Rerurn to Romy the Cat's Site


In the Forum: Horn-Loaded Speakers
In the Thread: Deep End DIY - Australian take one Macondo.
Post Subject: Please consider it.Posted by Romy the Cat on: 10/24/2025
 anthony wrote:
So, this year I've been steadily finalising the installation of Macondo and Melquiades into my room. It seems as though there is always more to do for perfection, but the end is very close now.  There have been plenty of building and measuring and even purchasing...here is the summary:

Melq Bass Channel - Bass Cannons 8 x Scanspeak woofers per side in time aligned position adjacent to horn stack.
Melq Midbass Channel - 3 x Vitavox K15/40 per side set behind listening position not quite in time aligned position (45cm/18" delayed)
Melq Upperbass Channel - 110Hz horn with Fane Studio 8M
Melq Fundamentals Channel - 250Hz tractrix horn with Vitavox S2
Melq Midrange Channel - 400Hz tractrix horn with Vitavox S2 using YO186 DHT
Melq HF Channel - RAAL Lazy Ribbon

Infrabass channel driven from preamplifier - SS buffer with gain to 4 x 18" Dayton Ultimax II woofers in sealed boxes each with a 400w Class AB amplifier...not time aligned...third order low pass at 15Hz...set behind listening position.

Horn stack and Bass Cannons are meticulously time aligned using acoustic measurements, namely alignment of excess phase in the passbands of each channel.

No digital signal processing.

Room acoustic treatment is complete with even reverb times down to bass frequencies.

Room SPL response is good from 10Hz+ with subs engaged, 30Hz+ without subs.

Immersive speakers (7) installed and Trinnov Altitude 16 processor in place with aim to experiment with Auro3D and other immersive audio.  Two channel system must first be completed with all i's dotted and t's crossed.

Have recently upgraded my dac and preamplifier...will talk about the preamplifier later, but it has been a revelation for more than one reason.

I am so pleased with the sound right now, but have had glimpses of other improvements and wish to explore them.  My next post in this thread will be about recent steps/improvements, current compromises and next steps.

Anthony,

I have much to share about my impressions and feelings regarding your project, but I’ll save that conversation for another time. For now, I’d like to address one specific technical point. I wrote about it quite some time ago, but given the volume of my earlier posts, it’s entirely possible that message was lost in translation.

You see, the topology employed in the Melquiades/Macondo system—its internal filtration approach—is both beautiful and problematic. It uses line-level filtration, where the filters also serve as bias-voltage providers. The elegance of this design lies in the fact that there are very few components directly in the signal path, and the filtration works against a constant, steady impedance. This allows for the implementation of an exceptionally precise, phase-consistent response curve. Everything seems ideal—at least in theory.

However, there is a subtle complication. The input side of these filters does not experience a truly stable impedance from the source. Of course, preamps generally maintain a consistent output impedance, but that consistency differs from one design to another. For example, if my preamp has an output impedance of 5 Ω, yours has 100 Ω, and someone else’s has 1 kΩ, then the Melquiades filters will behave very differently with each of them. In practice, this means the filter’s performance effectively needs to be recalculated each time the source impedance changes significantly—from, say, 5 Ω to 500 Ω.

I learned this lesson the hard way. At one point, I owned several preamps, though the Placette remained my favorite—and it still is. The original version I used (not the newer one from the last decade, which I don’t particularly care for) had a default output impedance of about 20 Ω, if memory serves. When I decided to commit to that preamp, I sent it back to Guy for modification, asking him to reduce the output impedance as much as possible.

He built a more substantial power supply, placed the output stages on larger radiators, and biased them to run significantly hotter. We debated whether to target 3 Ω or 7 Ω, and ultimately he settled on something in between—around 5 Ω to 7 Ω—assuring me that the current assembly could reliably drive as low as 3 Ω. When the preamp returned, I didn’t find the sound objectively better or worse, but it had changed. The difference wasn’t bothersome, yet it was clearly audible and worth acknowledging.

So my point is this: when you experiment with multiple preamps in front of your main system, the changes you hear are not solely attributable to the preamp’s own character. You are also—perhaps without realizing it—hearing how the Melquiades input stage reacts to varying source impedances. Please keep this interaction in mind when evaluating or comparing components.

Rerurn to Romy the Cat's Site