My preveous phonostage was one of my first tube designs. Since I built it I have learned of ways that it could be improved. This design incorporates many of those improvements. It also lets me experiment a bit.
A phonostage does two things. It amplifies the output of the cartridge, and it reverses the pre-emphasis curve of the LP. A moving magnet cartridge has a rated output of about 5mv, while most power amps need 1-2 volts for full output Also, when an LP is recorded, the bass frequencies are attenuated, and the trebel is boosted to save space and combat noise. A phono stage reverses this pre-emphasis with an equalization network.
If you have looked at many tube phonostages on the web, you may notice that this one is unusual in that it uses a pentode. I chose this tube mainly because I had several of them on hand. Also the specs seemed to indicate it would work well for a phonostage.
Compared to triodes, pentodes have a very low input capacitance. This seems to be an advantage for a phonostage. Pentodes also have much higher gain than triodes, useful if you want to keep the design down to two stages.
On the down side, pentodes are less linear, have a much higher plate impedance, and tend to be noisier than triodes. In the first stage of a phonostage, distortion shouldn't be much of an issue since the signals are small. Like with triodes, pentode's distortion tends toward zero as the signal diminishes. With careful design, the high plate impedance could be an advantage, and the equalization network will take care of at least some of the noise.
The equalization network consists of the .032 uF and .094 uF capacitors, and the 3.4 kOhm resistor. The 25 kOhm pentode load resistor is part of the network as well. Since the pentode's plate impedence is so much higher than the load impedance, the output impedance is quite stable even if the plate impedance varies some. This allows me to include the output impedance as part of the network, eliminating a resistor.
The output stage is fairly ordinary except for the way it is biased. The cathode is connected directly to ground, and the grid resistor is much larger than usual. This is called a grid leak bias. The grid always captures a few of the electrons released by the cathode. If the grid resistor is large enough, the electrons can give the grid a negative charge. This charge can bias the tube as long as the signal coming in is not too large.
Most of the improvement over the last phonostage involve improvements in the quality and placement of the capacitors in the RIAA network. The new capacitors are simply Sprague Orange Drops; not especially exotic, but in my limited experience they work better in tone circuitry than metallized film caps. They are placed before the coupling capacitor, rather than after it, so they are biased at about 140v. I have heard this can affect a capacitors sonic performance. The one tube related improvement is these tubes draw more current than the preveous tubes. This seems to make it easier for them to drive the RIAA network, volume control, crossover and other downstream loads.
In spite of the pentode front end, hiss is not too bad. It is noticable at full volume, but not at normal listening levels.
This phonostage has a wider bandwidth than the preveous one. Highs are clearer, and lows are more powerful and deeper. It also has at least 10dB more gain, and more overhead. Loud passages of music are cleaner sounding than before.