My custom Modula MT project

[toolsresearch]

Greetings, everyone. I’m new to this forum and would like to build a “Modula MT.” While I entirely trust that the series/parallel crossover will be wonderful, the “mad scientist” in me can’t resist trying my own, something between the Modula MTM and Modula MT crossovers here, both in terms of complexity and crossover frequency. Also, I am using LspCAD 5.25 and hybrids are out of the question, so I have in mind a parallel topology.

Right now, I have a nice LR7 acoustic at about 1.5k using the Jon Marsh measurement files. Here’s the question: At that frequency, how far back is the acoustic center of the RS180? In other words, I am looking for what LspCAD calls the “dZ” parameter in mm.

Measuring to the center of the front plate in the motor is a real joy in this case!? So, I tried to extrapolate from the Modula MTM design. I modeled everything, and varied “dZ” to get the deepest anti-phase null. Something like 68mm seemed right. Also, is anyone here familiar with this method:

http://www.speakerdesign.net/sbarticle.html

I think driver manufacturers should give a plot of acoustic center offset as a function of frequency. Ok, maybe not that, but at the very least the distance to the center of the front plate!?!

Thanks in advance!

[JonMarsh]

About 40 mm

[toolsresearch]

Thanks!

About 40 mm

Wow that was a fast response. Thanks for correcting my figure.

Below is an attempt at a crossover for the 180and 28. This post is meant to invite comments.

I made the assumption that the measurement files in post #20 above were made under the same conditions, i.e. that the 0db level in both cases are the same so that the 180 and 28 can be matched for amplitude using just the data without additional scaling. Please correct this assumption if it's wrong.

The woofer circuit is just a re-optimized version of the Modula MTM with two exceptions. First, it has what I call a poor man's zobel (PMZ?), namely, resistors in parallel with the driver to swamp voice coil reactance at all frequencies. It let's you get smaller coils taking up less money and less space. Also, you can tweak BSC by swapping resistors. However, series coil resistance and the PMZ do make for a kind of L-pad, so keeping series coils big will help. Jantzen 15 gauge coils burn up about 2db with this approach. A standard zobel wouldn’t be that much more expensive, although the coils would be more, and if sensitivity were a big issue you would want the standard zobel. Second, the bypass cap on the second inductor tunes the notch to 8k. This gets less attenuation in the crossover region, but the filter is now at something like -90 there, and -65 at 6k.

The reduced attenuation in the crossover region also make possible a simpler tweeter circuit. It places a standard 4th order circuit inside the legs of an L-pad. Again the shunt R swamps voice coil reactance. Then too the larger caps, necessitated by the small load the 4th order components "see," swamp some of the woofer circuit's inductive rise. Z phase is tame as a result. The tweeter has got it pretty good being somewhere around -60 at 500 cycles.

Mark it up!

[cjd]

tools: might be helpful to toss your own thread out there for this.

Off the bat, what the HECK are the 15R in the mid-woofer circuit?! Other than a huge heat source...

[toolsresearch]

R for Z

Off the bat, what the HECK are the 15R in the mid-woofer circuit?! Other than a huge heat source...

Thanks for the note.

Yes, they will get hot. I'm not wedded to the unorthodox idea. Their purpose is explained in the post: they make the impedance less reactive for one and second they allow for smaller inductors. In the drawing, top trace is just the driver, bottom trace is with the parallel R.

It would take both an LCR and a CR to get that result normally. This is a shortcut.

[toolsresearch]

Cabinets

Worked out interior box dimension using the following ratios: 1 : 1.6 : 2.33 (from Sepmeyer) to keep modes out of each other's way. A 9" baffle--thus interior dimension of 7.5 wide, 17.5 tall , and 12" deep using 3/4" MDF--gets me close to an appropriate volume.

Used a single vertical brace from Birch plywood.

Glopped everything inside using a borosilicate microsphere mixture. I like to call the spheres NASA dust; it's from these folks:

http://www.hytechsales.com/

Also, went with a 1.5" baffle from two sheets. I put T-nuts on the back of the outer sheet and laminated them inside the two pieces. That way, the chamfer on the back of the inner sheet could go 360 degrees!

Ping!