|
Mid-December 2014 Thar She Blows!
It
is about time I showed more than a glimpse. At long last, here's a system
pic, in the current state, which does change fairly frequently: You can see my waveguides, driven by Selenium D2500Ti
neodymium compression drivers with modified (damping coated) diaphragms at
this article that are enhanced with foam both in the throat and as a
ring, improving the roundover at the mouth within
this article on top of my ultimate
bass cabinets in cherry On the same shelf is the blue box, which is a temporary (though it's been that way for some time) set of transformer volume controls acting as my preamp. These are based on Stevens and Billington TX102 transformer volume controls, and are about as close to a perfect linestage as I'm aware of. Normal resistive attenuators, whether stepped or potentiometer based, have high output impedance and require buffer circuits implemented along with them if they're to drive lower impedance amps. Transformer volume controls don't have this issue, as they trade voltage for current like all transformers, and thus can drive extremely low loads if they're being operated in step-down mode. While this means that the preamp is at its best when -10 or more dB from the source level, it's an extremely clean way of controlling drive level without needing active circuits and the challenges those introduce. Modern systems have plenty of gain, too much in fact, so I've long advocated eliminating additional gain stages where unnecessary. For reference, in most systems I've used over the years, my volume demands have ranged from -30dB to -50dB for most use. I've only ever needed gain in a linestage when utilizing a low-gain phonostage. There are TVCs made by a number of manufacturers, including Stevens and Billington, Sowter, Silk, and autoformer solutions from Intact Audio, which are on the inexpensive side relative to other options (as TVC solutions are relatively expensive) but are reputedly excellent. See reviews of the Intact Audio stuff here. I recently received a selector switch from Elma, so re-boxing this puppy is on the short list, I've begun the chassis work already. The holey-fronted amp is the Motorola
-- a stripped down five channel power amp with a higher-end implementation of
Tripath Class D amplification. It is not rare to see works in-progress in
my reference system, and this unit is no exception. It has quite a bit of
spare chassis space, designed to allow me to implement any input controls and
logic into the unit. This space will be the focus of at least one future
article, maybe more. It is giving me somewhere on the order of 100 Wpc in
the four out of five channels I'm using in it, but I never use anything near
that much, given that each amp channel only has to drive either a high-impedance
woofer or tweeter of 16 Ohms or more and over 96dB/W/m sensitivity. Hidden from view behind the left speaker is the power distribution and second rack. On this unit is the Linn LP12, with Ittok LV.II tonearm and Dynavector 17d12 cartridge, sitting on top of a multi-layered foam-core board platform, which is light, stiff, and well-damped -- an ideal match for the lightweight construction of the Linn. On this rack are also the DAC (a 192kHz/24-bit TI-board based unit), a Behringer DEQ2496 digital equalizer with many other useful functions, a Marantz (Pioneer-based) laserdisc player, and lastly the Dayton SA1000 plate amplifier. In addition to the rough n ready platform for the LP12, there are a pair of Neuance shelves utilized for the LDP and temporarily for spare bits, which follow a similar light, rigid, well-damped philosophy. The rack frame is a custom three-leg arrangement with 0.75" brass all-thread rod for the legs and tubular steel shelf support structure. At the very bottom is a ballast shelf comprised of layered maple, damping rubber, aluminum, and steel. This is different in philosophy than I usually apply for vibration control, but is very effective and the high mass helps keep the rack from becoming top-heavy; with the three leg arrangement racks like this can be prone to tipping. The SA1000 on top of this extremely heavy shelf prevent this issue entirely, keeping the center of gravity low. The vast majority of the wiring is of my own design, but I do find myself using some fairly pedestrian wire on occasion. Good wire matters, but not as much as a crossover, and mine are forever changing. Speaking of which, here's a shot of the interim solution for crossovers I'm using a simple 3.9mH/10uF 2nd order filter on the BD15s, and 4R+5uF in series (charge-coupled for the 5uF) with a parallel 0.25mH on the Seleniums. Response is very good but I still will continue working on it to bring up a little crossover dip I'm not happy with. Note: Thank you to my wife for tolerating this much
gear, which occupies our main living room space and is the first thing one sees
when entering my home. I hope this lends some context to what I write-
since it changes so often, there's not been a time when I've wanted to put this
sort of description together, but for now, thar she blows!
|
|