3 way open baffle with Eminence and Skytronic

[SimontY]

This thread is to document the design and build of my 2nd open baffle speakers.

The first were a proof of concept using all cheap drivers - My dipole experiment

I then started a thread to discuss good, affordable drivers for a 3 way open baffle, dipole design - Which drivers for my next open baffles?

These threads and my experience, reading and simulation so far has led me to the following driver line-up, which may of course be altered if it doesn't work out:

* Eminence APT80 compression tweeter (comes with small horn)
* Eminence Beta 8 midrange driver
* Skytronic 902.222 poly cone woofers

So we've got PA style tweeter and midrange, supplemented with twin 12" woofers of the cheap variety. The woofers were chosen because of their extreme low price (£30/pair new on Ebay) and suitability for open baffle loading (Fs=29hz, Qes=0.71, 3mm xmax - enough with 2 drivers per side).

Bass will be enhanced with Graham Maynard's T-bass (transformer bass) circuit, which passively provides up to 6dB of boost at LF. It also provides some cut of upper bass, which is essential for a passive open baffle design. Driver energy storage is reduced and reproduction of low frequencies is greatly enhanced.

The goal is to produce a speaker with great timing, dynamics and tone. They also have to fit into a tiny room, which has limited me to the use of 12" woofers.

I've attached an initial sketch of how they might look.

Simon

[SimontY]

Let's start by looking at the mid driver. I've chosen a driver based on a few factors:

* small mids tend to sound wrong to me
* high sensitivity is a good thing
* PA drivers always sound interesting and detailed
* Eminence parts are affordable
* I've heard rave reviews of the basic Eminence ranges used as mids
* In the few 8" mid or mid/bass speakers I've heard the midrange comes across well compared to 6.5", 5" or 4" cones
* an 8" cone should barely move when producing midrange, so distortion will be low

It has been noted that the Eminence Beta 8 has a rising response, but I've noticed this is almost exactly the opposite of the roll-off after the dipole peak in my speaker shape.

I think this response may very well be to my advantage. I've taken a screen shot from the Beta 8 official response curve and overlaid it on the same frequency scale as my Edge plot for that shape of baffle and driver.

See attached pic.

[SimontY]

Here are specs of the Beta 8:

http://www.eminence.com/proaudio_spe...8&SUB_CAT_ID=2

Here's some info on the APT80:
http://www.eminence.com/tweeters.asp

and here (without horn):
http://www.eminence.com/proaudio_dri...ail_link=APT50

The tweeter doesn't seem overly fussy about where it's crossed over but to avoid a harsh sound I wish to avoid crossing over too low. I'll start at about 4Khz 1st order, and also try 2nd order.

Everything will be passive.

[SimontY]

Here's what Edge thinks about two 12" woofers on this baffle (wings included as width for the simulation).

[SimontY]

And here it is with some boost, to show (very, very roughly) what T-bass will be doing.

I think a 2nd order low-pass filter at about 100hz or so should be a good starting point to match up to the midrange.

I might invest in some speaker software to make filter design easier, but I'll be making measurements using ARTA at the very least.

Simon

[SimontY]

Here's an example of the T-bass circuit for anyone who's wondering what it looks like.

I've already sourced some 500VA 2 x 50v secondary Airlink brand transformers for the job. The low winding resistance should be ideal to achieve maximum boost and sound quality.

Simon

[SimontY]

I've posted this twice before in other threads but it's very interesting - the X-Baffle simulation for some woofers, including the Skytronic ones I'll be using here.

See how I'll get a flatter response to work with from the Skytronic woofers than my previous ones (7" sub driver x4).

Credit to Thorsten Loesch for writing the spreadsheet and Panomaniac, as well as others, for putting me onto it.

Baffle Construction

This is an area I'm open to suggestion on, but with the following constraints:

* cost must be relatively low
* complexity must be low
* materials must be readily available (e.g. from Arnold Laver timber merchants)

I know full well that 18mm MDF doesn't really cut it in open baffle speakers, which are massively excited by multiple woofers rocking the whole assembly back and forth.

I have heard of MDF & chipboard being an effective combination. I could have a cheap, pretty and hopefully solid baffle made from, say, 15mm veneered MDF stuck to 12 or 15mm chipboard...

Simon

[pacificblue]

Originally posted by SimontY
It has been noted that the Eminence Beta 8 has a rising response, but I've noticed this is almost exactly the opposite of the roll-off after the dipole peak in my speaker shape.

That is a nice coincidence. Maybe they were designed like that, because many guitar amplifiers are open baffle, and 8 to 12 inch is a common driver size for them?

[SimontY]

The more pricey Eminence "American Standard" lines (Delta, Kappa etc.) often have much more pronounced breakup, so I guess it's just a lucky coincidence. My Eminence parts were dispatched this afternoon so hopefully they'll be with me soon

[Moondog55]

Simon I wonder if there is any way to raise the total "Q" factor of those Skytronic woofers without robbing them of too much SPL??

Adding mass will lower Fs and raise "Q" but at the expense of SPL/efficiency.
From the little work I've been doing a woofer Q of around 1 seems to be the best compromise; above that the peak at resonance uses up all of the woofers excursion and below that there is less and less bass as the "Q" decreases.

What we need is a dedicated 12 or 15inch OB woofer that we all can afford. Yes I know I've said it before, I need to learn to accept compromise.