Zaph L18/27TBFCG design questions

[Zaph]

In the efforts of simplifying the TMM 2.5 way Zaph designed....

What would happen if you took the original woofer circuit in the TM design - and duplicated this circuit for the second woofer, increasing the series inductor for W2 to allow for BSC?

Impedance doesn't change significantly through the xover region so I'd expect the crossover point to stay as designed. Modeled phase alignment seems to hold true (well consistent with what speaker workshop predicts of on the original TM). Minimum impedance (not shown) is approx 3.3 ohms around the 100-200 Hz region.

Have a look at the following, and please feel free to point my lack of knowledge / ignorance (I'm just learning!).

Thanks,
David.

Hi Dave,

What you've got there is 2 woofers with baffle step compensation, and is closer to a 2-way design than a 2.5 way. To edge that closer to a true 2.5 way design, the 2.2mH coil will have to get smaller and the 3.0mH coil will have to get bigger. But once you start to do that, you'll begin to see that the simple coil plus notch starts to not work as well anymore, as there aren't enough components in the circuit to shape the response as required. That's probably how I came up with the brute-force method previously mentioned in this thread. It could probably still be improved upon though.

Not by me however, I've got other projects on my plate.

Where did you get your data? That dip/peak centered at 550 shouldn't be there as the L18 is very smooth in that area. It looks like what might happen if there is insufficient damping in the box.

John

 

[Dave Bullet]

Hi John,

Thanks for taking the time to reply.

I got the raw data driver here:
http://www.htguide.com/forum/attachment.php4?attachmentid=7036

I believe it was produced by Seas. I believe these are aechoic and don't know the conditions of the test, so my simulation is really rather rough (I know).

Thanks,
David.

 

[Dave Bullet]

I've been tinkering with the 2.5 way TMM concept for a L18/27TBFCG speaker, leveraging off John K's work.

I've come up with the following design. It doesn't really simplify John's existing 2.5 way design (infact there is an extra crossover component). What it does show (model anyway):

- A similar response to John's original L28/27TBFCG TM design - but higher efficiency. (The white frequency line is John's original TM system on tweeter axis freq. response). I used John's original TM measurements as a reference since I plan on using exactly the same baffle dimensions and TM driver placement as John's TM speaker

- Excellent crossover phase characteristics (bottom picture is tweeter in phase with woofers giving crossover freq null)

- Higher order highpass filter on the tweeter to provide more low frequency protection (since I need to to play approx 4 - 5 dB higher due to the extra 0.5 woofer)

Let me know what you think. I plan on ordering 2 extra 2.2mH inductors and 3.3 ohm resistors so I can make up Zaph's original design if this doesn't pan out. The rest of the components in the design are in common luckily.

Porting options.... I'm thinking of only porting the 0.5 woofer in its own enclosure, and running the full range woofer in its own sealed enclosure. Is this typically how ported 2.5 way speakers are built? (or are both woofers usually ported)?

Thanks,
David.

 

[Dave Bullet]

Here's the crossover for the above...

(note - driver offsets were set to
0 inches for tweeter
-1.316 inches for full range woofer
-2.817 inches for 0.5 woofer

to match the 1 foot on-tweeter axis frequency and impedance curves from the original in enclosure measurements)

 

[Dave Bullet]

Modelled frequency response with tweeter polarity reversed (expected reverse null response)

 

[Dave Bullet]

Ok - here's the bad news. Apart from the more costly crossover, impedance dips to approx 3.0 ohms between 100 and 200 Hz

 

[Jonasz]

Dave, did you ever build these?

 

[Dave Bullet]

Hi Jonasz,

No I didn't. Reason being, I changed to wanting to try a 3 way with Seas drivers. I wanted a driver dedicated to the mids that was more of a mid driver. I ended up choosing an L15. In hindsight it may not be optimum (will require a 2KHz xo to the tweeter) but it is what I have to work with now (I've bought the drivers!).

Also since then I have invested in a measurement setup so I can more accurately design.

I am now working on a design with an L26 / L15 and 27TBFCG
here.

The above link is empty, as I am about to post some information on a sample crossover and baffle geometry.

Personally, if you still wanted to build the 2.5 way... I would recommend you build Zaph's untested L18/27TBFCG 2.5 way design (which rather than mine) for reasons too obvious to need specifying

I see Zaph has pulled the page on his website with his 2.5 way modeled xo. I have a copy and am happy to post it here (unless Zaph objects).

Cheers,
DAvid.

 

[Dave Bullet]

Looking at my 2.5 way design... the 0.5 woofer doesn't roll off fast enough. I am concerned too much of it is playing in the midrange space of the 1.0 woofer - potentially "smearing" the sound.

David.