|
|
| Elliott Sound Products | Speaker Box Project - Part 2 |
Main Index
Articles Index
Part 1 - The Cabinets
The speakers are complete, as described in Part 1 of this article. The next phase (actually completed some time ago now) was to finalise the Linkwitz Riley crossover, tweeter amp and re-establish my phono preamp. The idea was to make the final design a complement to the VP103 valve preamp, both in looks and function.
To this end, one power switch now brings my preamp and all seven power amplifiers on-line (six for the stereo tri-amped speakers, and the sub-woofer amp). There are four amplifiers in my original power amp unit, a further two in the tweeter amp, and the separate sub amplifier.
All mains switching is performed in the tweeter amp unit, which has a 20A relay. This is powered from the 12.6V heater supply from the preamp - I added an extra output for the supply, and there is a corresponding input on the tweeter amp unit. There is zero power usage when the system is off.
Construction was not without difficulty. For example, neither the toroidal power transformer nor the heatsink would fit in a 1RU (1 "rack unit" is 44.45 mm, or 1 3/4 inches) rack case, so some modifications were performed so that they would. I was able to utilise many of my workshop "toys" to bring this project to fruition - always a good thing, since it justifies their purchase to SWMBO (She Who Must Be Obeyed :-)
Figure 1 - Base of the Rack Case
Figure 1 shows that there is a cutout for both heatsink and transformer, and both project about 10mm below the bottom of the case. The bottom plate was cut and bent to provide mounting flanges for the heatsink, and a separate panel was attached for the transformer.
Figure 2 - Top View of Baseplate
In the view in Figure 2, the transformer can be seen, now recessed into its little cavity. The large screw to the left is for earthing. The power supply has been wired at this stage, and the filter caps were mounted on a piece of blank PCB material that I mechanically etched using a hand-held engraver unit. The power amplifier is actually the prototype of the Project 3A amplifier board - before I decided to extend the PCB for the output transistors.
The supply voltage is +/-25V from a 120VA toroid, and the amp is used only above 3,500Hz, so loading is light, even at very high volume. As you shall see, the main supply is also used for the preamp power supply (Project 05), which powers all the smaller boards.
The next photos shows the unit in a very advanced stage - it is almost complete. All the circuit boards are wired and installed, as well as the switching module (bottom right) and muting relays (bottom centre). Standard IEC mains connectors are used for mains input and switched output, with all mains connections properly shrouded to prevent contact.
Figure 3 - Rear View
The inscrutable box on the left side houses the phono preamp. Being high sensitivity, the last thing I wanted was hum or other noise, so complete shielding was the answer. It is also made with adjustable gain, since I use a (relatively) high output moving coil pickup most of the time.
The earth connection for the turntable is on the extreme left of the rear panel. That is followed by phono inputs and outputs, and the gain switch. The next set of connectors is the input from the preamp, followed by midrange and low frequency outputs. The "ordinary" phono socket is for the DC switching voltage. The set of binding posts/ banana sockets is for the tweeters - not surprisingly.
Figure 4 - Front View
From the front, you can see the Linkwitz-Riley crossover - two boards stacked on top of each other on the extreme right. Just to the left is the power supply unit (Project 05), and you get a clearer look at the muting relay board (just right of centre at the top of the picture).
Although it all looks neat and tidy, there were a few stages where it looked like a rat's nest, while I moved wiring about to get the lowest noise levels. Because there is so much packed into a small case, this was not a simple task, and even moving a cable by a few millimetres made a substantial difference - especially the wiring to the tweeter amp inputs.
In this view, the power supply regulator is again visible, and no - the heatsinks do not touch each other, it just looks that way in the photo. They are actually held apart with a small dab of hot-melt glue, to make sure that they cannot short out.
As you can see, there is no power switch, just one green LED on the front panel (plus the obligatory ESP logo, of course).
The final listening tests with this system have been a true revelation. It is quite possibly the most revealing system I have heard, with nothing - ever - disappearing into the mix. Even at quite astonishing sound levels (about 110dB peak!), a violin does not tear out your ear drums and leave them bleeding on the floor. It just sounds like an incredibly loud violin. Other instruments fare equally well, as do vocals, both male and female. Normally though, I have found that my preferred (serious) listening level is up to about 90dB (average unweighted), and usually less.
After some measurements, I can give the following (completely useless) information. For an average listening level of 85dB SPL, I measure just under 3V at the midrange and woofer terminals - i.e. about 1W for each. The tweeter voltage is around 1V, or 125mW. From this, I deduce that the total power per speaker is about 2W on average - the tweeter power can be ignored as insignificant. With speaker efficiency of 90dB/W/m, this means that I have a room loss of about 10dB at the listening position, based on applied power versus measured SPL. (Listening distance is about 2.5 metres.)
One thing that this system has highlighted is just how bad some recordings really are, and the artifacts of compressors and limiters (for example) are immediately audible. Although I prefer to listen to the music, rather than the recording technique or the system, sometimes it is just not possible when every flaw is revealed so clearly. As a result, I am re-assessing my CD and vinyl collection to some degree - in many cases to hear details that I have never noticed before. As always, there is a good and a bad side to having such a transparent system.
In fact, the sound is so realistic, that from another room voices sound as if there are people in my lounge room. On more than one occasion, I have had to get up and see who was there when I am working in my study, only to find that it is just the TV, which (naturally!) is connected through the hi-fi as well.
Overall, I am very satisfied with the speakers, crossover and indeed the complete system. It has not been a cheap exercise, but compared to the cost of purchasing a system that could come anywhere near it, I feel that I have a real bargain. While the system was put together (and documented here) in 2001, as of 2021 (20 years later) it's still in operation, with the only major change being to remove the dome tweeters and install ribbons. The electronics as shown are (almost) untouched, and have never failed. The only change there was to install the Project 110 remote control.
Main Index
Articles Index
Part 1 - The Cabinets