Project Black Diamond MTM

[Tweaker]

After a few fits, starts and research, I’m on to my next project that I’ll be calling Black Diamond MTM. I’ve looked over many projects on this site and others and I’ve recently built a pair of Nano Neos and Overnight Sensations. It’s time to go it on my own and start from scratch. With all of the incredible design tools available and with input from this board I feel confident in doing so.

Some background; several months ago I built a home theater setup using the buyout Jamo 5” drivers paired up with a Vifa DQ25SC16-04 tweeter.
14400
This tweeter is what Zaph uses on his ZA5 series of speakers. I really like the way that tweeter sounds and according to Zaph, they can be crossed over low at 2.2kHZ. Perfect for an MTM center channel laying on its side. With my existing speakers I’m getting decent horizontal coverage as far out as 45 degrees. My target is closer to 35 degrees. Those were the first speakers I made a serious attempt at designing and building and the sins I learned along the way hopefully won’t be repeated, Ha!
At the time, I figured the tweeter is a proven entity and at $4.50 each for the drivers I didn’t have that much to lose if I ended up being unhappy with the setup. The Jamo drivers sound nice enough but have no low end. I’ve got everything paired up with a Yamaha YST-SW216 sub sitting on the floor behind the monitor stand which sounds good for the absolute bottom end stuff, but I’d like to be able to get more from the screen channels too. I don’t get a sense of low frequency detail, it’s just lots of boom up to around 120HZ. It’s not a setup issue, it’s just that the Jamo’s don’t have it in them and frankly, the Yamaha may not be the best sub in the world either, but hey, that’s another project. At any rate, I'm not happy with the way these guys look and my gal asked if they could be made smaller. I agree with her, these are ugly and overwhelming.

My requirements for the home theater system require that the l/r channels be located right up next to the screen and that the center be located below the screen. I tried both the Nano Neo and the ONS as a center channel laying on their side, but I’m not getting as much volume as I want, plus the fact that I have a physical limitation of a maximum of 5-3/4” below the screen. I modeled up a MTM arrangement using both the Dayton ND105-4 and ND105-8 as well as the HiVi B4N. The ND’s sound great but have some drawbacks for me, in particular the pincushion frame and the enclosure size. I built the Nano Neos with the driver mounted from behind the baffle. It kind of worked, but was a real pain in the you know where. The challenge is carving out enough material behind the baffle to bring it forward while still leaving enough material to securely mount the driver. I ended up leaving about ¼” of thickness around the speaker and then chamfering the opening from the front side of the baffle. I’m just not happy with the way it looks and the mounting screws don’t have much to grip. I may play with that concept more someday, but for now I’ll use the HiVi B4N flush mounted.
The Nano Neos will be replacing the ONS at my desk and the ONS will be put into service as surround channels.

OK, on to the new project.
One of my favorite things to do is ski. I always say, let gravity be your friend! The snow has been incredible here in the NW and Friday I spent the day hitting run after run of deep power on the Black Diamond slopes. I figured that’s a great name for these speakers.
I’ve been playing with PCD and Win ISD modeling the enclosure and crossover. I’ll post that in later posts as I move further along. I’ve included the drawings and renderings in this post. I was able to get most of the wood cut yesterday and I’ll put those pictures up too after I tweak them for the web.
So, here goes!
I use AutoCAD for my design work and for this kind of work I usually draw as either ANSI-C or ANSI-D. These drawings are ANSI-C so if you print them on a letter sized sheet scales noted will be halved.

This is a rendering showing the speakers around my HT system. They'll be close to the screen on all three sides so I don't plan on doing much in the way of BSC.
14395

This drawing shows sections and dimensions of the enclosures. All three enclosures will be identical and I'll be making a stand of some kind for the left and right channels. The front baffles are oriented to place the tweeter away from the screen. Since I've got them up so close to the screen I'm hoping that will provide the best sound stage. It seems to work OK with the existing setup.
14396

This last sheet is the materials cut list. I was able to get all of the wood cut and the dado/ rabetting done yesterday as well as cutting out the braces. I'll hold off on cutting out the baffle openings for now.
14397
I'll put the PCD and WinISD models up later this evening along with some pictures of the panels being cut. For now, it's a non-rainy afternoon here in the NW and I've got to take advantage of it!
Regards,
Greg D.

[Tweaker]

I've attached the PCD models showing the crossover design. I'd appreciate input if someone sees something wrong, or perhaps a better way.
14412

Following, are the crossover components.
14413

This response model is 30 degrees off axis and still looks pretty good.
14414

And now things fall apart at 45 degrees off axis.
14415

My worse case scenario is about 35 degrees off axis so I'm comfortable with how things look.
This is the crossover schematic.
14416

I appreciate any comments,

Thanks,
Greg D.

[andykriech]

I can't really weigh-in on your design, but wonder if you will have enough room from the back wall for your rear-firing ports? your drawing makes it look to be very close to the back wall. Maybe front-firing or front slot port might work better.
Just a thought.
Andy.

[Mark65]

Since the design looks to be about 90ish dB sensitive, I'm thinking that you don't have any BSC incorporated. I like about 4dB, I think it's a good compromise for varying environments, but season to taste, for sure. Also, I think that peak at around 4k is going to be annoying, perhaps even sibilant (not 100% sure on the sibilant part). I'm pretty sure that the B4N's breakup is to blame there. You might try a tanked 2nd order electrical on the LP, remove the first coil from the low pass, but put the dcr of your coil in that space, then add an LCR notch before the xo, using your coil value, a small cap (.1uF or so) and 9999 for the resistor. You also want to put some resistance in series with that, 2-4 ohms is usually enough.

I like the looks of the design, though!:)


Mark

[bmaupin]

When I design a crossover, I target an individual response for the woofer, and one for the tweeter, so they will sum well. In the following pic is a 2.2k LR4 target and the resulting B4N x2 response (using my measurements).

https://home.comcast.net/~bradmop/B4N@2200.jpg

You might try these values and try to target 86 to 87dB in the passband like Mark recommended.

L: 1.6mH
C: 0.15uF (small)
RC: 1ohm & 20uF

The spl on my measurement is off about a dB and you can tell I need to have a little less BSC. The tweeter filter is for the DQ25, but needs to be simplified - too much $ in xover for a cheap tweeter.

[Tweaker]

From Andy:
I can't really weigh-in on your design, but wonder if you will have enough room from the back wall for your rear-firing ports? your drawing makes it look to be very close to the back wall. Maybe front-firing or front slot port might work better.
Just a thought.
I thought about that too. I think I'll be OK though. In the CAD drawing I placed a flat surface behind the detail to help render a decent drawing with a "dramatic" background. Ha!
In reality, the center channel vents will straddle a center column with over a foot of free air behind. The system is setting at about about a 20 degree angle to the family room and set into a small corner that is nothched out of the room with a wall and fireplace to the left and a bay window to the right. The left side speaker is angled up against the corner of the notch with the front baffle forward enough to provide about 2' of clearance to the left and about 5" clear behind the speaker. You can see the distance to the screen in my pictures. The right side will have about 8" of clear air behind the speaker and typically when we're watching video we'll have the curtains drawn. The window faces east luckily and is only a problem for me during football season.:rolleyes:


From Mark:
Also, I think that peak at around 4k is going to be annoying
I think you're right. This is something Paul Carmody mentioned on his Overnight Sesations MTM. He placed a .22uF cap across a 1.1mh coil to compensate for what he called the "God awful" breakup at 4k.
I was thinking the crossover had pushed the woofer down enough to avoid the problem. When I model the tweeter and remove the woofer the bump is still there.

In this screen capture I'm showing the tweeter alone.
14524
This capture shows both drivers with raw response, the filtered response and the overall response.
14525

You might try a tanked 2nd order electrical on the LP, remove the first coil from the low pass, but put the dcr of your coil in that space, then add an LCR notch before the xo, using your coil value, a small cap (.1uF or so) and 9999 for the resistor. You also want to put some resistance in series with that, 2-4 ohms is usually enough.
This is how that models. The bump at 4K is still there. But I now know how to place a tank circuit! I wondered how you do that in PCD.
14526
I really think the bump is from the tweeter alone.
check out the trace Zaph provided of the tweeter in the following link.
http://zaphaudio.com/temp/Vifa_DQ25SC16-04-FR-0-15-30-offaxis-horizontal.gif
I traced this graph by placing a screen capture into AutoCAD to draw the response curve and graph box with solid clean lines, then copied this into SPL Copy to create my .frd file.
It could be something in my trace that's causing the bump to appear so pronounced, yes/ no? Also, this is a $13 tweeter......

Brad,
I'll play with your version later and see what happens.
Best regards,
Greg D.

[Mark65]

Well, now you need to pad the tweeter more, and a lot of that peak will go away. Also you still have room to shape the knee of the tweeter response. You're getting there!

Mark

[Chris Roemer]

I thought about that too. I think I'll be OK though. In the CAD drawing I placed a flat surface behind the detail to help render a decent drawing with a "dramatic" background. Ha!
In reality, the center channel vents will straddle a center column with over a foot of free air behind. The system is setting at about about a 20 degree angle to the family room and set into a small corner that is nothched out of the room with a wall and fireplace to the left and a bay window to the right. The left side speaker is angled up against the corner of the notch with the front baffle forward enough to provide about 2' of clearance to the left and about 5" clear behind the speaker. You can see the distance to the screen in my pictures. The right side will have about 8" of clear air behind the speaker and typically when we're watching video we'll have the curtains drawn. The window faces east luckily and is only a problem for me during football season.:rolleyes:


I think you're right. This is something Paul Carmody mentioned on his Overnight Sesations MTM. He placed a .22uF cap across a 1.1mh coil to compensate for what he called the "God awful" breakup at 4k.
I was thinking the crossover had pushed the woofer down enough to avoid the problem. When I model the tweeter and remove the woofer the bump is still there.

In this screen capture I'm showing the tweeter alone.
14524
This capture shows both drivers with raw response, the filtered response and the overall response.
14525

This is how that models. The bump at 4K is still there. But I now know how to place a tank circuit! I wondered how you do that in PCD.
14526
I really think the bump is from the tweeter alone.
check out the trace Zaph provided of the tweeter in the following link.
http://zaphaudio.com/temp/Vifa_DQ25SC16-04-FR-0-15-30-offaxis-horizontal.gif
I traced this graph by placing a screen capture into AutoCAD to draw the response curve and graph box with solid clean lines, then copied this into SPL Copy to create my .frd file.
It could be something in my trace that's causing the bump to appear so pronounced, yes/ no? Also, this is a $13 tweeter......

Brad,
I'll play with your version later and see what happens.
Best regards,
Greg D.


As was already stated, your 4k bump looks to be your tweeter mostly. You can see that your high pass Xfer function has a peak right there (not coincidentially, you also have a very low impedance dip in that very same range). Drop your tweeter level a few more dB, and fiddle with the Q of your HP shoulder until that smooths out.

Actually, you've got the same hi-Q problem with the Xfer func. on your LP as well. Do you have a Zobel on it?

Chris

[Tweaker]

I'll keep working on the crossover but in the mean time I've started the build.
Finaly! a dry weekend in the North West.
I don't cut MDF in the garage so I need to wait until I get a dry couple of days.
Welcome to my outdoor "shop".
14572
I hook up the shop vac to the saw even outdoors and I also wear a dust mask. Even with the shop vac I still find this stuff is hard on the lawn.

I find the best approach is to rip my boards to the width I need and then cross cut the panels to length.
I built this panel cutting table years ago.
14573
I made two oak stips the width of the guides and about 1/8" talled than the saw table surface. With the strips in the guides, I screwed down the plywood as square as I could, slide the panel table back and raise the saw blade up and cut into the panel table. Push the table forward to within 8 inches of the near edge and you've got a straight cut.
Use the cut as the line to add the guide fence by setting a framing square against the cut line. Easy! You've now got a way of cutting panels with perfect 90 degree corners every time.
14574
Place the panel table on the saw and raise the blade.
14575

14576
Set up the table by placing a framing square against the guide fence and setting the cut mark against the blade. Clamp a block against the fence and you'll cut identical length panels each time. Just be careful that saw dust doesn't build up at the inside corner of the block or your cuts will be shorter than you want.
I'll post more photos in my next post.
Greg D.

[Tweaker]

With the panels cut, I changed over to my dado blade and cut dados and rabbets next. Here' s the finished cut panels.
14577
14578

I originally cut the internal braces by running my jig saw between these holes. It worked OK, but unfortunately the brace is too large. I'll never be able to reach in to mount the drivers.
This is a picture from the rear.
14579
I recut the braces with another "home brew" jig for my router.
I used a roudover bit and moved the fence back so it would leave about 5/8" of material. I set the depth of the bit so that the roundover would be equal on both sides.
14580

This is a picture during glue up with the "adjusted" internal brace.
14581
The next post I'll show how I did the rear ports.
Greg D.

[Tweaker]

Speakers are cooking now with either pink noise, or 65Hz. Sounds like the airport in my garage!
14608
Time to set up the rear panels and add the ports.
I like to flare my ports in the hope of lessening the chance of hearing "whoosh".

I start by using a forstner bit to a depth of 1/4" at the same diameter as the outside of the pipe. I'll be using 1-1/2" schedule 40.
14609
I use a home built router table with my old Black & Decker fixed base router turned upside down and mounted to a piece of plexiglass. I made the base so it can flip up and fold away for storage. I've got a fence for it that you can see in an earlier post but it's been removed for this.
14610
I'm using a 1/2" round over bit that I set the depth so that it just touches the outide edge. When I route out the port I like to start a bit low and then slightly raise the bit till I just slightly leave an outline of the outside edge of the bit.
14611
I tap in a piece of the pipe into my inside mounting point. This will be what the router bit bearing will ride on.
And here is the result.
14612
A bit of light sanding and it will finish out nicely.
Regards,
Greg D.

[Tweaker]

It's been a very busy few weeks but I'm now getting back to the project.
I've changed the crossover and I think the response looks much better.

I put the LF drivers and tweeter in the enclosure and checked the raw driver response using an Earthworks calibrated mic, Neutrik MiniRator with pink noise and SIA SMAART-Pro RTA. I'm liking what I see and in spite of that breakup at 4K these sound really nice. Using two LF drivers, the tweeter isn't too anoying, but it's definitely louder. With a proper crossover, I think these will sound wonderful.

I see the tweeter xover as follows:
10uf
.25mH
30uf
and a 1ohm series, 5ohm paralel L Pad.
This combination smoothed out nicely.

The LF xover as follows:
.5mH
35uf
.15mH
and a 4ohm/ 10uf Zoebel across the drivers.

This is the summed response chart:
15107
This is the summed impedance chart:
15108
More to come, I'm really struggling with this wet weather we've been having and the finish of the enclosures. I've come sooooo close, but I get a few bubbles in the paint now and then. These 100% humidity days are tough on spray painting.
More to come.
Regards,
Greg D.