Looks good, it's about time we get a few more designs with the Seas U18. I like this woofer quite a lot.
Just looking at the phase tracking on the simulation above. Wonder how it would look like if you only countersunck the tweeter, or gave the baffle a few degrees tilt? And maybe like Mark K put a 60 ohm (or there abouts) resistor over the woofer series inductor?
Just looking at the phase tracking on the simulation above. Wonder how it would look like if you only countersunck the tweeter, or gave the baffle a few degrees tilt? And maybe like Mark K put a 60 ohm (or there abouts) resistor over the woofer series inductor?
Last edited:
Thanks very much chaps, I much appreciate your suggestions. I'm not sure though but on the response curves above the big dip at around 500-600Khz the same thing that Toels Graveson is describing on his U16 design and also what markbakk mentions? See: SEAS CURV. Also having a first order filter on the tweeter worries me as occasionally I do like to play some modern music quite loud and would worry about damaging it. I also would like to see at least one other design using the U18, it seems that not one commercial manufacturer has ever used it. It has a very smooth response so should be a very easy driver to get the most from. I can only imagine that most builders of the IDUNN are quite happy and do not realise that there is something wrong with the response as it manifests itself very differently with different recordings and styles of music.
There are in fact quite a few manufactures using this woofer, or a oem very close to this. Electrocompaniet, Sonus Faber, Gold Note, Apertura. Probably more.
In a single woofer 2 way I think the U18 would run out of steam quite a bit before you have to worry about the DXT tweeter. I have run the 27TDFC in a waveguide together with 4 woofer 6,5 inch woofers in a 2.5 way in to stupid loud levels, using a 3,9uF cap as the only crossover component for tweeter. And still the limitation of that system was the x-max of the woofers.
In a single woofer 2 way I think the U18 would run out of steam quite a bit before you have to worry about the DXT tweeter. I have run the 27TDFC in a waveguide together with 4 woofer 6,5 inch woofers in a 2.5 way in to stupid loud levels, using a 3,9uF cap as the only crossover component for tweeter. And still the limitation of that system was the x-max of the woofers.
Humps and dips at the crossover are easy to get wrong. The odd rising slope of that driver, mixed with the baffle and room gain look like a challenging task. Very small AC offsets can play hell right there.
What does not make sense is the tweeter rising. Way more than the wave guide should give. I have had to roll off the top end of every 27T series. TBCG, TDFC, TDAC. The series is very sensitive to harmonic distortion exciting the breakup.
Have you played with tilt and toe with respect to the listening position? I think the lead on that you should be listening off-axis.
What does not make sense is the tweeter rising. Way more than the wave guide should give. I have had to roll off the top end of every 27T series. TBCG, TDFC, TDAC. The series is very sensitive to harmonic distortion exciting the breakup.
Have you played with tilt and toe with respect to the listening position? I think the lead on that you should be listening off-axis.
I hate the U18. I have it paired with the DXT and could never get it to sound good whatever I did. It has the dip in energy around 800Hz that totally ruins vocals for me.
Ha ha, all this is highly subjective but I would rate the reproduction of human voice as one of it's strongest points. Watching movies with my stereo system is a blast.
I have often found PP midwoofers sounding a bit "dull" for my taste, but this U18 is still smooth sounding as a PP should be but it is also more exciting and dynamic sounding than I'm used to from PP woofers.
It doesn't have very high QMS, but it still pretty punchy in the midbass. These are actually "growing" on me, and I'm really a paper mid kind of guy...
I have often found PP midwoofers sounding a bit "dull" for my taste, but this U18 is still smooth sounding as a PP should be but it is also more exciting and dynamic sounding than I'm used to from PP woofers.
It doesn't have very high QMS, but it still pretty punchy in the midbass. These are actually "growing" on me, and I'm really a paper mid kind of guy...
Not a poly-pro fan either. Their best attributes are being easy to implement and cheap. Seas went to a lot of trouble getting it to work almost as well as paper. Kind of pointless to me. Sticking with my Seas reed paper cones for now, but those new CSS have my eye.
You are correct. The lack of such correction pads in most designs of course has an economical ground. And in a lot of two way designs the attention is focused on the crossover region too much. Funny how history repeats itself, I used these LCR-circuits in cross-overs back in the 90’s, while measuring in a deadroom back then the dip was very noticeable and I didn’t take long to realize the big series coil wasn’t the way to go.
Last edited:
When you are using a very narrow cabinet you have to take that in to account in the crossover design process. Take a look at Audiexcite's version of a U16 design. A simple and cleverly executed design, utilizing the benefits of the smooth roll-off off the U16.
I think that is the key to make these PP woofers sounding good. Simplicity..
www.audioexcite.com >> Prestigious Two – Monitor DXT
I think that is the key to make these PP woofers sounding good. Simplicity..
www.audioexcite.com >> Prestigious Two – Monitor DXT
BTW- my simulation is done using cabinet adapted plots, and approximated Z offsets.
I first heard this wonderful driver in Maynard Goff's 'MalBecs' design back around 2010, and it has a wonderful sound in the mids. The weight is in the right place in terms of spectral tonality, and it rolls off easily without breakup. It's smooth and easy without losing details.
In terms of the 800-900Hz dip, a lot of woven drivers have this. Carbon fiber, Curv, glass-fiber, etc; and a lot of them sound very natural to me. Too much at 800Hz can sound whiny or nasal, and we definitely don't want that as the opposite problem.
I'd also like to try the U18 with the HiVi RT1.3WE, but we'll see how the first one goes.
Later,
Wolf
I first heard this wonderful driver in Maynard Goff's 'MalBecs' design back around 2010, and it has a wonderful sound in the mids. The weight is in the right place in terms of spectral tonality, and it rolls off easily without breakup. It's smooth and easy without losing details.
In terms of the 800-900Hz dip, a lot of woven drivers have this. Carbon fiber, Curv, glass-fiber, etc; and a lot of them sound very natural to me. Too much at 800Hz can sound whiny or nasal, and we definitely don't want that as the opposite problem.
I'd also like to try the U18 with the HiVi RT1.3WE, but we'll see how the first one goes.
Later,
Wolf
Very small changes in offset can make big differences in simulated or near-field measures. I find the AC using an old "pulser" kit.
You hear a bump far more than a dip.
I use OEM and nearfield no-crossover just as a selection guide. I use in-box-in-room for the actual design, then tune by ear guided by listening position measurements. The only measurement I no longer fool with is raw driver measures right at the surface. I also average my readings to 1/3 and one octave as they can unmask gentle slopes masked by the higher resolution measures.
You hear a bump far more than a dip.
I use OEM and nearfield no-crossover just as a selection guide. I use in-box-in-room for the actual design, then tune by ear guided by listening position measurements. The only measurement I no longer fool with is raw driver measures right at the surface. I also average my readings to 1/3 and one octave as they can unmask gentle slopes masked by the higher resolution measures.
As others have stated, tvr, I usually can guess Z offset pretty well and come out smelling like a rose.
The dip at 800 Hz for me is less of a concern. The dip at 1-2k is more troublesome and audible.
I never smooth to more than 1/6th octave for a look at the best-fit curve, but xovers are designed without smoothing.
Later,
Wolf
The dip at 800 Hz for me is less of a concern. The dip at 1-2k is more troublesome and audible.
I never smooth to more than 1/6th octave for a look at the best-fit curve, but xovers are designed without smoothing.
Later,
Wolf
Basically the design has two flaws, both regarding with the polar pattern. One is the less-than-perfect baffle step correction, the other is the converging output of the U18 combined with the DXT that radiates almost half-sphere at 2kHz, which gives a bump. The former requires a more sophisticated crossover, the latter a bigger waveguide.