Codename: “N.E.H.D.”
Article composed: 8/18/2009
Designer: Ben “Wolf” Shaffer

Codename is derived from .pdf file of Purpose A below, and is a supposed mistranslation from Danish to English. Documentation stated,”No ears have died…” which is possibly meant as, “Aural presentation was pleasant...”, and thusly N.E.H.D. is the resultant moniker.

Purpose A: To use the Tymphany LAT250-004 in a design, due to its unique aesthetic and construction. This stated unusual Line-Array-Transducer has been delineated as ‘limited use’ and this view has been an unproven verdict for its possible applications. Hopefully, this unit will have a favorable reputation and application once this experiment is concluded.

Purpose B: to use the HiVi TM1a in a design, as it is irrelatively and infrequently used. This mentioned mid and tweeter combination unit has not had favorable or many reviews, and had not been reviewed by this designer. This driver would hopefully integrate well with the unit in Purpose A, and provide a seamless and smooth, albeit odd, final aesthetic for this project.

The previous will self-destruct in 10 seconds…….


Pertinent dimensions and important information:

The LAT250-004 (LAT) is a weird driver in regards to its parameters, as its construction is also atypical. It has a series of ten 2.5” diaphragms, and two opposed neodymium motor structures positioned one at either end of a cylinder. Every other diaphragm is linked via a connection rod to one motor, and the other five diaphragms are linked to the opposing motor. This entire cylinder’s construction is mounted axially, across its diameter. There are slits open to the front and rear of the driver, which are cascaded to allow proper air movement without cancellation. Since these multi-pistons move in opposing directions, the typical vibration of an average conical driver is mitigated.

I was able to measure the parameters for this driver that did not relate to compliance or moving mass. The slits and opposing drive directions would not allow for a valid delta-mass test, so I took the three or four Vas specs I was able to find and made an educated/averaged guess. Since it was a guess, I will not state that information. These drivers are also very stiff in suspension until they break in a bit, so I had to run them a while before testing for real data.

Upon modeling, it was very apparent that a vented box, no matter what alignment or stuffing, would create a very peaky bass response. This was unfavorable, and directed me to use a sealed box with moderate fill to smooth out its issues. Unibox helped me spit out a volume of 15 liters for a trial run. F3/F10 was said to be 90/68 Hz, and whether that was accurate had yet to be seen.

I constructed a box for the drivers with outside dimensions of (HWD) 22" x 7" x 9.75" with 3/4" MDF for all but the left and right sides of the cabinet using standard ****-joints with biscuits. Sides are five-ply Pine. There is a 50 degree chamfer along the front vertical edges, which I did on the table saw. There is one horizontal brace half way in the vertical dimension. I also made the bottom panel removable to facilitate installing the crossovers inside the cabinets. Neither LAT nor TM1a allowed enough clearance to mount or install the crossover boards. Enclosures are stuffed.

Tactical placement:

Tweeter is positioned 2.75” from the top-edge, and the LAT woofer is mounted below the TM1a with 7” being the distance from the midrange on their respective centers. All drivers are centered horizontally on the front baffle.

Visuals:

I sealed the MDF with Elmer’s white glue, and used Duplicolor High-Build Primer from the auto parts store, followed by Krylon Champagne Nouveaux Satin Metallic spray paint. The sides were treated with Minwax stain pretreatment, and then Minwax Ebony stain. When the entire surface was dry, I laid on 2-3 heavy coats of spray gloss polyurethane to allow for a good and durable shiny finish. I used the 8-32 zinc-plated hex-headed screws for these three simple facts;
-I could use any 1/4” interchangeable-drive screwdriver socket in my cordless clutch-drill to drive them.
-I had to install 36 screws to mount the drivers, and the above made it easier. (30 in the LAT alone!)
-It looked really industrial and burly when complete.

Network manipulation:

The one main benefit to using 2 dome drivers and an axial-mounted woofer, is the resulting zero acoustic offset. All of these drivers radiate sound from the same vertical plane in space. Since this is the case, relatively shallow electrical slopes could be employed. The LAT is wired in series, and whether it matters or not I don’t know, but I placed the first voice-coil in the chain below the midrange. I’m uncertain about this due to the multi-diaphragms and their orientation. Both the mid and tweeter used a 12dB electrical high-pass plus attenuation resistors, while the mid and LAT low-passes resulted in damped 12dB and damped 18dB electrical filters respectively. The midrange also strongly required both an LCR shunt to nuke its self-resonant Fs, and a zobel to flatten its impedance rise to blend well with the tweeter. The tweeter has a parallel resistor to further adjust level, and attenuate its Fs a bit. You can solder the midrange negative to the tweeter positive and save yourself a connection from the crossover. These 19 parts do the job rather well for these drivers, and is not an overly high amount of components. I used L-pads as adjustable single resistors to dial-in the damping resistor values in the shunts of the midrange, but this is not a requirement. Normal fixed resistors can be implemented. Note that shunts with coils have resistors that include the coil’s DCR. Shunt caps can be NPE, but all series capacitors are polypropylene. All coils are 18/20 AWG air-core. The added DCR in the coils on the LAT were a benefit, so ferrous-cores should not be substituted. I integrated 4 dB of baffle-step compensation, and with the 3rd order electrical filter and box alignment, the LAT did the best it could.
Acoustic roll offs are 4th order Butterworth at 775 Hz, and 2nd order Bessel at 4.4 kHz. Sensitivity is set at about 85dB. Impedance minima is about 4.5 ohms at 2.8 kHz.







Considerations/Conclusions:

The LAT initially has a resonance around 1.2 kHz, and is to be highly avoided, followed by a steep roll off in its response. This is due to the 90 degree off-axis inherent response of the driver at the listening position. I would think it could be used up to about 950 Hz or 1 kHz if steep enough filters are employed. This opens options for other midrange-domes like the Dayton RS52, etc, but this of course requires an adjustment of the crossover for that change, or any other driver substitution. The LAT has the lobing of a typical planar tweeter, as it has minimal vertical dispersion and wide horizontal dispersion when mounted in a vertical manner. In other words, this is a planar-woofer, and would probably also mate well with the BG Neo8, and further to the BG Neo3 for a full-planar setup.

With this said, this design should more than likely be listened to on the midrange axis to minimize the mid to LAT lobing and integration/off-axis issue. I will also say that this issue is fairly minimal if on the midrange level.

The LAT bass definitely exceeds that of the simulation, and I would guess extends to about 55-60 Hz for the F3. The bass is ‘different’ in timbre, but also natural and tight. It does not quite sound like a conical bass driver.

The midrange (DMN-a) has a terrible harshness when not using the LCR to tame its Fs. I would not advise omitting it from the circuit, despite cost. With the LCR in place, it’s not as open as an open-back mid, or even a chambered Morel MDM55. I would assume this is due to the capped-rear instead of a chambered back. There are 2 iterations of this driver, capped and chambered. However, it does do a lot of things right. Vocals are clean and distinct, and saxophone has the same heavenly sound as other dome mids I’ve encountered. One caveat, it’s sticky! But it’s nothing a small piece of packing tape can’t remove in the way of dirt attractions.

The tweeter (unchambered TN28) is kind of a sleeper. It’s fairly clean, but not uber-clean, and not sssibilant or splashy. Integration on the midrange to tweeter is really nice, so that exceeded my expectations. The caveat for the tweeter is the terminal-mounts. They are plastic and flimsy. I melted one, and then snapped a terminal off when I removed it later. I was lucky to have another unit handy to replace it, and I glued it fast once I soldered it. The 25 ohm parallel resistor can be placed right across the tweeter like where I put it for ease of attachment.

All of these drivers placed together in this system really do sound rather good. If you have a lot of quick transients or plucks in your music, these have the ability to reproduce them as quickly as recorded. I’m guessing this is due to the small diaphragms and low moving mass of the 3 transducers. Fast acoustic guitar and flamenco just sound fantastic played through these. Tight and articulate are the 2 words I would use strongly, coupled with engaging, and moving. I’m really glad I experimented with the oddball LAT250, as it can be used well, and make some really good sound.

End Transmission, mission successful…