Lil' Brit... The "LS 3/5-MEH" Monitor build thread

[mzisserson]

So I ordered 30 of the sony buyout drivers, and the more I played with it, the better it looked in a small sealed cab with a Q of about 0.875-0.900. What better to put with a buyout driver than a big ol' fat paper cone tweeter? They have their advantages....

The is also the first venture that I am undertaking that is relying 100% on my own measurements. Here was the method:

1. Break in woofers, measure T/S with WT2
2. Design cabs around T/S
3. Build 4 cabs in one day (Why build just ONE pair? )
4. Measure Woofer IMP in cab
5. Measure Fq response mounted in cabinet 1M at 1W (middle of living room)
6. Get help from Jeff.
6.5 Remeasure
7. Realize my stoopit mistake and fix.... Throw it all in PCD and Design.

Below is what I came up with so far. The tweeter has really funky response. It is diffraction related for sure. If I tweak it I can get a 40dB reverse null, but I get a 5dB dip at 1.2 K doing so. So I stuck with the flatter FQ response and 20dB null...

Any thoughts? I did not include my "z" axis because it is accounted for in the measurements. Wondering if I need to move the woofer x-y?

This is DIY for the sake of DIY. It is fun!!!!

Subjectively speaking, this lil buyout is one heck of a driver. It will not handle much, or do deep bass, but it is super smooth with a fair amount of detail... running full range! The Bass it does have is "warm" and articulate. It is actually very nice.

[tom_s]

I like your shop gnome!

Looking good! 4 cabs in a day...that's productive. You could go quad for your desktop system!

[Jeff B.]

Any thoughts? I did not include my "z" axis because it is accounted for in the measurements. Wondering if I need to move the woofer x-y?

If you extracted phase, then you no longer have any Z offset shift accounted for in your measurements and would need to enter one into the program if you want an accurate model.

If you are running with only the measured phase from OmniMic, then your phase relationship may not be exact since Omnimic adjusts to the peak of the impulse response. If you measured each driver individually then their impulses may have a fixed delay between them, but you can not see that in two separate impulses, so that information is lost when the phase is adjusted within the program.

Either way, you need to enter the Z offset to be accurate.

This is why I detailed the three-measurement method in an earlier post recently (the dlr method). This finds the offset for you and you know that everything matches in terms of phase.

[mzisserson]

If you extracted phase, then you no longer have any Z offset shift accounted for in your measurements and would need to enter one into the program if you want an accurate model.

If you are running with only the measured phase from OmniMic, then your phase relationship may not be exact since Omnimic adjusts to the peak of the impulse response. If you measured each driver individually then their impulses may have a fixed delay between them, but you can not see that in two separate impulses, so that information is lost when the phase is adjusted within the program.

Either way, you need to enter the Z offset to be accurate.

This is why I detailed the three-measurement method in an earlier post recently (the dlr method). This finds the offset for you and you know that everything matches in terms of phase.

Thanks Jeff,
I did take a total measurement as well, so its a easy fix. I'll have another go tonight.

[arlis_1957@yahoo.com]

thats pretty cool. gonna try a 2.5way while your at it?

[johnnyrichards]

S
This is DIY for the sake of DIY. It is fun!!!!

When it stops being fun, it is time to go skiing or take up knitting or something. It is all about being fun - the fact we have something tangible when we are done is secondary to the experience.

How are those Goldwood cone tweeters?

[mzisserson]

When it stops being fun, it is time to go skiing or take up knitting or something. It is all about being fun - the fact we have something tangible when we are done is secondary to the experience.

How are those Goldwood cone tweeters?

Thanks Johnny. I agree. This project is like rehab for me in this hobby as of late.

It is. It measures well with a very damped impedance peak at 1.2K. From some folling with it, it seems to me nice and smooth, with decent extension. Laid back maybe, but typical cone. There is nothing blatently "Wrong" with it and for $3, it may just be a new good budget tweeter istead of neo-domes.

thats pretty cool. gonna try a 2.5way while your at it?

I am dying to. I need to get these little ones finished, then I am thiking of a MTMWWW/WWW (3 on fron 3 on back push/pull) 2.5 way. I am really likeig the way these sound. I have just been playing with them full range! Cannot wait to get the tweeter involved.

[johnnyrichards]

The cone mids and tweeters I have used over the years all seem to do pretty well with a single cap plus resistor. Eventually, it is what led me to realize that those cheap monkey coffins used such simple networks not just because the bean-counters demanded it.

[Deward Hastings]

The cone mids and tweeters I have used over the years all seem to do pretty well with a single cap plus resistor. Eventually, it is what led me to realize that those cheap monkey coffins used such simple networks not just because the bean-counters demanded it.

Tang Band W3-1053SC 3" Full Range Driver plus a Dayton Audio DC160-8 6-1/2" Classic Woofer, sealed, crossed at 300-400 Hz with a series crossover. Bog simple (just a couple crossover parts), sounds surprisingly good. Gets the good bass of the DC160 while avoiding the midrange mush, and while the W3-1053 doesn't have the "extended high end" of a cheap dome it does have a reasonably smooth and pleasant extension and the mids are reasonably clean (and uninterrupted by careless crossover design). Easy load for any amplifier (and only needs one per speaker, unlike the better quality 2-ways with active crossovers). Certainly plenty good for surrounds or "easy listening " bookshelfs, better than many of the two-ways that I've heard with much pricier drivers. I'd take it over the BR-1 any day. Used to be able to get the drivers on sale quite cheap, too (though I haven't seen that for a while), and it does take a fairly large inductor (watch the buyouts).

[billschu]

Looks like a fun one Mike. You'll get some inherent directivity with that cone tweeter which could actually be a plus in some regards depending on your goals.

Looks like I'll follow suit with an MTMMM for an on-wall HT speaker mounted near the corners using the Dayton ND16 in their elliptical waveguide.

The 80Hz rolloff should be OK for HT. The MT XO would be about 3 to 3.5Khz due to the loading limitations of the small waveguide. Center to center spacing might be an issue but likely not too bad 17 feet back.

The best part is that I'd just have to paint it with the room's trim paint using a mini roller and add some white grill cloth. Might be a weeklong project if I wasn't so busy with so many things.

Bill

[mzisserson]

Attached is the update with the z offsets and baffle geometry accounted for.

The crossover is more complex than I hoped for. 3BW tweeter 2LR woofer. Zobel on both. Cannot account for the HUGE diffraction signature I am seeing around 1K otherwise. Meh.

Thoughts?

[Deward Hastings]

The crossover is more complex than I hoped for. 3LR tweeter 2LR woofer. Zobel on both. Cannot account for the HUGE diffraction signature I am seeing around 1K otherwise. Meh.

Thoughts?

Consider trying the Tang Band W3-1053SC crossed lower (much lower) . . . it's more forgiving (of a simple crossover) and probably sounds a lot better than that "really funky" (your words) tweeter.

Oh, and what's "3LR" ? ? ?

[mzisserson]

Consider trying the Tang Band W3-1053SC crossed lower (much lower) . . . it's more forgiving (of a simple crossover) and probably sounds a lot better than that "really funky" (your words) tweeter.

Oh, and what's "3LR" ? ? ?

3LR is what you get after a 14 hour workday consisting of facilitating continuous improvement plans and trying to get buy in from a union.

3BW is more like it. the "Really funky" (It's a technical term) part is the diffraction, not the tweeter. I have them, for $3/ea I will stick with them. I enjoy a good challenge.

[Jeff B.]

Attached is the update with the z offsets and baffle geometry accounted for.

The crossover is more complex than I hoped for. 3BW tweeter 2LR woofer. Zobel on both. Cannot account for the HUGE diffraction signature I am seeing around 1K otherwise. Meh.

Thoughts?

Are you sure about those offsets? They make no sense to me at all.

[Deward Hastings]

I have them, for $3/ea I will stick with them.

GIGO. But "just for fun", if you really don't care what they sound like, sure . . .

[mzisserson]

Are you sure about those offsets? They make no sense to me at all.

I measured and repeted twice and got the same results both times? What doesn't look "right"? The tweeter is offset 1/2 inch from center, and the woofer is about 4.25" on center from the tweeter?

GIGO. But "just for fun", if you really don't care what they sound like, sure . . .

I apologize in advance for trying to have some fun. I forgot, this is all work no play in this hobby. I can guarantee they will sound good, and if I did not care what they sound like, why build them? Classy guy, Deward, to insult someone right in their own build thread.

[Brewski]

Mike I'd guess it's the z offsets Jeff is wondering about I think the recommended default for a 6.5" is .024-.025 ish I think I've seen .022 used for 5.25" drivers so you z offset is higher than I'd expect to see.

Take it easy
Jay

[CW Lindsey]

The crossover is more complex than I hoped for. 3BW tweeter 2LR woofer. Zobel on both. Cannot account for the HUGE diffraction signature I am seeing around 1K otherwise. Meh.

Thoughts?[/QUOTE]

Not trying to appear overtly ignorant here,but what is meh ?

[bkeane1259]

Not trying to appear overtly ignorant here,but what is meh ?

My kids use it all the time....

http://en.wikipedia.org/wiki/Meh

[Jeff B.]

I measured and repeted twice and got the same results both times? What doesn't look "right"? The tweeter is offset 1/2 inch from center, and the woofer is about 4.25" on center from the tweeter?

Ok. Correct me where I am wrong here, the image is very difficult to make out on my screen, so I could be missing something.

However, if the speaker is meant to be place horizontally with the tweeter beside the woofer then these numbers would make a bit more sense. So, if that's the case, let me know. I was just going by the cabinet pics you showed above. I say this because you have the woofer offset .120 horizontally from the tweeter, and you have the tweeter offset .013 vertically from the woofer. That also means neither driver is at the X-Y zero point, which isn't typical. On top of that you have a Z offset for both the tweeter and the woofer, although the relative difference between them looks about right. This adds excess delay which make it much more difficult to deal with the phase response on the chart. Now, maybe I am missing a minus sign due to the resolution of the image, but if these Z offsets are positive values then you have told the program that the woofer is closer to the listener than the tweeter is, and the opposite should be true.

When dealing with minimum phase sources, like loudspeakers, it is best to model them with one of them, usually the tweeter, being the reference "zero point" for the model. This would mean that the tweeter would have X,Y,Z coordinates of 0,0,0 and in this case, if the speaker is oriented vertically, the woofer would have .013, -.120, -.021.

So, again, if I have missed something, let me know, maybe I have it all wrong.