Seas W18 MTM: Want more. W22 MTM & Wave guide ?.

[Branwell]

Hi,

My current speakers are MTMs with the Seas W18EX and Millennium tweeters with the bottom end augmented by line array subs.

The crossover is a DEQX set with crossover slopes at 48db slopes @ 105hz and 96db slopes @ 1300hz.

I really like these speakers, but want more.
The main issue I have with them is that I would like a little more strength in the 100 to 300hz range and a little more strength in the lower tweeter region.

Am thinking an MTM with the Seas W22 and the Millennium with a wave-guide.

Am thinking the wave-guide would give the tweeter more low-end authority and the W22s would give the power I am looking for lower.

So the questions become.

Are these assumptions logical ?.
Is a crossover point of 1300hz, maybe 1200hz, too high given a 6” wave guide stuck in between the W22s ?

Thanks

Branwell

[Paul H]

The first thing I'd look at with your setup is moving both crossovers higher with the existing drivers.

The excel can sound good with a 1600-1700 high-slope lowpass, which should let the millenium 'breathe easier'.

I'm not sure about the lower end - what sort of subs do you have?

Paul

[Dennis H]

FWIW, a 96dB LR highpass will exercise the tweeter more than a 24dB slope. Plot excursion vs. frequency and you'll see it's true because of the sharper shoulder of the 96dB filter. Me, I'd move Fc a bit higher and reduce the slope to reduce tweeter strain. SL's LR4 at 1400-1500 Hz would be a good place to begin with the Millennium.

[AJINFLA]

Here's 2 metal 8's with a WG "stuck in between" : http://www.4sptech.com/Continuum_AD_features.php - although unfortunately not a dipole.
There was also a discussion here before about using LR24 with a notch.
Just two things to consider.

Cheers,

AJ

p.s. let Oldude know how this turns out. Might be the platform for the top panel he needs

[Branwell]

Hello,

Thanks for the responses.

On crossover frequency’s.
I’ve had a real love hate relationship with the Seas Excel drivers until I got the DEQX and was able to get very steep crossover slopes.

What I’ve found, at least to my ears, is that without the insanely steep slopes and crossed low, certain recordings seem to get the W18 to sound less than pleasant.

With the steep slopes ( 24db is not enough ) and low crossover point, it sounds amazingly transparent without all the fuss at its upper end. Makes a big difference.

So, moving the XO point much higher solves one issue ( tweeter stress ) but causes another. Much higher than 1400hz and the issue referenced above appears.

As to moving the bottom crossover up, just doesn’t sound good with the subs I have.

Tried a pair of 12” Peerless XLS drivers per side in sealed boxes.
Tried a single 15” pro driver per side in ported boxes.
Tried a line array with 8 x 6.5” drivers per side in ported boxes.

It comes down to transparency and resolution. The W18 is just so much clearer then the subs that crossing them higher just makes the lesser quality sound of the subs obvious.

So, we are back to where we started.

An 8” MTM with subs, if the wave guides and driver spacing don’t cause any problems, or perhaps a more conventional three way using a W15 mid to some sort of a high quality 10” or 12” bass driver might get the sound I’m after.

Thanks for the suggestions and if you have any more, please keep them coming.

Branwell

[Paul W]

Even with xo points under 1k, I've found improvement with attention to the Excel breakup. If they are not currently notched, consider trying it.

[JonMarsh]

Actaully, Branwell, I think your idea has a lot of merit, and I'd encourage you to persue it. The W22 has very low distortion up to 1.4 kHz, and the waveguide should make things easier for the tweeter- reduce the LF excursion for a given SPL.

That's why the current development for the "Uber Arvo" is moving that direction.

~Jon

[JoshK]

I am sort of on the same page myself. I have a pair of W22s for my variation on the Arvo project. I am working on a waveguide implementation, but as of yet I plan to use the H1212, not the Millenium. The constrained is really getting the waveguide in the works without loosing the c-t-c spacing ideal.

According to Zaph's work it looks like it should take a lot of the strain off the lower end of the tweeter's range, thus reducing distortion.

[cjd]

If you're, you know, bored, you may want to look at the Dayton RS265 or even RS315 HF subs. They may bring back some of the transparency you miss with the Peerless, yet allow a higher crossover frequency.

This may in turn allow you to push the W18 *slightly* higher - I have no data to back me up, but my ears tell me that removing some of the higher excursion bass duty helps clean up the top end nastiness - not a lot, but perhaps enough.

You could also mess with just a 48dB slope on the mid-tweeter. Or perhaps even a composite slope.

C

[Branwell]

Hi JoshK,

Do you happen to know what the spacing formula is ?.

Thanks,

Branwell

[JoshK]

Yeah...

Based on the speed of sound at 340.29m/s = 13397.22in/s and freq is cycles/s (Hz), you have:
wavelength = 13397.22(in/s) / freq (hz)

I believe you want the c-t-c spacing to be no more than 1 wavelength of the xo freq.

[Jonasz]

That's why the current development for the "Uber Arvo" is moving that direction.

~Jon

That's the most interesting piece of information I've seen regarding the Arvos for some time. Any thoughts about wich tweeter to use? :P

Wouldn't a waveguide be useful to boost the low end of a low q tweeter like the Crescendo?

[Branwell]

Hi JoshK,

Thanks……and please bear with an idiot if you can……….

If I’m interpreting what you are saying correctly, then:

@900hz, CTC would be 14.88”
@1000hz, CTC would be 13.39”
@1100hz, Ctc would be 12.17”

If this is correct, then:

Wave Guide = 6”
Center out of W22 = 4.35”

So:

4.35 * 2 + 6 = 14.7

This would give a crossover frequency of 910hz in an ideal world, not including a little room between drivers.

Does this look about right ?.

Thanks,

Branwell

[oneoldude]

AJ - Thanks for thinking of me! :lol:

[cjd]

If you offset your tweeter, it will allow closer spacing of the mids which helps things some. In general it will also help diffraction. Off-axis response on the "wide" side will also be smoother than the short side.

C

[JoshK]

yeah that is precisely right. Being a investment analyst who plays with spreadsheets all day, I did what was natural and plugged it into a spreadsheet. I then played around with what freq and offset do you need in order to keep the ctc spacing "tight" (as we have defined) while providing a real world spacing and a freq doable for the tweeter. This means about a 20º offset.

I don't remember off the top of my head what freq I used for such derivation, but it was somewhere in the 1100-1250 range. What I don't know yet, is whether the waveguide will allow a bit lower freq with Jon's CE style filter or a DEQX equiv. John Krutke has done some of that work, but I think more needs to be answered to point to where is the lowest reasonable cutoff with the applied filter in mind. Then the needed offset can be determined.

I'd rather have less offset than more (relative to the mids, not the baffle), as I can only presume this would provide more consistent response vertically and off axis symetrically.

[cjd]

I'd rather have less offset than more (relative to the mids, not the baffle), as I can only presume this would provide more consistent response vertically and off axis symetrically.

Vertical response is going to be limited by the topology (i.e. two mids) long before offset is an issue. Horizontal response, I have found an offset tweeter can provide improved response over a wider region to one side as compared to centered. John Krutke got me thinking about this - I think he found the same thing on one of his TM's. I've since read about someone else "discovering" this bit. So, I would re-evaulate your presumptions. I believe the data suggests you are incorrect. Either that, or you're not including all the relevant conditions in your statement.

C

[JoshK]

Are you talking about center of the baffle or centered with respect to the mids? Just for clarity sake, I was talking about offset relative to the mid all being offset on the baffle vs centered relative to the mid all being offset on the baffle. So I am not speaking to the prior at all. I think there is merit to the prior, as I have read from Jon and elsewhere. The later, I don't profess to know but I was postulating a guess.

[mikec]

I don't want to dissuade anyone from trying out a waveguide for themselves, but my experience with one was short lived. I used the same MCM waveguide that John and AJ used with a XT19. It certainly did help with the low end response of the tweeter but there were some aberrations that even my DEQX could not completely rid.

My waveguide measurements.

[oneoldude]

mikec,

What was the SPL at 1 m for this measurement?

Also do you have a without WG and with WG comparison graph? Or even better the without WG subtracted from the with WG. That will show what the WG actually did in a normalized way.

Interesting stuff!

[mikec]

mikec,

What was the SPL at 1 m for this measurement?

Also do you have a without WG and with WG comparison graph? Or even better the without WG subtracted from the with WG. That will show what the WG actually did in a normalized way.

Interesting stuff!

The SPL was approximately 85db. I don't have a curve w/o the waveguide but the graphs I've seen of the XT19 are flat as a board. I do have another graph with the overlaid response of a Fountek JP2.

[oneoldude]

Thanks

[cjd]

JoshK: As I thought, I lacked certain bits of information. You are likely correct about the value of keeping the mids and tweeter on the same centerline - that I have not done any research on.

C

[ralphs99]

Hi Branwell,

www.zaphaudio.com has a new project with 2x mid-woofers and a waveguide loaded tweeter that should be interesting reading for you.

My experience with waveguide loading suggests that a 6" waveguide is only good down to 2.5kHz or so. If you want to cross lower you'll need something larger. As a rule of thumb I use diameter = wavelength.

Cheers,
Ralph.

[Feyz]

It certainly did help with the low end response of the tweeter but there were some aberrations that even my DEQX could not completely rid.

My waveguide measurements.

In that plot, is the top curve the 1m waveguide response? If so what are the other curves? Just curious..