All-Tube Preamplifier

Tubes are quite strange, they do something to recorded music that I can't quite explain scientifically, but I will say that to have a genuinely high-end sound system, you need tubes somewhere. Personally, I don't think there's any one thing about tubes that makes them sound better than transistors, I think it's a combination of higher operating voltages and a space charge of electrons inside the tube that provide a faster response to transients, the tendency to "round-off" sharp transients that would clip on a solid state amplifier, and the all-around higher build quality due to the use of more rugged, higher voltage parts.

I started this project way back in 2006, but I didn't really finish it until early 2014. Every time I'd start working on it I'd come up with a new idea, which would require me to order some more parts, and by the time the parts got there I had already sidetracked onto something else. I have a very short attention span, you see.

The amplifier employs five tubes, a 12AX7 and 6DJ8 provide amplification to line level and RIAA equalization for a standard 47kΩ moving magnet phono cartridge, while a pair of 6DJ8s comprises the line stage and reduces the output impedance enough to drive a power amplifier.
A single 6BY6 double diode serves as a rectifier/voltage doubler. You could omit this tube and replace it with a pair of silicon diodes (type 1N4007) but it wouldn't look as cool, and it seems like tube rectifiers are more effective at filtering out noise than their silicon counterparts. If you opt for silicon, you may also want to add a delay timer of about 10 seconds to give the signal tubes time to warm up before the high voltage is applied to avoid premature wear (cathode stripping).

Materials

I constructed mine point-to-point on perf-board. The phono and line stages are on a single board with ceramic PCB mount sockets for the tubes. All of the capacitors are high quality Japanese-made Panasonic polyester film types. I also used some Japanese made Nichicon and American made Cornell-Dubilier electrolytics. All of the resistors in the phono stage are low-noise +/-1% metal film types, while the other sections should be OK with standard +/-5% carbon film. All of the interstage coupling capacitors are French made Solen caps. I used 62% lead, 36% tin, 2% silver solder with a brush-on rosin flux. Use the finest wire solder you can get, I used 0.022", and be sure to use an adequate soldering iron, prefereably 100W.

General Notes on Construction

I would build this into a well ventilated steel enclosure, the heavier gauge the better. The phono stage has such high gain it tends to pick up strong radio stations if it's not properly shielded. V5 may also need a shield depending on how much external interference you have. The power supply is a separate board, I situated it, and the power transformers as far as I could from the phono stage to reduce noise, and as an extra precaution I installed an alumiminum shield around it. The chassis is a standard ventilated steel box, which was a Dish Network model 2700 satellite receiver in a former life (I have about 10 of them, and they're obsolete so there's nothing else I can do with them).

Since the L7806 voltage regulators I'm using for the heater supply have a design rating of 1A @ 6V, you would think you could get by with two regulators running two tubes each, since the 6DJ8 and 12AX7 both draw about 0.3A each. Unfortunately, things don't always work out so well. Heating elements like the ones in tubes present an almost dead short when first powered on cold, without proper cooling the regulators overheat and shutdown before the tubes have warmed up. You have two options, you can either give each tube its own regulator, or do what I did and bolt chunky heatsinks to your regulators and install a small 12V fan at the back of the chassis. I chose the fan because it's cheap (about $3, or free if you rip it out of an old computer like I did), doesn't make a noticeable amount of noise, and has the added benefit of keeping the other heat generating parts (resistors, tubes, transformers) cool as well.

As built, the phono stage tubes V4 and V5 are only powered on when the amp is in phono mode (S2 is on) to prolong tube life. S4 should be a high grade make-before-break type (usually called "Form D" or sold as a "telecommunication relay") to avoid hearing a loud POP every time you switch sources. If you only plan on listening to the phono stage, and don't need an external auxiliary input, you can omit switch S2 and relays S3 and S4, just connect V4 and V5 directly to their heater power and high voltage.

Grounding

Pay extremely careful attention to the way you ground this! The number one source of noise in a high gain amplifier is the low hum introduced by ground loops. Capacitors C13 and C14 are essential to AC couple the heater and high voltage grounds with the mains ground, without these capacitors you will have a massive increase in hum. I suggest using a stiff copper wire, 14 gauge or larger, as a grounding bus. Bend the wire so that it follows the outside of the board and solder any component needing a ground connection directly to it. Each "section", power supply, preamp, phono preamp, should have its own bus, which should be connected at a common point. In my design the common point is a copper ring I made out of 8 gauge wire, bolted to the chassis with a conductive screw. The connectors for audio in/out should not have their metal shields touching the chassis, for panel mount jacks, mount them on a non-conductive plastic or phenol plate, then screw that to the chassis.

Download Schematic (PDF, 577KB)