DIY monitors question

[TomasJ]

Hello there,

I am pretty much an amateur when it comes to electronic and speakers. I would like to try to build my own studio monitors. First of all I am not expecting to create replacement for any $500+ monitors (or even cheaper ones). I am just trying to get a solid sounding system* as part of my hobby.

So I read few articles abut DIY studio monitors and also checked some schemes of original studio monitors. I came up with this solution:

Woofer:
80 W/160 W (RMS/Peak)
89 dB
4 Ohm
45 - 3 000 Hz
8''

Tweeter:
50 W/100 W (RMS/Peak)
91 dB
8 Ohm
1 200 - 20 000 Hz
1''

Crossover:
Linkwitz-Riley, Second Order (2 000 Hz)
Tweeter part
L1 = 1.27 mH
C1 = 4.98 uF
Woofer part
L2 = 0.64 mH
C2 = 9.95 uF

My question is. Am I doing anything wrong? Or am I on a good way to build something that would actually sound anything like studio monitors?

*solid sounding system means that anybody who listen to it would say: "Not bad!" 

 

[audiomixer]

to be honest, none of the specs listed in your post relate to speaker quality / suitability at all.... sorry.

if you want to build a pair of speakers - fine. if you want to be successful at the first attempt build a kit. if you want to fail and learn from your mistakes go ahead.

a few hints:
woofers live in a ecosystem made from the box (volume, stiffness, damping, size ratio), the tuning (closed box, bass reflex loaded, other options) their thiele and small parameters, the electronics (active, passive) to name a few. the usable range is dependant of the driver itself, the type of cone, the make of the cone, the surround and many other criteria. frequency response will be non flat, so a mathematical approach to x-over design is usually not fruitful. impedance is non flat too, by the way.
tweeters have a usable range starting well above its resonance frequency. power handling is usually not so critical. choosing a crossover frequency is usually a trade off between both drivers.

so to get the best you could get a kit like this, build it and then see what modifications do... learning by doing. having a possibility to measure frequency response and impedance would be nice too, even if that is a field in its own right.... you might want to start with a cheap but decent mic and ARTA.  this seems a decent kit for starters:
http://www.parts-express.com/dayton-audio-br-1-6-1-2-2-way-bookshelf-monitor-speaker-kit-pair--300-640#lblProductDetails

don't forget to buy a few capacitors / inductors as well. caps can be paralleled to get bigger values, inductor don't scale that easily.  3w to 5w resistors come in handy as well, values around 1 to 22 ohms are used most.

- michael

 

[TomasJ]

Yea, the cabinet is another story, I know. My question is just about theoretical correctness of picked components. As I am a beginner in this field, the speakers are cheap ones from local store (Slovakia). Please tell me what other parameters are important when picking speakers (please don't say brand or price...these are obvious). So in theoretical terms this speakers + crossover can create a studio monitor right?

Tomas.

 

[pvision]

Brand and price are irrelevant except to help you fit within a budget

If I wanted to design a decent speaker I'd set a design target,  choose drive units based on their Thiele-Small parameters, model the performance of the speaker and enclosure using software, design a crossover that suited the units and that gave the required frequency response curve & unit sensitivity matching, then build & measure a prototype

If I wanted to build something and learn along the way I'd buy a kit of parts from someone like Wilmslow Audio in the UK, or follow designs from one of the online speaker design websites or a manufacturer like SEAS

To be honest, I think building something from random drive units will be a waste of time & money

A few links
http://www.wilmslow-audio.co.uk
http://www.troelsgravesen.dk/Diy_Loudspeaker_Projects.htm
http://www.humblehomemadehifi.com
http://www.lautsprechershop.de
http://www.linkwitzlab.com/orion-photos.htm
http://www.seas.no/index.php?option=com_content&view=category&id=66&Itemid=365

Nick Froome

 

[TomasJ]

These are not random. I've was looking especially for woofer with the highest frequency of 3kHz and tweeter with the lowest frequency of 1.5kHz as some of most popular studio monitors used to have this kind of relationship between woofer and tweeter (e.g. Yamaha NS-10). The same I did with crossover. NS-10's use 2kHz crossover, little bit differently set but this is the type of crossover I get calculated online. So my question is just whether I am missing anything or not. As far as T/S parameters, what other parameters are important in choice of drivers?

Tomas.

 

[TheJames]

You want a "solid" sounding speaker system.  You haven't stated budget, but are implying that you don't want to spend much on them.  You haven't specified what drivers you're planning on using, just some vague specs.  Nor have you specified target performance goals.

I would specify some goals, and then look to see how they can be reached.  A pair of speakers, built for $400, that reach -3db at 50hz and will reach SPL's of 110db.

You'll find that to be a fairly tough target to reach.  Don't forget to factor in amplifiers too.  What's worse is that if you start trying to do recording/mix work on these monitors and the results aren't turning out like you expect, you'll always be questioning your unknown quantity, DIY speakers.

Take a look here:

http://www.seas.no/index.php?option=com_content&view=category&id=66&Itemid=250

Those are some pre-defined designs that are a known quantity from a quality manufacturer of speakers.  Look over what they have and see if something suits you.  I would suggest building one of those before designing your own.

With all that said....With the price of some of yester-years actual studio monitors....I don't know if it's worth it.  A great pair of Tannoy's could be had for cheap.  Or even some modern powered systems like Events or KRK's.

 

[audiomixer]

If you give us a hint of what you can get in your country it might be easier to get you into the right direction. as far as I understand you are mainly interested into DIY, not second hand. for the learning experience this makes sense. but keep in mind: you will need some kind of reference, otherwise tuning is very difficult and prone to failure due to  listener preference, which is usually non flat,

a reference could be a good set of quality speakers, a few well known recordings, a measurement system.

hope this helps,

- michael

 

[TomasJ]

If you give us a hint of what you can get in your country it might be easier to get you into the right direction. as far as I understand you are mainly interested into DIY, not second hand. for the learning experience this makes sense. but keep in mind: you will need some kind of reference, otherwise tuning is very difficult and prone to failure due to  listener preference, which is usually non flat,

a reference could be a good set of quality speakers, a few well known recordings, a measurement system.

hope this helps,

- michael

This sounds good. Here's the web page I was picking drives from. It's possible to change language so you'll understnad:
woofers:
http://www.bsacoustic.com/woofers/
tweeters:
http://www.bsacoustic.com/tweeters/

if you can help me with this I'll be really glad. Maybe you can chose two (or more) different combinations and tell me why this and not other. Thanks! It's kind of you!

Tomas

 

[audiomixer]

for the sake of self education do the following: try to find an online vented / closed box calculator that uses thiele small parameters. identify the interesting woofers, plug the number in and check with your expectations. keep in mid, some drivers like closed box instead of vented, some might not give meaningful results in a vented box or vise versa. so check for both.

look at the frequency response graphs and try to identify resonances in the cross over region. again, keep in mind that manufacturers might try to obfuscate their weaknesses....

come back with three candidates....

- Michael

 

[TheJames]

I did a quick look and picked a potential woofer as a starting point based on some of the criteria I posted above and added some design considerations....

When you're looking at individual drivers...you want to check the response curve for any odd response characteristics.  Take for example this driver:

http://www.bsacoustic.com/woofers/200510/

That's not a bad driver, but it has an odd peak at about 3.5k.  If you were to use that driver, you'd probably need to use a first order crossover around 1-1.5k and add a filter to counteract that peak at 3.5k in the crossover of the woofer.  Ideally you want it to taper off into the background without announcing itself at some point beyond the crossover frequency.

It should be able to hit the 110db SPL at the end of it's power handling.  One parameter that isn't included in the specs is "Xmax" which tells you the cone excursion.  The resonant frequency of 46hz is something to note as well as the various Q parameters.  If you don't have modeling software...You'll definitely want it to allow you to easily compare enclosure alignments.  There's quite a bit of math involved, and modeling software makes short work of it.

This particular driver brings up a problem with integrating a tweeter though as you'll need to find one that goes pretty low.  I didn't find a tweeter that I really liked for matching up with the above woofer.  The closest was this guy:

http://www.bsacoustic.com/tweeters/ar-26tg55/

If you were to use this tweeter, you'd want to use a rather aggressive third-order crossover design perhaps slightly above the crossover point of the woofer.  You might like that peak at 10-11k, but it would most likely give you a false sense of "air" in the speaker response. 

There are potentially dozens of other combinations...But that's a basic example of where you could start with a design.  I don't think I would keep working with those drivers unless I knew that they sounded particularly good.  Most likely I would look at a different woofer.

The first challenge in a speaker design (after determining the design type, 2-way, 3-way, MTM, horn, etc) is finding the right low-frequency component to deliver your goals.  Part of that is looking at how the woofer integrates with the overall design (my issue with the peak in the woofer above) and any design requirements the woofer imposes on the other components of the system (in my example, the lower frequency point of the tweeter).

Everything is a compromise.  You want this?  You give up that.  Repeat.

 

[TheJames]

You might want to look over the projects on this site:

http://www.zaphaudio.com/

The author/designer steps through many of the design goals and challenges in the design of the speakers.  It might assist you in a virtual "looking over the shoulder" kind of way.

 

[alexc]

Some great tips and methodical approaches for beginners  described here    

Thanks for the thoughtful posts and links most useful. 

I too am a speaker build beginner.

My interests lie more in the high-sensitivity, low power, full range boxes for studio big monitoring  (12s)

But no doubt, I can follow some of the advice given here and get my understanding of the basics to a decent level.

Cheers
AC

 

[TomasJ]

@audiomixer

That's a great idea but I am not sure I am getting this right. I've found few online calculators but each of them is calculating only with one speaker in a box. How should I calculate box for 2 drivers - woofer/tweeter? I am sure I don't understand this right so what am I getting wrong?

@TheJames

Thanks that's really helpful. I will try to put this drivers into the calculator as well to check their behaviour. And yes I've check this site already but it's not easy to get this drivers here, and the shipping is way too expensive. So I have to work with what's available here in my country.

 

[audiomixer]

you only need to align the woofer LF response with a box (vented, sealed system), so the tweeter comes in later. you will think of the tweeter later on....

- michael

 

[TomasJ]

So I tested two different woofers and I am not sure what's wrong, but one gives me extremely big value for box volume and other gives me also big but not unreal big.

Calculator 1: http://www.ajdesigner.com/subwooferonline/sealedonline.php

First driver:

http://www.bsacoustic.com/woofers/i8-80/
http://postimg.org/image/oulmworoh

Second driver:

http://www.bsacoustic.com/woofers/200510/
http://postimg.org/image/85k70rv35

Calculator 2: http://www.micka.de/org/en/

First driver:
I skipped "your own box" option. And there's also no vented box plot as it says that Qts must be between 0.2-0.65 even the vaule from drivers webpage is 0.7
http://postimg.org/image/4bmspj33b
http://postimg.org/image/lq70xt087

Second driver:
Also gave me same 'error' even Qts is now 0.47
http://postimg.org/image/wzr8oe1cd
http://postimg.org/image/4y8lwy6od

 

[audiomixer]

some woofers are designed to be installed in a vented box, some for closed (sealed) box. you will get unsatisfactory results if you mix that up. as a general rule of thumb you could say if Qts is below 0.5 you will tend to try a vented box first, above 0.6 your better off with a sealed design. I gave your second option a run here:
http://www.mh-audio.nl/reflexboxcalculator.asp I just picked one....

this gives more or less realistic results. try a bunch of speakers and note Fs, Qms, Qe, Qts, Vas, SD/diameter (SD = r^2 * PI).. then run through the caclulator and note the box volume, F-3, Fb....
you can rerun your drivers with a sealed box design if you are keen on that. this will give you a idea on which speaker you might want to use. a hint: volumes above 60 to 80 liters are quite unmanageable in real life, need lots of bracing and are probably nothing for you....

using a spreadsheet and double checking the individual values entered a helpful. do not assess mid / high frequency indications of the spec sheet right now, find a suitable candidate first.

good luck,

michael

 

[TomasJ]

Ok so there's absolute winner (the one TheJames suggested):
http://www.bsacoustic.com/woofers/200510/

Here's its frequency response in sealed box with appropriate dimensions:
http://postimg.org/image/4q3c3rawj
W=183mm
H=249.5mm
D=164mm
V=10.489l

I was trying to get similar frequency response to yamaha's ns10:
http://dt7v1i9vyp3mf.cloudfront.net/styles/news_large/s3/imagelibrary/y/yamahans10fig1.jpg?b2_vAaRyZSK1aICJYKYroeM7xi9qOg4q=&itok=tJTboLOh

I set the volume of internal driver and parts to 0.4l. Am I close to reality? Or use to be this number much different?

And I am also not sure about performance parameters:
Total Q=0.950
Frequency at peak db = 140.11Hz
peak = 0.964db
f3=75.39Hz (NS10's have f3 at about 70Hz so I hope I am not far from standards)

I have two more questions before I move to tweeter.
1.There use to be split cabinet for woofer and tweeter in some monitors. Am I right? So are my calculations correct then? Shouldn't I calculate with for example 1/2 volume of cabinet for woofer part?
2. Is there any calculator that calculate closed but filled cabinet's behaviour? I am asking just because some monitors have filled cabinets.

Thanks guys!

 

[audiomixer]

OK, you are after a NS10 'replacement', I didn't get that until now - smile.

you will see that with sealed boxes the actual LF response is less prone to drastic change with different volume as with vented enclosures so you have some play there. usually you account for damping by making the actual volume slightly smaller, let's say around 20%, because the damping acts as if you were placing the driver in a larger cabinet.  make two or three different cabinets, that'll teach you ;-). anyway, +/- 1 liter is probably not an issue and you can / have to finetune the damping anyway.

this driver also nicely fits into a vented box (but in a noticeably larger one) so you can test that too if you want. you'll get extended LF response but different behavior in the time domain so that might be worth trying. buy four drivers anyway, so you can select the two best ones for the final project.... (if you can fund it of course...)

you can look up the formula for volumes (basically a cylinder for the magnet and a cone for the cone) and get a rough estimate of the displaced volume that the driver occupies. doing that in decimetres (1dm = 10cm) gives you liters (1dm x 1dm x 1dm = 1l  is the same as 10cm  x 10cm x 10cm )

tweeters that are thoroughly sealed can live in the same volume, but its better practice to use a separate internal volume for the tweeter. take this volume into account when calculating the internal volume for the LF driver. the crossover or a compartment for an amp module takes some space too.

there are many different  characteristics defining a driver, but within your limitations this might be an alright choice to see how far you can get in terms of rebuilding this classic. the NS10 Woofer has / had a paper cone, your driver of choice has a PP membrane, not that comparable but hey, that's what you can buy!


by the way, I would try to allign the box to a Qt of 0.5 (Bessel) for a little slower rolloff.

cheers,

Michael

 

[TomasJ]

Ok just to be sure I understand right. Your english is little bit complex for me.

So, the damping makes driver to behave similar way as putting it in bigger cabinet right?

Now the part I am really not sure about is tweeter part. Basically I don't need to calculate any cabinet volume for tweeter performance as I did for woofer, yea? But what I should try, is to separate woofer volume from tweeter volume. That makes smaller volume for woofer doesn't it? Should the separation plate/board have a whole in the middle? I think I saw something like that somewhere on the internet. I mean the volume of cabinet wasn't really separated but there was a separation plate/board with big round whole in it. Is that what I should try?

As far as NS-10 example, I am not really trying to do a replacement. They are just so popular it's very easy to find any parameters about them. I haven't found any other monitors with so many 'revealed secrets' on the internet, if you know what I mean. So that's why I consider them as guild line when working on DIY monitors.

Thanks for help!
Tomas

 

[audiomixer]

Ok just to be sure I understand right. Your english is little bit complex for me.
* sorry
So, the damping makes driver to behave similar way as putting it in bigger cabinet right?
* yes, you can put it that way.

Now the part I am really not sure about is tweeter part. Basically I don't need to calculate any cabinet volume for tweeter performance as I did for woofer, yea? 
* exactly. many tweeter have an internal volume behind the dome. fancy ones have a hole in the magnet assembly and a plastic cover to create a small volume.
But what I should try, is to separate woofer volume from tweeter volume. That makes smaller volume for woofer doesn't it?
* yes. calculate it and bake the enclosure bigger by the same amount. I would still try to make a separate volume for the tweeter.
Should the separation plate/board have a whole in the middle? I think I saw something like that somewhere on the internet. I mean the volume of cabinet wasn't really separated but there was a separation plate/board with big round whole in it. Is that what I should try?
* this would be bracing - this makes the cabinet stiffer and reduces or changes the resonance of the cabinet. if you knock on a speaker it has a distinct 'tone'. the less pronounced the better.

As far as NS-10 example, I am not really trying to do a replacement. They are just so popular it's very easy to find any parameters about them. I haven't found any other monitors with so many 'revealed secrets' on the internet, if you know what I mean. So that's why I consider them as guild line when working on DIY monitors.
*thats ok, you decide what you like. just be aware that a vented enclosure is a option with that driver too.

Thanks for help!
Tomas