[audionut71]

It feels like this build has been going on forever... but it is finally nearing a level of completion!

I decided a while back to do something different from my normal project. When I got into the DIY scene it was because of meager funds available to dedicate to my system. As a result, i used a lot of closeout and bang-for-the-buck drivers. As time went on my skills, tools, and disposable income got better and left me wondering how much better the higher dollar stuff could sound. Being a tightwad at heart, I still didn't drop $200 on each woofer, but i did do a bunch of research and decided to build my front LCR's with the Vifa XT25 ring radiator tweet and 2 Peerless HDS Nomex woofs.

Here you see the TMM (left and right) cabinets with Cherry veneer on the sides and top and maple on the back. This is after 3 coats of Danish Oil (i think they got 6 total).

Here they are with the freshly routed baffle on them.

And this is how they look finished up.

The tweets were bought off ebay (I loved the truncated faceplate that Polk used) as were the woofs. Madisound designed the x-over, although since I started the project I have seen a few other folk's designs that are similar, so I am still playing with the crossovers.

I am starting on the center channel today so I'll throw some pics up as they become available. It snowed overnight here in St. Louis, so I won't mind being sequestered in the basement today!

 

[audionut71]

Now on to the center channel. I wanted to minimize the lobing that is inherent in the MTM, but really did not like the thought of a tweeter out on the end of a TMM setup. I got an idea from one of those ads for Paradigm that run at the top of the pages here @ the shack.

here is the enclosure in progress. The shiny stuff is epoxy, somehow I forgot to put glue on that joint (doh!) between the brace and the back of the box. To fix it I drilled into the joint at a 45 deg angle a few times on each side then put a fillet of 10 min epoxy down both sides of the brace & shot 3 screws thru the back. It is solid.

Here is a pic with the baffle after routing. I included the beer bottle for scale.

Here is the test fitting of the drivers. I use nail polish remover to get the Polk logos off of the tweeters.

Has anyone done a stacked MTM like this before? Any tips on x-over designs to minimize lobing?

 

[Anthony]

Well since you already have the drivers and cabinet, you are a step up on a lot of crossover designers. most of the time you are guessing at driver response, baffle response, etc. The downside is, you are stuck/limited to the drivers you already have -- but you can definitely work with it.

First off, I would get a full test rig for doing REW measurements (ECM8000 mic, phantom power preamp, and a soundcard with a line in). Then you can do nearfield and farfield measurements of the speakers, one driver (or the pair of woofers together) at a time. This way you will know the actual frequency response of your drivers.

Next, you need to know the impedance curve of your drivers individually. Parts Express sells a woofer tester that runs a sweep and reads out impedance. I haven't used it, but apparently it works pretty well. There's also speaker workshop, a free program that has tools to test drivers and model crossovers. You have to build a jig to test the drivers, but there is a lot of help on the internet about doing so. A final option is that someone may have already measured the impedance of your drivers. Not ideal, but it is a good starting point.

After that, you can model the crossover in a couple of free programs like Speaker Workshop or FRD Consortium free Excel spreadsheet tools (I couldn't find the link though, so they may have moved sites). From there you can play around with 2nd order, 4th order, crossover points, etc. Depending on how deep your tweeter can go, you are probably looking at a 4th order crossover to keep the bass out of the tweeter.

Great work on the cabinets!

 

[Anthony]

I'm dumb. here's the FRD page:
http://www.pvconsultants.com/audio/frdgroup.htm

 

[audionut71]

Anthony,
Thanks for all of the info, I guess my question was worded to broadly. I already have a full speaker workshop/ TrueRTA rig that I use for testing and x-over design. I also paid Madisound to use LEAP to design one for me (I did this prior to my aqusition of the tools).

What I was trying to ask was if anyone had designed a speaker with a similar driver layout and if a specific x-over topography had proved to be better suited to the (fairly unusual) tweeter-over-side-by-side-woofer arrangement.

 

[Anthony]

Ah got it. I'm not sure about beaming and the like with that pattern. I know you will get some beaming at frequencies above the center-to-center distance of the woofers (which you've minimized), and I know plenty of commercial speakers poke the tweeter up high like you did, so I know the alignment works (but not having seen the crossover of those . . . )

I would do the start simple and fix a problem approach. Linkwitz-Rielley alignment (2nd or 4th) should have an even power through the crossover region, which minimizes the beaming (if I recall the stuff I read in D'appolito's book a long time ago correctly). Treat it like any other speaker, wire it up, then do some off axis measurements in the horizontal and vertical. You've already solved the primary problem, which is driver separation, so I think you'll be in good shape.

Then again, I'm more of an envelope planner, built, test, rebuild, test, repeat kinda guy. If I overplan, the project never gets started.

 

[audionut71]

I would do the start simple and fix a problem approach. Linkwitz-Rielley alignment (2nd or 4th) should have an even power through the crossover region, which minimizes the beaming (if I recall the stuff I read in D'appolito's book a long time ago correctly). Treat it like any other speaker, wire it up, then do some off axis measurements in the horizontal and vertical. You've already solved the primary problem, which is driver separation, so I think you'll be in good shape.

Then again, I'm more of an envelope planner, built, test, rebuild, test, repeat kinda guy. If I overplan, the project never gets started.

I'm the same way, I'd rather build and rebuild than suffer the old Paralysis-by-analysis. I learn more that way too.

 

[audionut71]

Update:

I am finally getting back to this project and have made some progress:

The top and sides have cherry veneer to match the left & right mains, and quarter sawn oak on the back and bottom (I like the contrast of two differing veneers).

Here you can see the baffle fully routed and rounded-over with it's first coat of sanding sealer. Many more to follow.

Next I'll finish the finish on the baffle and start the x-over tweaking, stay tuned...

 

[zero the hero]

Nice looking speakers. I hope that's a 2 1/2 way XO?

 

[audionut71]

Zero,
I have 3 different x-overs that will be auditioned and measured, a 2.5 way and 2 different 2 ways. When I started this project, I paid Madisound to design a x-over (they did a 2 way), but since then I have stepped up my knowledge and tools. I will document the measuring and listening of the x-overs as soon as I get to it.

I'm pleased that you like the cabinets, I'm quite happy with them as well.:T

 

[zero the hero]

I'm very interested in seeing your results of the different crossovers. All the theories tell us that you shouldn't use a 2 way XO in a TMM/TWW config due to the distance to lower woofer, but then I wonder how companies like Klipsch can get away with TWWW 2 way in their big RF-83's (which I confirmed with Klipsch are nothing more than a 2 way).
That tweeter looks interesting, I'm guessing a Vifa ring radiator made for Polk?

 

[audionut71]

Zero,

I have been doing a bunch of research on commercial TMM's and it seems most are a standard 2 way (the Klipsch's and M&K's old LCR series). The Polk LSi series is an mtm but uses a 2.5 x-over. So I will be doing some serious testing, along with off axis (vertical and horizontal) sweeps to see what that tells me. I've had the L&R mains in use for the better part of a year now, and love them, but the engineer within thinks that they might be able to sound even better.

The stacked MTM for the center is an exercise in curiousity as well. I want to see if it ameliorates the horizontal lobing that plagues the MTM

The Polk ring radiators are indeed the Vifa XT25TG-36 according to the label on the magnet. They are used in their LSi series. They also have a stacked MTM inwall (LCi-RTS-C) center that uses this tweet that they give only sketchy info on the x-over (unlike the free standing LSi's where they give all of the x-overs details). I wonder why...

 

[JCD]

First, that is some really fine woodworking skills.

As for the XO, I'm not sure what can be done about lobing using just a passive crossover unless you're going with a MTM D'appolito system. I'm very curious to see how your different crossover's perform.

 

[audionut71]

Jacen,

Thanks for the kind words. Just like everyone else, I see only the flaws. Of course, I know where the bodies are buried...:innocent:

I would have preferred to use the vertical TMM that I used for left and right mains for the center, but that will not work with my current setup. I intend to do testing on the l&R mains and the center @ on axis, 15 and 30 degrees off axis on the horizontal axis. We shall soon see. I just finished my 3rd coat of sanding sealer.

 

[zero the hero]

what is the crossover point?

 

[audionut71]

Zero,
The designs that I am going to start with are:

Tweet @ 2.4k @18 dB/oct
Top woof lp 2.4k @ 12dB/oct
bottom woof lp 200 @ 12 dB/oct

Tweet and woofs in parallel 2500 @12 dB/oct

And the last one is not defined but can be found here. It is an 3rd order HPF on the tweet, 2nd order LPF on the bottom woof, and a 1st order LPF on the upper woof.

I'm all ears if you have some insight or knowledge to share!

 

[zero the hero]

Nothing much to add here, looks like you've got a pretty good handle there. You're obviously aware of the distortion issues that tweet has under 2k, and I think your center design is as good as its going to get for a horizontal 2 way with the woofers that close together. Off axis is not going to be great no matter how you slice it, since that's an unfortunate weakness of that tweeter. But lobing shouldn't be too bad even with that highish crossover point. Still, you might be better off with the center also using the 2 1/2 way if you have people sitting off axis...

 

[alexlee]

Oh, nice work. I really love it:daydream:

 

[Owen Bartley]

Looking good, audionut. I like how the drivers snug right up to each other with the flattened sides. The baffles and the veneer look great. Centre channels are tough to design, but it sounds like you've done your homework and yours should come out fine... keep us updated with your impressions!

 

[audionut71]

Finally wrapping this project up (7 months later, can you tell that spring and summer are my busy months at work?). I have been playing with crossover options for the Center and opted for a 2.5 setup with a 3rd order on the tweet @ 2.3k, a 2nd order on one woof @ 2.3k and a 1st order on the other woof @ 500. I'll grab some screen caps of the TrueRTA measurements when I get my laptop working again.

Here is the baffle after 2 coats of high build primer (from Autozone out of a rattlecan) and a few coats of thinned wood glue on the round overs and driver recesses.

This is a close up highlighting the speaker baffle and the flush mounting. I am very happy with the way that it came out.

Done and done.

This is the front stage of the theater setup as it sits.

Here it is in it's new home (I really should've dusted that shelf better:doh