Kits for CA-801 Vacuum Tube Headphone Amplifier
Create your own sound: The greatest deal, the most fun, and the best collection
If you would like to find a headphone amplifier under $1000, which can manage high impedance headphones like HD800 or T1, I believe CA-801 will be the best deal out there!
Design and Build Quality
The input stage has a triode single-ended structure with transistor constant current source at one end, which provides good linearity and enough amplification. The output of this single-end stage is directly connected to the PowerMOS Q2 through a emitter follower. This provides enough driving capability to headphones. By using direct coupling, we not only reduced cost, but also eliminate quality issues (for example: leakage current, DF problem) from coupling capacitors. However, since it is direct coupled, the instability of DC is something we need to take into account.
PowerMOS has similar characteristics of a vacuum tube, which has high bandwidth with over 300 MHz and high input impedance. Adding an emitter follower in front of the PowerMOS can decrease the input capacitance loading-effects. The Ciss of PowerMOS IRF610 is about 140 pF. By inserting this emitter follower, we can improve the bandwidth significantly.
Measurement shows that the bandwidth was tremendously increased with our implementation. With high bandwidth and low harmonic distortion, the amplifier was able to reproduce full range melodies and the original vocals or music played by instruments that come out from the headphones.
Adding a resistor on the emitter side of transistor (Q1) solves the current imbalance problem from the two transistors. The two transistors (left and right channel) share one 13V Zener diode. This Zener stabilizes the voltage to 13V, which is necessary for the tube filament (12AU7 filament is 12.6V; its current measures about 150mA). Adding a parallel capacitor next to the Zener diode allows the capability of reducing breakdown noises from the Zener diode.
This amplifier is designed to use transistors with Zener diodes rather than regulator ICs. And the slightly feedback network is added, this does not only reduce the level of distortion and increase bandwidth, it also greatly reduces the out-put impedance. Without the feedback, the gain variance could easily go over 1dB. The tests show that the distortion generated by one-side of the tube is barely detected, which creates a vast difference from that of a regular tube amplifier. A non-feedback tube amplifier generally has large harmonic distortion than feedback amplifier. Moreover, the gain variance between the two channels is also within 0.1dB, while the bandwidth is over 200kHz!
Fig2. The output is reduced by 17 dB when it’s at 100KHz in comparison to at 1KHz.
Fig 3. When feedback is added, the output is only reduced by 1dB when it’s at 100KHz.
Fig. 4 When both feedback and an emitter follower (buffer) are added, the output is only reduced by 0.1dB when it’s at 100KHz!
Fig2, Fig3, and Fig4 show the simulation (please pay attention to the difference in ratio at Y-axis) that has non-feedback, with feedback, and feedback with addition of an emitter follower, respectively.
Components
Let’s start with the input terminal. VRs from ALPS Audio grade are used at the input terminal, as they are some great minions to reduce noise.
WIMA, 2.2uF/50V, the input coupling capacitor, is famous for its delicacy and exquisite appealing.
The choice for transistors, without any hesitation, only goes for “made in EU” or “made in US”, as they are known to produce vigorous sounds.
IRF610 is used in the PowerMOS. In terms of products from Vishay, IRF610 has a lower Ciss than either IRF510 or IRF710. Having a specification of 200V and 3.3A is more than enough to manage the system.
The output current source uses MJE15030 from On-Semi. This is a fantastic transistor—but it’s by no means cheap.
2N5210 and 2N5087 have been selected. These two types of transistors are recognized to produce incredible sounds with low noise, high linearity, and high ft (over 100Mhz when it runs at 1mA).
Zener
A better specification zener diode from On-Semi is selected—having the lowest impedance (2.5 Ohm).
The output coupling capacitor adopts the FM series from Panasonic. It’s great in its characteristics—a long life expectancy (7000Hrs@105C), ultra-low in high frequency impedance（18mOhm@100kHz）, which is only 1/2 ~1/3 of a regular capacitor, and extremely high ripple capability. The bypass capacitor for the cathode also uses the same series.
The power input capacitor uses Nichicon products, designed specifically for Audio. The Nichicon capacitor (3300uF/50V) shows a high capacity (the higher the capacity, the lower the ripple noise), high voltage appliance, and high quality; such a capacitor seems to be the only choice to fit in a small headphone amplifier.
The capacitor in the passive component uses the capacitor made in Germany or Japan, to ensure its high quality and to avoid heaviness.
There are 3 de-coupling capacitors in the circuit. Considering the overall volume and high frequency impedance, ceramic product is selected—in fact, a good ceramic capacitor is costly. The filament in the vacuum tube is also decoupled with a capacitor that has a good performance in high frequency.
The collection of components is selected without taking the cost into consideration as the top priority. Clover Audio holds the belief and determination to make the best products.
Selection of the vacuum tube
Many well-known brands make 12AU7(ECC82). This gives audiophiles some flexibility and more importantly, the fun, to try different vacuum tubes. The fascinating part about vacuum-tubed amplifiers is that the sound is different when a different tube is used. 12AU7 from JJ, Electro-Harmonix, Genalex Gold Lion, SIEMENS(E82CC), TELEFUNKEN all perform a decent job. The choice over the tube should be based on personal preferences and how deep your pocket is J. The one made by Black Sable Mullard is surely one of the most expensive 12AU7 you can get!
The sound is more cohesive and distinctive!
In addition to the higher density and resolution, coherence in between each beat has also been improved noticeably, especially when brass instruments are played. On top of this, the resolution seems to be much higher after hooking up this amplifier; I am able to get more details that I didn’t notice before. It’s quite peculiar I have such a feeling; as we know, an amplifier does not create more than what the source has. It does make it strange to believe the resolution has been increased.
I try to listen it very carefully and give this feeling a second thought. I think this perception rises because the music has been altered into a more distinctive rhythm, rather than cohesive unity, that each beat and note have stand out individually. This seems to enlarge all details in music and simultaneously increase the contrast. The strange thing is that such increase in contrast does not bother listeners; rather listening to more details seemingly hallucinates listeners to believe the music has been played with a greater resolution.