Aunt Corey's Homemade Buffered Passive Preamplifier Page 4
A design like this is pretty simple for the experienced constructor, so there aren't too many guidelines to lay out. Definitely make sure you use RCA jacks which come with isolation washers, like the Tiffanies or the Vampires, in order to keep the jacks from touching the chassis; the only ground point in the entire circuit should be that single point ground mentioned earlier. Mount the 470µF supply bypass caps (with their 1µF bypasses) as close to the BUF-03s as you can to ensure a solid path to ground for any noise. And even though they may look more "professional," printed circuit boards degrade sonics; hard-wire everything together with high-quality wire and you'll get far superior sound.
Resistors R5 and 6, 10 ohm Holcos, are optional. John Curl recommended them to me in light of the BUF-03's tendency to ring at around 30MHz in the presence of highly capacitive loads, and the resistors are there to isolate the load and minimize the danger of ringing. Alternately, Mike Moffat uses the BUF-03 without isolation resistors in his Theta units (footnote 6), tying the BUF's output lead directly to the output jack. I initially elected to go with the resistors, but an interesting situation recently arose when I got my hands on a pair of the new Grado headphones. Although my preamp seemed like a natural to drive the 'phones, the resulting sound quality was pretty underwhelming, not at all what I'd heard from them with other gear. I couldn't for the life of me figure out why until I learned that the Grado headphones have a 40 ohm impedance; with the best 10 ohm Holcos on the outputs, my preamp had a 12 ohm output impedance, hardly the best match for the Grados.
So I removed the resistors, tied the outputs of the BUF-03s directly to the RCA jacks, and Dios mios! Now the Grados came alive; the bass became leagues tighter, more powerful, and the smeared quality I'd heard throughout the upper midrange to lower treble was gone, replaced with the extreme transparency I'd heard on previous occasions (footnote 7). Even the sound over my Spicas/Muse Model 18 was clearer, with an audible increase in detail. So...even though John Curl has a point about the potential for ringing, I recommend leaving the series output resistors out unless you plan to use very long, high-capacitance cables.
As you may have noticed, I didn't include a balance control in Aunt Corey's Homemade Buffered Passive Preamp. Personally, I've never needed one, but if you do, try using two mono pots for volume instead of adding a dedicated balance pot to the circuit.
Trimpots VR3 and 4 are there to null out any residual DC offset from the BUF-03s; after letting the unit warm up for an hour or so, measure the offset with a voltmeter set for DC volts on the lowest scale, with the red probe contacting the signal side of the RCA output jack and the black probe touching ground. Tweak the trimpot until your meter reads 0V DC, and repeat for the other channel. You may want to check the offset again in a few weeks to make certain it hasn't drifted, but if you use high-quality trimpots like the specified Bourns unit this shouldn't be a problem.
Make sure you attach heatsinks to the little BUF-03s, as these class-A buffers run real hot. You can find suitable TO-99-size heatsinks at most electronics huts including Radio Shack.
And lastly, use either Wonder Solder or the identical Ersin SN62 (available at Radio Shack, albeit in the tiny-tunes gauge, part #64-013) for all soldered connections; solder is audible, and Wonder/SN62 sounds better.
Good luck, and happy listening!
Note that the homemade buffered preamp's apparent simplicity is deceptive when it comes to its actual construction. If you have never built an electronic component before or if you have no soldering experience, successfully completing this project will be difficult. At the minimum, you will need a good 15-25W electric soldering iron, cable-strippers, needle-nose pliers, a small side-cutter to trim leads and wires, normally-closed tweezers to clip to the IC leads while you solder, a solder-sucker to undo erroneous connections, a multimeter, a steady hand and eye, and a willingness to learn from your mistakes. I also recommend buying one of those "third-hand" part holders from Radio Shack.
Part of the fun in constructing any electronics project is the hunt for the specified components or suitable alternatives. A minimum list of suppliers is mentioned in the sidebar; Stereophile cannot help readers obtain parts for Corey's buffered preamplifier. Nor can we help with faultfinding or repair of completed preamps that fail to work. However, Corey welcomes feedback from readers who have successfully constructed this design, particularly if they've achieved sonic success with different parts or cables, or even done something completely gonzo such as use a separate power supply for each channel. Write to him at Stereophile, P.O. Box 5529, Santa Fe, NM 87502, Fax: (505) 983-6327, with "Homemade Preamp" written prominently on the envelope or Fax cover sheet.---John Atkinson
Footnote 6: In a conversation I had with him, Mike informed me that the Thetas have provisions on their circuit boards for output resistors, but he doesn't include them unless there is a real problem with a specific system/cable setup; ie, very long, highly capacitive cables causing the aforementioned ultrasonic ringing.
Footnote 7: To drive the Grado headphones with my preamp, you need to make up a cable for them: use a high-quality four-conductor microphone cable like Canare, VTL, or give Joe Grado a call and order some of his. You'll also need a pair of RCA plugs and a good stereo phone jack like the Switchcraft. On the preamp end, strip away the outer jacket and cut away the exposed braided shield entirely; at these low impedance levels, shielding is unnecessary, and you don't want the added capacitance it adds to the cable anyway. On most 4-conductor mike cables, the four legs are spiral-wound with color-coding of white-blue-white-blue; what you want to do is use the two white legs for signal R, and the two blue for signal L. Avoid using two directly adjacent legs for each signal side, as this raises the inductance of the cable; pair each leg with its opposite-side mate. At the headphone side of the cable, strip the outer jacket away again, along with the braided shield, and, with the help of a multimeter set for "continuity," determine which of the four conductors are signal L and R, and which are ground. On the phone jack, the solder lug that corresponds to the tip contact is always left, so solder the blue signal leg to tip, the white signal leg to "ring," and both ground legs to the common ground lug.