How to convert 2 way design into 2.5 way.

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Dear all,

First off, I'm not an expert in designing x-overs and that is why I'm here, To seek help.

I'm in the verge of making a new speaker, I already have the peerless HDS tweeters, I'm planning to use them in a 2.5 way design using 2 peerless 7 inch woofers per side.

When I was surfing the web, I found a 2 way project from our fellow member Jay_WJ using the drivers I'm planning to use, I then downloaded the dataset from his site ( thanks Jay !).

I added 1 woofer and inductor in the simulation using speaker workshop , to make it a 2.5 way, But I'm not sure what I did right or wrong in the simulation.

Below is a screen shot of the simulation I did.

My questions are:

Is it gonna sound good?
Are there wrong component values in x-over circuit?
Do you have any opinion about it?
Is there something wrong with the design?

Btw, The box design will be tilted to about 10 degrees, so I reduced the woofer offset in the simulation to -1 inch, Is this the correct way of using the offset feature in the speaker workshop? The x-over freq. is about 2.4Khz

Thank you , In advance.


  • 123.jpg
    47.8 KB · Views: 675

I'm not sure how you have modeled it but something is seriously
wrong. The level of the tweeter is the same as in Jay_WJ 's design
and simply put it should not be, it it should need to be 5 to 6dB
higher, depending on the relative levels of BSC of the two designs.

Edit : Sorry missed the resetting of R9, but 10R||2R is not the way to do it.

One way of doing a 2.5 way is a lowpass filter for both drivers, this
connected directly to the bass/ mid driver and via an inductor to the
bass only driver, for the given c/o frequency strongly recommended.

FWIW I cannot see the point of the LCR across the tweeter.

"...but 10R||2R is not the way to do it."

Thank you for pointing that out, I should have reduced the 10R , so that both R would share power, I forgot about that.

"FWIW I cannot see the point of the LCR across the tweeter."

Yup, You're probably right, I just copied it from Jay, It is to avoid the rising response of the tweeter, But I dont know whether the rise in response in the top octave is audible , Or whether the effect of putting additional component to make the FR flatter have some ill effect in the subjective SQ.

PS. This is where the design of the X-over came from. In this link. Jay 's Peerless HDS 2 way.

The top end stuff depends on the design axis, i.e. you can design
them to point straight ahead and be listened to off the direct axis.
This can have some power response advantages.

My main advice is still to change the low frequency topology.


A 2.5-way xover is a little more difficult to optmize than a 2-way xover. Below is my quick modeling of a 2.5-way xover. I predict that this will sound nice even without fine-tuning. Nonetheless, I recommend to voice at least the tweeter network to your taste.

An externally hosted image should be here but it was not working when we last tested it.

An optimal xover point for these drivers in an LR4 TMM design should be in a 1.75 k to 2 kHz range. A 2.4 kHz point is too high for this 7" woofer. No problem with tweeter's power handling at 1.75 kHz LR4.

Do not worry about a wide dip in the 200 to 400 Hz range. It's there because this sim uses a short TM baffle rather than a TMM baffle of a floorstander. If you include a correct baffle diffraction/loss sim result, this dip will be filled.

If you use a 10 degree slanted baffle, the woofer acoustic offset should be about -.3 to -.5 inches for the midwoofer, and -.75 to -1 inches for the .5 woofer. I used these values in my sim.

The CLR series notch filter in the tweeter net was not used without a reason. This was used according to my voicing experience of my Usher/Peerless 2-way. Including these components won't take much cost.

Good luck,

Jay_WJ has come up with the goods once again. Note the change
of topology to the more usual 3rd order treble / 2nd order bass,
this is good for the higher (nominally 6dB) output level required
from the tweeter (comparing a 2 way to 2.5 way same drivers).

Also note that the 0.5 way driver is fully suppressed, its output
does not "catch up" with the midrange driver above its c/o point.

FWIW my only minor point would be for the tweeter response
shaping a series L||C||R is better IMO than a parallel L-C-R.


(A definitive article on simulating 3-ways is yet to be written .....
No-one has ever detailed the myriad BSC possibilities available .....)
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.