Wet Tantalum Capacitor and Electrolytic Capacitor
It was also in "MJ" magazine that I saw an article about an experienced self-made amplifier enthusiast who described the sound quality of "Wet Tantalum Capacitor" (or "銀タン" for short in Japanese) was far more excellent compared to the usual electrolytic capacitors.
It might be nearly half a century ago. Of course, at that time DigiKey and Mouser Electronics didn't exist, and I didn't know how to get them, which was developed for the space exploration by NASA, I heard.
I had been also informed, it was extremely expensive, so it had been a long time away for me. However, its sound quality is extremely good and has been engraved in my memory for a long time.
It wasn't that long ago that I finally experienced the sound of this “Wet Tantalum Capacitor”, about 20 years ago now.
I learned about the existence of an old parts shop in Japan, and thankfully I was able to obtain this capacitor at a very reasonable price. In addition, this shop handles a wide range of parts, resistors and capacitors, mainly for the US military, and has greatly expanded my experiences and knowledges in this field since then.
First, I tried using it as a cathode bypass capacitor for the WE310A, the first stage tube of my PX25A non-feedback stereo single amplifier. A comparison with electrolytic capacitors, which had been considered to be of relatively good quality up until then, showed a very clear difference.
In short, it has a very clear sound, rich in contrast and a glossy tone. Also, I couldn't help but marvel at the three-dimensional expression that allows you to see the front and back of the sound space.
It was certainly a shocking experience for me at the time. Encouraged by the remarkable results, I also replaced the first stage and the second stage cathode bypass capacitors, C62 and C64 in the circuit diagram, of my Model 7 K phono amplifier with this “Wet Tantalum Capacitor”.
The results obtained here were equally impressive, and fortunately, I later had the opportunity to make this modification for four Marantz Model 7s, which were Handoko-san’s from Indonesia, Alan-san’s and Roy-san’s from the US, and Ryszard-san’s from Poland.
After replacing the bypass capacitors with “Wet Tantalum Capacitors”, I think that the improved realistic expression of the analog LP playback sound with our - Alan-san, Roy-san, Nishikawa-san and I - most favorite MM cartridge, the Shure V15 Type 3 + VN35MR, reflected the remarkably positive character of this capacitor.
I suppose the reason that the richness of transmitted information in the mid-high frequencies (especially) of the “Wet Tantalum Capacitor” enabled such a more realistic reproduction, despite the roll-off of high-frequency response by the unique K-K feedback RIAA equalizer of Model 7.
In short, it can be said that the replacement with a “Wet Tantalum Capacitor”in the Model 7 phono amplifier had brought a desirable result both from a somewhat qualitative point of view, and in terms of actually improving the audible ability. About the “Wet Tantalum Capacitor”, please refer to this article in Wikipedia.
Well, for the past 7 or 8 years, I have been using a handy LCR meter, to measure the capacitance, Tan δ, and ESR of almost all capacitors that were used in my own studies, or that were installed in the amplifiers, which I modified and built up for a few eager readers of my articles on Jeff’s Place.
This simple LCR meter can measure such values at each point, 100 Hz, 120 Hz, 1 KHz, 10 KHz, and 100 KHz. As a result, for example, I have learned from the experiences that the measured ESR value of electrolytic capacitors is closely related to the actual audible characteristics. Well, I wrote something like this about electrolytic capacitors in “My Adventure with My Old Marantz Model 7” Part 8 HERE.
“I have had so many discussions about my experiences with electrolytic capacitors, but I hadn't yet organized well the stories to tell you about, but as I said before in my preface, “I often think and wonder that if the electrical characteristic of the usual electrolytic capacitors were close to an almost ideal level, such as with a well-built mica capacitor, how much we could fully enjoy the music playback with more realistic and organic sound by the audio devices.”
However, in order to honor the reputation of electrolytic capacitors, which are probably the most widely used capacitors in the world, I’d like to clarify my thought about the merits or advantages of electrolytic capacitors that cannot be replaced by other capacitors.
First of all, there is no capacitor superior to the electrolytic capacitor in terms of the capacity to store more electric charges. In addition, making use of this characteristic, electrolytic capacitors are mainly used for rectification in power supply circuits, decoupling of various power supplies, and cathode bypass capacitors, which is the theme of this time, in vacuum tube amplifiers.
The advantage of using electrolytic capacitors in the tube amplifiers is not only that they have a large capacity, but also that the ESR in the low range, for example around 100 Hz, easily shows a value of 1 ohm or less.
I have never seen such a low value in any other capacitor. This low frequency ESR value is effectively the output impedance of the capacitor, and so it can be considered that lower ESR value could bring out the richer low frequencies. Well, please see Table 1.
These capacitors are the ones I actually used for decoupling and bypass capacitors. Among them, only the blue PVC Sprague ATOM TVA 1906, 20uF /500 VDC would be still available, but the Mallory 30uF /450 VDC is really hard to find now, made in the 1960s for the US military.
Let’s compare the ESR and at 100 Hz, 1K Hz between these two, ATOM; 1.8 ohm, 0.9 ohm, Mallory; 0.9 ohm, 0.49 ohm.
You should be able to see Mallory's superiority right away, but most important thing would be actual audio hearing differences between the two.
It is true that the band from 100 Hz to 1 K Hz is also the region of the fundamental tone, and most of the musical energy is concentrated, so it seems that some difference in physical quantity will be able to appear in the sound quality difference.
Actually, when I replaced the Atom with the Mallory as a decoupling capacitor for the plate power supply of WE310A in my PX25A stereo amplifier in the past, I could hear the clear improvement in sound quality, or rather the richness of the musical relief.
In other word, the soft and rich low frequencies filled my sound field, creating an indescribably organic atmosphere. Other than this, I also tried the Atom in other situations, but the impression that its sound character was a little solid did not change.
So I concluded that it was not due to mere capacitance difference of 20 μF and 30 μF, but due to low ESR or low output impedance of Mallory.
So my obsession with vintage Mallory grew deeper and deeper. And you will also see vintage Mallory 100μ / 350 VDC in a paper tube.
Sorry, but we can’t find it anywhere at present, but I was so astonished this lowest ESR value.
By the way, this splendid Mallory was installed as a cathode bypass capacitor of the power triode in my friend Kajiwara-san's PX25A mono SET, just two years ago.
The electrolytic capacitors in this paper tube are probably from early or before 60s. I also experienced the earliest Atom in an orange colored paper tube about 15 years ago. I remember well that the sound of this old Atom was impressive.
What I'm trying to say here is that the old electrolytic capacitors in paper tubes might be rather superior in terms of the characteristics and the sound quality compared to modern ones, and I proceeded in my thinking like this.
Recent capacitors have explosion-proof construction for safety measures and regulations, and the PVC outer skin that is used by shrinking is thicker than the old Mallory.
And, depending on the case, I came to the hypothesis that this is causing the deterioration of the sound quality.
Then I decided to cut the blue PVC sheet off this Atom TVA1906 and made it naked. As a result, at 100 Hz, 1 K Hz, and 10 K Hz, ESR improved by 21%, 26%, and 27%, respectively. And also, the loss tangent, Tan δ, also improved by 0.4 degrees, 2.6 degrees, and 8.2 degrees.
After that, I stripped off the other four types of the PVC outer sheets on another Mallory, a Roe made in Germany (with excellent characteristics and sound quality), and some more recent capacitors. I then measured the characteristics, ESR and Tan δ, before and after.
The differences were not as large as that of Atom, but all the values were improving.
In addition, there was virtually no difference in the low range, 100 Hz at before and after in the Mallory and the Roe, because of the thinner PVC outer skin, I suppose.
In any case, from the point of view of characteristics or sound quality, it seems that it is not desirable to tighten the aluminum can, which is also the negative electrode of the electrolytic capacitor, with the shrinkage of the PVC film.
Needless to say, I would like to make it clear that I am not recommending stripping the PVC skin of capacitors to Jeff’s Place readers from a safety standpoint - it might be dangerous as an explosion hazard. If you try this, the process and results are strictly your own responsibility.
Well, what I would like to argue in this section is the fact that it is becoming increasingly difficult to obtain high quality or desirable electrolytic capacitors in the market, especially for tube amplifiers.
Requirements for electrolytic capacitors are likely to become more stringent, such as detoxification of electrolyte components, thorough safety, and stricter explosion-proof construction.
I also understand that it is difficult to develop new electrolytic capacitors for the professional manufacturers, that are specialized for tube amplifiers, due to sluggish demand.
Sorry, but I have to admit that this is a major stumbling block in trying to build a particularly simple and high quality tube amplifier in this day and age.
As I might said before, the final sound quality of an amplifier with a simple configuration decisively depends on the audible quality of only one or a few passive parts.
Well, now back to “Wet Tantalum Capacitor”, and let’s study the ESR matter of an electrolytic capacitor, I just mentioned above.
Now, let's take a quick look at the physical properties of “Wet Tantalum Capacitor” from Table 2.
First, it should be noted that this test sample is part of what I have used so far, and the measurement time spanned a long period of time, so I would like to say that it was not necessarily measured under the same conditions.
Therefore, rather than comparing the details of the measurement results, I would like to see what kind of characteristics or trends they show. In addition, these wet tantalum capacitors are NOS products, and I suppose that they might be stock disposal products of Sprague in the old days.
The first thing I notice is that the capacitance value above 10 KHz is more than halved compared to 100 Hz and 1 KHz. Furthermore, the value of tan δ deteriorates rapidly in synchronization with this.
In short, compared to around 100 Hz, the higher the frequency, the more the original character of the capacitor changes completely. Also, perhaps due to structural restrictions, it seems that the physical characteristics of small-capacity, high-voltage capacitors are superior to those of 100 µm or more, large-capacity, low-voltage capacitors.
Honestly, I took a closer look at these physics and was completely confused, because I praised the sound quality of the wet tantalum capacitor in the previous section.
What's more, how could the physical data be related to the excellent mid-high range, where I could really feel the dense overtones?
Certainly only ESR maintains a relatively good value even at high frequencies. So sorry, but I can't qualitatively describe or find out the secret of the excellent audible quality in the Wet Tantalum Capacitor yet.
Going back in time, in the cold season of early 2022, when I was soldering the wiring and parts in the final stage of the PX25A mono SETs, I was concerned about this gap between the actual audible quality and the physics of this wet tantalum capacitor.
The sound of the two PX25A mono SET was a familiar one for me, but I was beginning to feel that the tone was slightly biased towards the mid to high frequencies, but maybe I was just seeking for a new type sound.
At that time, what came to my mind was the idea of connecting in parallel to this wet tantalum capacitor with an excellent electrolytic capacitor, such as Mallory or Roe, which have satisfied rich low frequencies.
Of course, I thought that this modification would reduce the ESR value over a fairly wide band and increase the capacity to be able to enhance the low frequency range.
Fortunately, strengthening the cathode bypass capacitor brought out a truly substantial sound quality improvement. I was amazed at this remarkable positive change that exceeded my expectations.
However, at times like this when I was in a state of excitement, I should have calmed down, as I wanted to recognize the change from an objective point of view.
On April 22, 2022, I had the opportunity to have the final hearing session of Craig-san's and Ryszard-san's PX25A (DA30) mono SETs with Nishikawa-san, a close friend of mine, just before the shipping.
About his impression, I would like to translate his email, sent to me the next day, so as to be as accurate as possible so that I could convey what he really felt.
Nishikawa-san said, “Yesterday, I had a chance to listen first Craig-san’s amplifier, which added a Mallory to the first-stage cathode bypass capacitor. Even before the addition of the bypass capacitor, I was so impressed, but with the addition version, I really feel deep density tone and the beauty of the tone gradation at the same time. I was also amazed at the realism, such as the strength of the transparent mid-low range. After lunch, I listened to Ryszard-san’s amplifier, which was freshly made just before the shipment to Poland, with WE310A mesh. It's hard to describe in one word, but the room was filled with highly refined and pleasant music that is irresistible to music lovers, I suppose. Well, I also felt a new aspect of the WE310A mesh.”
Certainly, the addition of just one capacitor has revealed a truly rich musical expression before us.
