The Latest Improved U4
Now, this may be the main theme of iPhone Adventure follow-up, the sound quality improvement based on Honda-san's circuit analysis results of Douk Audio U4's final stage.
It would be a natural idea for me to consider the final stage as a Tube Adventure, as it has such a large impact on sound quality, and I thought that the analog approach that I had been involved with for many years would result in an effective sound version upgrade.
What surprised me when I saw this circuit diagram is that tube in the output stage is not a cathode follower, which is normally used in the final stage. This is because, although the cathode follower has almost no gain, it can reduce the output impedance to a sufficiently low value.
However, this use of tubes was equivalent to the first stage of a so-called RC coupling circuit, which is commonly used in tube amplifier circuits, where the vacuum tube is used as a plate follower. Even though it is possible to amplify the voltage to a certain extent, for example, the plate resistance (Rp) of the ECC83/12AX7, which has a high amplification factor (μ): 100 and is often used in Model 7, is about 60 to 80 k ohm.
Therefore, as a normal audio device, the output impedance is too high to be able to drive a power amplifier in the next stage. Therefore, it turns out that the 6N3 (5670, 396A), which was an unfamiliar tube, was selected because it has a heater voltage of 6.3 V, which makes it easier to increase the voltage, and a plate resistance of about 6,400 ohm.
Another thing I noticed was the fact that the output of this single tube goes through a coupling capacitor, but it is just output as is, and it is non-feedback circuit.
Generally speaking, it is true that feedback circuits in tube amplifiers have positive aspects such as improving high-frequency characteristics and lowering distortion, but also negative aspects such as making the sound quality too homogeneous, and worsening the rise of the sound.
The non-feedback circuit was the reason why I had such an interesting Tube Adventure in which I could clearly hear each sound or tonal character of each tube. I also felt that the lively musical expressiveness of Douk Audio U4 was largely due to this non-feedback circuit.
Well, I would like to introduce to you the latest version of the Improved U4, which I have managed to arrive at after more than a year of trial and error (above).
Here are the detailed and specific explanations of the changes or the parts replacements from the original.
(1) Replacement of the coupling capacitor between the plate of tube and the RCA output terminal with the capacitor that seems to be the best at present:
The original U4 uses a very small and red WIMA 1.5 MFD film capacitor, and the sound quality is fairly good without any peculiarities. However, what is required of a signal capacitor for coupling purposes is how large the amount of transmitted information is, although there is a certain correlation with the measured value of ESR (equivalent direct current resistance), it seems that the only way to confirm the amount of transmitted information, is through my own hearing.
In short, when you hear a sound that has a really large amount of transmitted information, whether the low-frequency volume is rich or poor, or whether the high-frequency extension is good or not so good, such audio expressions would never come to mind. You only feel real music itself.
From that perspective, the main capacitor I chose was Jupiter Condenser's Red Comet, an aluminum foil PIO capacitor (in mineral oil), 1 MFD/400 VDC.
You can download my honest impressions HERE, so please refer to it. Best of all, the tonal contrast across almost entire range is higher or clearer than ever, making the music feel alive and real to us. This is also a common impression among my close friends.
I also connected an Ultimate Ruby Mica, 0.016 MFD / 500 VDC in parallel to this Red Comet. The addition of this capacitor allows for vast spatial expression and a clear three-dimensional sound image, just like adding on an excellent super tweeter system to the existing speaker.
(2) Connect the metal film resistor (10 k ohm, 1 W) in parallel to the existing cathode resistor (chip resistor: 1.5 k ohm) of the tube:
Since the signal current flows directly through the cathode resistor, the sound quality depends on the quality of the resistor itself, as I mentioned in My Adventure with My Old Marantz Model 7, Part 10 HERE.
I had to abandon the task of replacing the existing chip resistor with a higher quality resistor because it was too risky for my poor technique. What I came up with was to solder a 10K ohm best quality metal film resistor to the 2nd and 8th terminals of the socket, and essentially connect it in parallel with this chip resistor, 1.5K ohm.
This would reduce the cathode resistance to about 1.3 k ohm, but the cathode current would increase by about 10%, so I thought it would be beneficial in terms of sound quality.
Fortunately, I was able to obtain one NOS resistor from IRC, International Resistance Company, which unfortunately no longer exists, but it was an RN70 (1W) 10k ohm, the best quality resister I had ever experienced.
Of course, chip resistors are not completely useless; as they do not have the inductance component that wire-wound resistors have, and my general impression is that they produce a clear sound in the mid-high range.
By the parallel connection with this NOS RN70 I wanted to achieve a rich and solid sound even in the mid-low range and below. This is because the frequency range from 100 Hz to around１k Hz is the most important band where musical energy is most concentrated.
(3) Strengthen the decoupling capacitors for the plate power supply and another circuit:
A decoupling capacitor is also called a local power supply in Japan, and I improved the sound quality by strengthening this capacitor, by increasing the capacitance, or by connecting multiple capacitors of different types in parallel to reduce over a wide range of ESR.
It means the same thing that the reducing the output impedance of the power supply as much as possible over the entire range. There are two small 10 MFD electrolytic capacitors mounted on the original board, but this time I left them and connected a rare NOS Mallory 100 MFD / 250 VDC large size electrolytic capacitor in parallel.
In addition to simply increasing the capacity, this parallel connection also reduced ESR. Also, this Mallory, which was produced almost half a century ago, boasts extremely low measured ESR values at 100 Hz, 1 kHz, and 10 Hz.
Furthermore, I have used this Mallory as the cathode capacitor of the output tube of Alan's PX25A (DA30) mono SET, and so I have already confirmed the sound with excellent quality, full of richness, but also powerful.
In the case of suitable and simple analog circuits, the inherent sound quality of the parts and their combination will determine the final sound quality, I’ve experienced.
However, sorry that electrolytic capacitors with such excellent physical characteristic and sound quality cannot be found in the current products and markets.
Well, the length of the lead wire I used was short, abut only 10 cm, but I wanted it to be of the highest quality possible, so I chose the following: Tinned Copper Wire Lacquered Cotton Insulation - Jupiter Condenser HERE.
This wire seems to have been developed using the famous Western Electric cable as a model, but the company's president, Cristopher-san, is particular about using the highest quality tin plating.
I have long highly praised the WE16GA, which is one of the excellent speaker cables, and the Belden 8402, which is an old microphone cable that makes for great interconnects.
I have also considered the quality level of the tin plating to be one of the major factors. Now I take my hat off to Christopher-san for his efforts in questing for the purity of tin itself and finding a higher quality tin plating by focusing on the skin effect of the plating.
Therefore, I used Jupiter 20 GA single wire around the coupling capacitor and Jupiter 16 GA wire for the decoupling capacitor.
Please go to the next page.
