CX120-8 easy build

[djg]

So these things really impress me after a few minutes of listening. Really the first coaxial design I've heard. For about $100, recommended if Dayton still sells these drivers. Even I can hear the precise imaging.

[Steve Lee]

Did you use the CX120 or CX150's - you weren't specific concerning the parts from Meniscus - will BobinGA's XO work for both speakers? (120 & 150)

[djg]

The CX120. I've never seen a CX150 design.

[wogg]

Those are sweet! Did you use Bob's crossover schematic?

[djg]

Yes, his XO, cab size and port, manipulated into a deeper shorter size. I did add 1/2" to the width to allow for a roundover, as BOB minimized the width (height) for his center channel use.

[BOBinGA]

I’m glad you like them. I measured them both horizontally and vertically and there was not a lot of difference, so they can be used either way. Your layout should work great.

[djg]

The final dimensions were influenced by the size of some leftover plywood pieces, but they worked out very well.

[cjbruce]

Thanks for the great design! I'm looking to create a set of in-wall enclosures to put behind an acoustically transparent screen, but due to a strange concrete foundation wall and a low ceiling I only have 0.5 cubic feet to work with. I also need something with off-axis response consistent out to 45 degrees. By any chance have you taken off axis measurements?

[davidB]

If you are putting them in-wall, you can look at the spec sheet off-axis plot, as it is done on a large, flat baffle.

Remember to use Bob's lower bsc values in the crossover.

[cjbruce]

This looks just about right for a small home theater. Nice work!

If you don't mind my asking, at what frequency are you crossing over to the sub on your AVR? I have some 0.8 liter 3D printed boxes for the TEBM46 2" BMR that sound decent crossed over at 160 Hz on the AVR. I was hoping that I could use the CX120s to go a little lower with boxes only a little larger, perhaps to a 100 Hz crossover?

[cjbruce]

Parts came in for the CX120-8 Easy Build today. Over the next few weeks I’m hoping to 3D print and cut parts for a 3.2 liter ported version. They will initially be mounted in 9” deep bookshelves as LCR channels for a 100” pull down screen. They will be crossed over to an RSL Speedwoofer sub at around 100 Hz. I will be using a Yamaha TSR-700 AVR to drive them. I don’t have a calibrated mic, so I will be equalizing using YPAO and my ears.

[djg]

Good luck. I used the YPAO mic with my Yamaha AVR, ended up changing the YPAO derived settings. 100 hz should be fine.

[cjbruce]

That's good to hear. I've been playing around with a bunch of 3D printed boxes for a single TEBM46 driver and I can't get a satisfactory sound with a crossover below 160 Hz, even with a WinISD-tuned box at 110 Hz. The 2" BMR just seems to be too small to make it work. I have a good feeling about the CX120 though. Its a small room with low ceilings and we keep the volume low during movies.

[cjbruce]

Here's the initial design sliced for 3D printing.

I'm planning to mount the crossover board in the slot against the left wall. The top and bottom will be made of an as-yet-undetermined hardwood, depending on what we have in the shop.

[cjbruce]

I did an initial for check for the 3D printed driver mount today. The fit is snug but not too tight. I’m going ahead with the full 52 hour print now. Fingers crossed for the next two days!

[cjbruce]

Just finished the 3D print and initial assembly. I added a little polyfill which seemed to help a bit in our initial listening. I'm happy with the sound on the test bench so far, but I will need to put it in the room with the subwoofer to see how it integrates.

TOP - 3D printed 2 mm marble PLA - Looks surprisingly good but doesn't fit well.

BOTTOM - Laser cut 3 mm baltic birch plywood - Fits well, but I would like to go with something a little nicer.

TO DO:
1. I am currently using a 1.2 Ohm resistor to attenuate the tweeter. I will need to listen to the speaker in the room to see if this tiny bit of tweeter attenuation is enough.
2. Neither the current 3D printed top nor the 3 mm plywood bottom will be in the final version. Need to figure out what looks and fits the best.
3. Need to design the magnetic grill and choose a grill fabric. I tried 3D printing a grill but it looked a little like a 70's Marshall amplifier grill -- not the look I am going for.

[djg]

Very nice. I think you would want thicker top/bottom pieces.

[cjbruce]

Absolutely. I feel like I am really pushing the limits of my 3D printer with an enclosure of this size. The sides of the enclosure are already eating up a roll of filament, and when I tried to do 10 mm thick top and bottom pieces I was getting air leaks and rattles trying to press fit the pieces together. The outside of the slotted port is not fitting well and is in dire need of some sort of rigid glue and sealant to stop buzzing.

I'm almost at the point where I want to give up and just make an MDF box.

[cjbruce]

I am having trouble with some peaks and dips around the crossover region. The speaker didn't sound flat initially, compared to some old Edirol monitors and a TEBM 65 BMR I put into a small box. I don't have access to a calibrated microphone, but I do have a Frederickson 2502 Function Generator in our lab and my iPhone 12 running the free "Decibel X" app. My uncalibrated phone mic is showing a big drop around 3 kHz compared to the TEBM65.


I didn't have the exact crossover parts specified in the "Reduced BSC" list, so I ended up going with the following:

C1 = 5.6 uF NPE
L1 = 1.0 mH steel core inductor
C2 = 4.0 uF NPE
R2 = 1.2 Ohm 10 watt

Not satisfied with what I was hearing I did a bunch of sweeps at 2.83 volts on the function generator, listening and watching for peaks and dips. Here's what I tried:

1. Inverting tweeter polarity - This replaced the big dip with a very unpleasant peak around 2-3 kHz.
2. Increasing the value of R2 - This did little to eliminate the dip, but padded down the tweeter so the highs started to disappear. With inverted tweeter polarity it helped reduce the 2-3 kHz peak a little, but caused a dip starting at 4 kHz.



I eventually gave up on the tweeter side and started looking at the low pass. I didn't have another inductor, so I started playing with C1 instead:

3. Adding capacitors in series to reduce C1. I was thinking by changing things around I could get rid of the some of the phase cancellation that was occurring between the woofer and the tweeter, but no matter what I tried I could get the big dip (normal tweeter polarity) or bump (inverted tweeter polarity) to go away.

Finally, out of things to try unless I buy more components:

4. Remove C1 entirely. With C1 removed suddenly the crossover region sounded better. The crossover is now a simpler (and cheaper) 1st order on the low pass with a 1.0 mH inductor. There is still a narrow dip at 3.9 kHz, but the speaker is measuring and sounding much flatter:



Now that I have a better baseline I would like to go back and redesign the grill. It is currently a 3D printed design with 3mm holes covered by speaker cloth. I suspect there is some diffraction happening due to the hole pattern that can be eliminated by opening up the entire thing and using cloth fabric only.

[cjbruce]

The speaker was tested close to the floor. This speaker will be used as a center channel below a pull-down screen below a 77" soffit. There is very little space so the speaker must be as slim as possible.

[cjbruce]

Testing is complete, and I've settled on first order filters for both low and high pass. The filters don't sum to perfectly flat on my iPhone microphone, and the speaker placement near the floor is causing some challenges that seem to disappear when I place the speaker a few feet higher.

Here is my final parts list:

Dayton Audio DNR-1.2 1.2 Ohm 10W Precision Audio Grade Resistor
4.0uF 100V Electrolytic Non-Polarized Crossover Capacitor
Jantzen Audio 1.0mH 15 AWG P-Core Inductor Crossover Coil

Total cost at the time I am writing this is about $20 for the crossover.

Here's the finalized design:

[cjbruce]

Finished! This is version 3 of the enclosure. Here is the link to build instructions on Thingiverse:

https://www.thingiverse.com/thing:5371462

Sound is good! The speakers compare well to a pair of Edirol monitors I have been using for the home theater, but these blend better visually with the gray fabric grills.

[LOUT]

I didn't expect the grills to look as nice as they do. I'm glad you went with the cloth, it seems to fit the look really well.
I also didn't expect to see the port covered by the grill for some reason, so that's kinda inspiring (I don't have a good excuse..just that thought never occured to me, lol).

[cjbruce]

Thank you! The cloth grill was necessary to make the room look cohesive. I'm happy with it too.

Its funny that you mentioned putting the cloth over the port. With my surround speaker design the air velocity in the port was so great that it was causing buzzing. I redesigned the grill cutout on all my speakers, including this one, to prevent this. I didn't occur to me that a simpler solution would be to not cover the port.

[Chris Roemer]

Looks nice.

[cjbruce]

Thank you! With the marble PLA there is no finishing required. The speckles in the material hide most of the 3D printing flaws.