, coil and cap design; sure, still have2-wasy that we made 20+ years ago, and they still sound very nice.
I would suggest that you find a generator, e.g. found an old HP tube gen that works beautifully for tuning and sweeping through frequencies (quickly). I would be dead in the water without it.
I mention quickly because some of the low cost transistor generators are just too slow to work with.
 

ngjockey

Member
2015-02-03 11:17 pm
My first question would be TM or MTM or TMM? Put another way, 8 ohm or 4 ohm? That depends largely on which receiver/amplifier you intend to drive them with.

Safe assumption is using either 17/18 cm (6.5") or 15's (5") woofers. Hard to go wrong with the SB as good value/performance and relatively easy and predictable to work with. Also safe to assume a vented design to get appreciable bass, unless you're also planning a sub or two.

If you do choose a basic TM arrangement, next question would be whether you want to go for a stepped/sloped baffle or vertical. The former offers symmetrical (ie: 3rd/3rd electrical) while the latter is more commonly asymmetrical ( ie: 2nd/3rd electrical), which, more likely, would require additional compensation for the low pass and cone breakup. For a 6.5", a baffle sloped about 15 degrees works with most woofers except those with large 3" voice coils (Morel, Dynaudio), with close woofer/tweeter spacing. For 5", about 10 degrees. Some adjustability with threaded feet.

With a 6.5", safer to go with a low crossover (below 2000 Hz), which leads to larger tweeters, like the SB29's.

Once these decisions are made, we can delve deeper.