diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Multi-Way (http://www.diyaudio.com/forums/multi-way/)
-   -   Cheap 2-way Speakers - Design advice (http://www.diyaudio.com/forums/multi-way/208361-cheap-2-way-speakers-design-advice.html)

jacobthellamer 7th March 2012 11:17 PM

Cheap 2-way Speakers - Design advice
 
Hi All,

I have picked up some cheap drivers last weekend and am wanting to design a cabinet for them. These are just a little fun side project, nothing too serious.

I wired the drivers in parallel and they sounded ok straight on, as both the tweeter and woofer are around the same sensitivity so neither one overpowered the other.

I am not sure what crossover frequency I should aim for and what cabinet type I should go for. My first set of speakers are actively crossed 3-ways so I am new to passive crossovers.

I would say my crossover budget is around $50

I was thinking about building a 60L vented enclosure. I am unsure on the best dimensions for this and where on cabinet I should place the drivers, I was thinking that the tweeter and driver should be as close to each other as possible?

16mm MDF with lots of bracing should be fine for construction?

http://www.surplustronics.co.nz/shop/product-SP2202.html
http://www.surplustronics.co.nz/shop/product-902.376.html

What information I could find on the driver:
Re - 7
Fs - 46
Qms - 2.82
Qes - 0.63
Qts - 0.52
Sd(CM2) - 211
Vas(L) - 33.5
Cms(uM/N) - 531
Mms(g) - 22
Rms B x L - 8.5
SPL_1W/1m2 - 89
Prms(W) - 60
Pmax(W) - 120
Xmax - 1.5


Any ideas and suggestions would be much appreciated.


Cheers,
Jake

PeteMcK 7th March 2012 11:47 PM

1 Attachment(s)
60L tuned to the mid 30s is about right, yes the tweeter should be close as possible to the woofer. Dimensions - Keep the tweeter about ear height, & internal volume constant, then make it any shape you like - wider is better for low baffle step (my preference), narrower is purportedly better for imaging.

Without a freq response graph for the woofer, it's hard to choose a crossover freq. I'd use the recommended 3KHz for the tweeter as a starting point. Because the woofer is a dual cone, it's response is probably pretty ragged up high, go for a 2nd order crossover to start with.... Then you may need to modify it to add some baffle step correction if you find you want more mids & bass.

To choose crossover components accurately, you'll need impedance plots for the drivers: Download ARTA: ARTA Your Page Title, Make this simple jig: ARTA Jig - AudioBlog: A simple loudspeaker measurement jig for ARTA

mdocod 8th March 2012 04:52 AM

Hi Jake,

Since I don't think there are any response or impedance charts available for the drivers (unless you can measure them?), you'll just be shooting for "functional" and nothing special here (I have no problem with this personally, especially if it's cheap. The chance to build something is always a good thing).

In my experience, the best results in these "shooting in the dark" crossover builds will come from a pretty simple 4 component x-over. An inductor on the woofer, cap on the tweeter, and a zobel circuit on the woofer. There's more room for error in a shallow slope build. Considering the fact that the woofer only has 1.5mm of Xmax, I highly doubt you'll run into power handling problems on the tweeter before running out of Xmax on the woofer even with the gentle single order slopes.

Use a crossover component calculator to come up with the values. 3K would be a safe bet, but I'd be tempted to push my luck with lower.

jacobthellamer 8th March 2012 04:55 AM

2 Attachment(s)
Cheers Pete!

That software looks very handy :)

My microphone is pretty average but I might have a go at frequency response soon.

Woofer:
http://www.diyaudio.com/forums/attac...1&d=1331184724

Tweeter:
http://www.diyaudio.com/forums/attac...1&d=1331184724

jacobthellamer 8th March 2012 05:03 AM

Quote:

Originally Posted by mdocod (Post 2937813)
Hi Jake,

Since I don't think there are any response or impedance charts available for the drivers (unless you can measure them?), you'll just be shooting for "functional" and nothing special here (I have no problem with this personally, especially if it's cheap. The chance to build something is always a good thing).

In my experience, the best results in these "shooting in the dark" crossover builds will come from a pretty simple 4 component x-over. An inductor on the woofer, cap on the tweeter, and a zobel circuit on the woofer. There's more room for error in a shallow slope build. Considering the fact that the woofer only has 1.5mm of Xmax, I highly doubt you'll run into power handling problems on the tweeter before running out of Xmax on the woofer even with the gentle single order slopes.

Use a crossover component calculator to come up with the values. 3K would be a safe bet, but I'd be tempted to push my luck with lower.


Cheers, I hope to get the best results possible from the drivers. They will be powered by ~15w amps and are unlikely to ever experience more than 60W.

I will have a go at getting some frequency response graphs up - hopefully my mic will not sway the results too much.

jacobthellamer 8th March 2012 07:06 AM

2 Attachment(s)
Fairly limited by my equipment, hopefully this will be useful.

Woofer:
http://www.diyaudio.com/forums/attac...1&d=1331193783
Tweeter:
http://www.diyaudio.com/forums/attac...1&d=1331193783

jacobthellamer 8th March 2012 09:10 AM

Steps seems to work better than Arta

Cross them at around 1.5kHz?

Full-range:
http://www.diyaudio.com/forums/attac...1&d=1331201058

Tweeter:
http://www.diyaudio.com/forums/attac...1&d=1331201058

giralfino 8th March 2012 09:33 AM

Hi, some random thought.

I assume you ran the impedance test in free air. If yes, it appears that the woofer has an Fs of 75 Hz instead of 46 Hz. Maybe you should input some 20-40 Hz tone for some hours to the woofer and re-run the test. And try also to calculate the full set of T/S parameters with arta. If the specs are way off to the real ones, you can't calculate a good box.

Second, the tweeter. It has a hard impedance peak at 800 Hz, meaning it probably doesn't have ferrofluid. Without good measuring equipment (more on that later), you should stay away from Fs and use at least a 2nd order electrical crossover on the tweeter. 3K Hz could be a good starting point. You could use a lower crossover point but you should tame the impedance peak, because if not it will appear also in the frequency response plot, and you will also stress the tweeter.

About the mic: what kind of mic is? Are you reasonably sure it has a flat response over a good range of frequencies (IMHO at least 100 Hz - 10KHz)? If not do you know its difference from flatness (you could input that into arta)?
If you don't know the response of the mic you can't use data collected through it for modeling a speaker.

About the frequency response graphs: what is the measurement environment? I mean did you ran the tests in free air or in a box or in an infinite baffle? It is difficult to draw some conclusions without this info.
And also you should gate the measure, because if not you are measuring also the room, and not only the drivers (and every time you move the drivers in the room you will have different data).
For now I'm ignoring the 10 dB difference between 300 Hz and > 1 KHz on the woofer.

Last. When you have reliable data for impedance and frequency response you can use some sw to model a crossover. For a good and free set of software see here: FRD Consortium tools guide

Ralf

jacobthellamer 8th March 2012 05:47 PM

Hi Ralf,

Thanks for the advice!

I did run the impedance test in free air, I will give the woofer some low frequency work tonight for a few hours and re-run the test.

The mic is a small cheap one from a headset - not the best I know but it is all that I had at the time, I will see if I can find another to use. The steps test gave far better results but it looks like the attachments dropped off. I will post these up again tonight.

I am measuring in a closed wardrome to try and eliminate noise from my room, the mic and drivers are 300mm apart.

sreten 8th March 2012 06:57 PM

Hi,

A sensible piece of advice is to remove the whizzer
from the dualcone, they have no place in 2-ways.

rgds, sreten.

I will state that the effort of making decent cabinets, is
a complete waste of time and effort for cheap drivers.


All times are GMT. The time now is 08:25 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2