Capacitors (CAP) are important components in audio systems. They have different voltages, currents and form factors. In order to choose which capacitors are best for sound, moderators need to understand all the CAP parameters. The integrity of the audio signal largely depends on the choice of capacitors. Therefore, when choosing the right device, you need to consider all the important factors.
Audio CAP parameters are specifically optimized for high-performance applications and offer more efficient audio channels than standard components. The types of capacitors commonly used in audio applications are aluminum electrolytic and film CAP, and which capacitors are best for audio in a particular environment depends on the circuits and devices used: speakers, CD players and musical instruments, bass guitars, and others.
History of the sound capacitor
The capacitor is one of the oldest electronic components. Electrical conductors were discovered in 1729. In 1745, German inventor Ewald Georg von Kleist discovered the Leyden vessel, which became the first CAP. Physicist Pieter van Mussenbrouck, a physicist from Leiden University, discovered the Leyden jar on his own in 1746.
Currently, the Leyden jar is a glass vessel covered with metal foil inside and outside. CAP serves as a means of storing electricity, and which capacitors are better for sound will depend on the capacitance, because the higher this indicator, the more electricity it will store. The capacitance depends on the size of the opposing plates, the distance between the plates and the nature of the insulator between them.
Capacitors used in audio amplifiers are of several types, such as the common CAP with metal foil for both plates and impregnated paper in between. Metallized paper (MP) capacitors, also called paper-oil capacitors CAP and metallized paper single layer capacitors (MBLP) for audio, which are used in AC, DC and pulsed current circuits.
Later, Mylar (polyester) and other synthetic insulators became more common. In the 1960s, metal CAP with Mylar became very popular. Two strengths of these devices are their smaller size and the fact that they are self-healing. These are the best capacitors for audio today and are used in almost every electronic device. Due to the huge volume of trade and production of these types of capacitors, they are quite cheap.
Another type of CAP is electrolytic with a special design with predominantly high and very high values ranging from 1 µF to several tens of thousands of µF. They are mainly used for decoupling or filtering in the power supply. The most common in amplifier designs are metallized mylar or polyester capacitors (MKT). Higher quality amplifiers generally use metallized polypropylene (MPP).
The best polypropylene capacitors for sound
Products with polypropylene dielectric are distinguished by their durable design, long service life, and operation in high voltage circuits without leakage. The element volume is small.
JB JFGC
JB JFGC audio film capacitor is made of polypropylene and metallized film. The top of the product is additionally covered with a layer of polyester and filled with resin. Used in acoustic systems inside a crossover filter. The capacitor belongs to the budget category, but produces sound no worse than its expensive counterparts. Supports the combined operation of direct and alternating current, which provides excellent harmony with filters in speaker systems.
The device operates at temperatures up to 100 degrees. Operating voltage – from 250 to 1250V. The line is represented by capacitors with different capacitance ratings from 0.047 µF to 36 µF. The dimensions are relatively small (diameter 0.8 mm), suitable for standard-sized acoustics. They have a laconic, stylish body.
Advantages:
- Low cost;
- Wide selection of denominations;
- High-quality sound;
- Minimum deviation from the stated values;
- Great looks.
Flaws:
- Small capacity.
FM MKP
The element is equipped with a polypropylene dielectric. It is made of a film coated with a metallized layer, which is applied by vacuum deposition. This technology ensures high reliability, accuracy of operation, long-term operation, while the dimensions of the capacitor are small. When manufactured using this technology, devices can repair themselves if the dielectric breaks down, practically without losing their properties.
FM MKR provides decent quality sound over a wide frequency range. The main purpose is to use it in acoustic separation filters and cascade audio equipment. Available in three ratings and operate at voltages up to 160V. The average dimensions are 21 mm in diameter, so you need to find a place for it in the diagram in advance.
Advantages:
- Operates in a wide frequency range;
- Inexpensive;
- Great sound;
- Minimum deviation from the nominal value;
- Strength.
Flaws:
- Large size.
MKP Jantzen Cross Cap
Film MKP Jantzen Cross Cap with metallization, coated with epoxy resin, is available in a wide capacitance range from 0.1 to 330 µF. The product is suitable for lamp technology and various acoustic systems. Operates at voltages up to 400V.
The manufacturing process takes place under the full control of the manufacturer in compliance with unique technology, so the capacitor is of high quality, reliable, and durable. The use of excellent materials ensures pleasant sound with high quality transmission without distortion or interference.
Advantages:
- Provides a pleasant soft sound;
- Small dimensions;
- Sufficient capacity;
- Minimum error;
- High quality construction.
Flaws:
- The appearance of the case is quickly lost.
Best radio systems
Component manufacturing technology
CAP technology largely determines the characteristics of devices, and which capacitors are better for sound depends on the class of equipment. High-end products have tight tolerances and are more expensive than general purpose capacitors. Additionally, these high quality CAPs are reusable. High-quality audio systems require high-quality CAPs to provide top-class sound quality.
The performance, or how capacitors affect sound, largely depends on how they are soldered to the PCB. Soldering places stress on passive components, which can result in piezoelectric stresses and cracking of surface-mounted CAPs. When soldering capacitors, you must use the correct soldering order and follow the profile recommendations.
All Dacron audio capacitors are non-polarized, meaning they do not need to be marked with either positive or negative terminals. Their connection in the circuit does not matter. They are preferred in high-quality audio circuits due to low loss and reduced distortion, if the product size allows.
MKC metallized polycarbonate type is practically no longer used. ERO MKC types are known to still be widely used because they have a balanced musical sound with very little coloration. MKP types have a brighter sound and also have a larger sound range.
A little known type of MKV capacitor is a metallized polypropylene CAP in oil. This is the best capacitor for audio because it has more powerful characteristics than metallized paper in oil.
Which capacitor is better for sound?
Charge stores or capacitors are used in many electronic circuits. Audio technology is no exception, so it also works with charge storage devices, which are used in different circuits. But when choosing a charge storage device for audio systems, you need to take into account parameters that can make the sound better. These include the material of manufacture, capacitance indicators, and permissible operating frequencies.
A wide range of different capacitors can lead to confusion for an inexperienced user. To make it easier to choose the right model, our team from the VyborExperta.ru project has developed its top best capacitors for sound. The rating is divided into 3 categories according to the type of dielectric. The best nominee in each group was:
- Film JB JFGC, characterized by a low price, decent sound characteristics, a wide selection of denominations, minimal deviation from the declared values;
- Paper Jensen Nos Aluminum foil, suitable for any audio equipment due to its properties;
- Electrolytic Elna Silmic II, showing stable reliable operation at all frequencies.
We tried to select the best representatives for the rating with good characteristics, customer reviews, and an affordable price.
Quality of passive elements
Capacitors, especially when they are located on the output signal line, greatly affect the sound quality of an audio system.
There are several factors that determine the quality of CAP, undoubtedly very important for audio:
- Tolerance and actual capacity required for use in filters.
- The dependence of capacitance on frequency, so 1 microfarad at 1,000 Hz does not mean 1 microfarad at 20 kHz.
- Internal resistance (ESR).
- Leakage current.
- Aging is a factor that will develop over time for any product.
The best choice of capacitor applications depends on the circuit application and the required capacitance:
- Range from 1 pF to 1 nF - control and feedback circuits. This range is primarily used to eliminate high frequency noise on an audio channel or for feedback purposes such as a Quad 606 amplifier bridge. An SGM capacitor in audio is the best choice in this range. It has very good tolerance (up to 1%) and very low distortion and noise, but is quite expensive. ISS or MKP is a good alternative. Ceramic CAPs should be avoided on the signal line as they can cause additional harmonic distortion of up to 1%.
- From 1 nF to 1 µF - coupling, decoupling and oscillation suppression. They are most often used in audio systems and also between stages where there are differences in DC level, vibration elimination and feedback circuits. Typically, film capacitors will be used in this range up to 4.7 microfarads. The best choice of capacitor for sound and audio is polystyrene (MKS), polypropylene (MPP). Polyethylene (MKT) is a lower cost alternative.
- 1 F and higher - power supplies, output capacitors, filters, insulation. The advantage is very high capacity (up to 1 Farad). But there are several disadvantages. Electrolytic CAPs are subject to aging and drying. After 10 years or more, the oil dries out and important factors such as ESR change. They are polarized and must be replaced every 10 years or they will negatively affect the sound. When designing the signal line electrolyte interconnect circuit, problems can often be avoided by recalculating the time constant (RxC) for low capacitance below 1 microfarad. This will help determine which electrolytic capacitors are better for sound. If this is not possible, it is important that the electrolyte is less than 1 VDC and a high quality CAP is used (BHC Aerovox, Nichicon, Epcos, Panasonic).
By choosing the best solution for each program, the developer can achieve the best sound quality. Investing in high-quality CAPs has a positive impact on sound quality more than any other component.
Best Paper Capacitors for Sound
Paper components are placed in a metal shell to improve durability. Used at different frequencies due to their wide range.
Jensen Nos Aluminum foil
This model is equipped with a dielectric made of oiled paper. Jensen Nos Aluminum foil has an aluminum foil coating on top, which provides a long service life of up to 3000 hours. The combination of materials results in balanced sound transmission at different frequencies.
The main application is for high-end audio equipment, including tube equipment. The frequency range is from 400 pF to 0.082 µF. Works with voltages of 400, 600, 630V, which depends on the selected model.
Advantages:
- Great sound;
- Best workmanship;
- Wide selection of containers;
- Operating time up to 3000 hours;
- Can be used on various audio equipment, including amplifiers.
Flaws:
- Not always on sale.
Jupiter Copper Foil - Paper & Wax
The high-quality element is made of durable, reliable materials. The dielectric is made of wax paper, and copper foil is used as a shell, providing the best sound transmission without distortion.
It is actively used for installation on premium segment audio equipment. The products have a wide range of capacitances from 1 nF to 12 µF, so they can be used on a wide variety of audio equipment. Capacitors are manufactured with an accuracy of up to 5%. Operates at voltages up to 600 V.
Advantages:
- Wide capacitive range;
- Excellent quality;
- Suitable for premium audio equipment;
- High precision sound transmission;
- Copper foil is used;
- Long service life.
Flaws:
- High price.
Alexander by Duelund copper
Quite an expensive option, distinguished by its unsurpassed characteristics. Oiled paper dielectric together with copper foil provide rich sound. Natural timbre of voice and instruments, presence effect, endless space, dynamic bass - all this is provided by a capacitor from Alexander by Duelund copper. Suitable for working on audio equipment, conventional electronics - stable results are ensured everywhere.
An important distinguishing feature is the use of pure oxygen-free silver-plated copper for the product leads. It additionally provides high-quality sound without interference, as well as a long operating time. Operates at high voltage up to 900V, capacitance error does not exceed 10%.
Advantages:
- Ensuring good vibration isolation;
- The conclusions are made of high quality material;
- Convenient design;
- High sound quality;
- Operates at voltages up to 900V.
Flaws:
- High price.
The best mixing consoles
Testing CAP elements for applications
There is a general understanding that different CAPs can change the audio quality of audio applications under different conditions. Which capacitors to install, in what circuits and under what conditions remain the most discussed topics among experts. That is why it is better not to reinvent the wheel on this complex topic, but to use the results of proven tests. Some audio circuits tend to be very large, and contamination in the audio environment, such as the grounds and chassis, can be a big problem for quality. It is recommended to add nonlinearity and natural distortions to the test by testing the remains of the bridge from scratch.
| Dielectric | Polystyrene | Polystyrene | Polypropylene | Polyester | Silver-mica | Ceramic | Polycarb |
| Temperature | 72 | 72 | 72 | 72 | 72 | 73 | 72 |
| Voltage level | 160 | 63 | 50 | 600 | 500 | 50 | 50 |
| Tolerance % | 2.5 | 1 | 2 | 10 | 1 | 10 | 10 |
| Error % | 2,18% | 0,28% | 0,73% | -7,06% | 0,01% | -0,09% | -1,72% |
| Diffusion | 0.000053 | 0.000028 | 0.000122 | 0.004739 | 0.000168 | 0.000108 | 0.000705 |
| Absorption | 0,02% | 0,02% | 0,04% | 0,23% | 0,82% | 0,34% | n/ |
| DCR, 100 V | 3.00E+13 | 2.00E+15 | 3.50E+14 | 9.50E+10 | 2.00E+12 | 3.00E+12 | n/ |
| Phase, 2 MHz | -84 | -84 | -86 | -84 | -86 | -84 | n/ |
| R, 2 MHz | 6 | 7,8 | 9,2 | 8,5 | 7,6 | 7,6 | n/ |
| Native resolution, MHz | 7 | 7,7 | 9,7 | 7,5 | 8,4 | 9,2 | n/ |
| Bridge | short | short | very low | high | short | short | high |
For you, audiophiles!
Today we will look at “audiophile” capacitors. This is quite a difficult matter - after all, some believe that the best capacitors are Telefunken, obtained from receivers produced in Germany between 1934 and 1944 (i.e. under Hitler). Some people believe that you need to wind the capacitors yourself from silver foil and the “correct” dielectric on the 13th on the new moon, turning your face to the south. Unfortunately, not only do I not have either the first or the second capacitors, I have never seen them in my life. Therefore, today there are only three contenders:
Metal paper capacitors K42U-2 and their outdated (but well “warmed up” for 30 years) version MBM. It is believed that paper is a very good-sounding dielectric, because it is made of living creatures and “responds” to beautiful music (I know well how a neighbor’s dog responds to music, but I just don’t understand how paper responds!). However, paper capacitors for amplifiers kosher.
And polystyrene capacitors K71-7. Polystyrene is a very successful dielectric with good properties. The big advantage of these capacitors is the low capacitance spread - for mine it is only 0.5% (for our metal-paper neighbors the capacitance spread is 10%, i.e. much worse). Such capacitors are well used in oscillators and precision (and complex) filters. Disadvantage: large dimensions. But the quality of the capacitors is excellent (and measurements confirm this once again).
When making measurements of this kind (almost at the limit of accuracy of the measuring system), the question of measurement repeatability arises. It is no secret that in the two months that have passed since last time, something in the (home) measurement conditions may have changed. And it really has changed. I repeated some of the previous experiments - the values turned out to be slightly different! But not by much, in the third significant figure, so the new results are almost comparable to the previous ones. So if the “audiophile” capacitors turned out worse, then this is so, the measurements have nothing to do with it! As proof, I present the result of a comparison of the capacitor K73-16, which took part in the previous test, and K42U-2, a new participant. These measurements were carried out practically simultaneously (with an interval of 5 minutes for resoldering the capacitors and the measurement itself) and under absolutely identical conditions. You can clearly see the difference:
Here is the same graph, only refined:
So, at least in terms of linearity, the paper is probably a little worse than lavsan.
1. Metal paper K42U-2
Kg = 0.0023%, K'g = 0.0078%
Not very bad, but not very good either. Maybe they have their own good side in some ways, but it is not visible here.
Conclusion: I didn’t find anything interesting for myself.
2. Metal paper MBM
Kg = 0.0014%, K'g = 0.0067%
Despite the fact that the spectrum of harmonics is somewhat wider, their amplitude is smaller, so the old one turned out better than the new one. Let me remind you that I take one capacitor at a time, which means I am not immune from unsuccessful copies. Maybe this happened because during 30 years of “warming up” the current through the capacitor only flowed in the “right” direction?
Conclusion: “From this side, it’s no better!” (Eeyore).
3. Polystyrene K71-7
Kg = 0.0016%, K'g = 0.0061%
But this is already quite good! Even good. Kg mainly consists of third harmonics. And the harmonic spectrum is narrow, indicating good linearity.
Conclusion: Very good quality with amazing accuracy. I simply don’t know any capacitors with the best quality-accuracy indicator.
Award ceremony for winners (ongoing)
Due to the obvious advantage of the polystyrene capacitor, I will not conduct a local rating, but will immediately give the overall result.
| Place | Type | “Regular” Kg, % | Place | Type | Normalized K'g, % |
| 1 | MBM | 0,0014 | 1 | K78-19 | 0,0049 |
| 2 | K78-19 | 0,0015 | 2 | EPCOS | 0,0053 |
| 3 | K71-7 | 0,0016 | 3 | K71-7 | 0,0061 |
| 4 | EPCOS | 0,0017 | 4 | K78-2 | 0,0064 |
| 5 | K73-16 | 0,0017 | 5 | MBM | 0,0067 |
| 6 | K73-17 | 0,0019 | 6 | K73-17 | 0,0074 |
| 7 | K78-2 | 0,0022 | 7 | K40U-2 | 0,0078 |
| 8 | FT-1 | 0,0023 | 8 | K73-16 | 0,0091 |
| 9 | K40U-2 | 0,0023 | 9 | FT-1 | 0,0098 |
| 10 | "Green" | 0,0025 | 10 | Imported "K73" | 0,012 |
| 11 | Imported "K73" | 0,0027 | 11 | "Green" | 0,024 |
| 12 | K10-17a | 0,83 | 12 | K10-17a | 2,2 |
| 13 | KM-5 | 2,1 | 13 | KM-5 | 6,1 |
Model characteristics
Ideally, the designer would expect the capacitor to be exactly at its design value, while most other parameters would be zero or infinite. The main capacitance measurements are not as noticeable here since the parts are usually within tolerances. All film CAPs have a significant temperature coefficient. Therefore, to determine which film capacitors are better for sound, testing is carried out with laboratory instruments.
The diffusion coefficient is useful in assessing the efficiency of an electrolytic power supply. This effect on the audio performance of signal CAPs is not consistent and may be quite minor. The number represents internal losses and can be converted to effective series resistance (ESR) if desired.
ESR is not a constant value, but tends to be so low in high quality capacitors that it does not have much impact on circuit performance. If high Q resonant circuits had been built, it would have been a completely different story. However, low dissipation coefficient is a hallmark of good dielectrics, which can serve as a good clue for further research.
Dielectric absorption may be more worrisome. This was a major problem with early analog computers. High dielectric absorption can be avoided, so mica audio capacitors can provide RIAA networks with very good sound.
DC leakage measurements should not affect anything since the resistance of any signal capacitor should be very high. When materials with a higher dielectric constant are used, less surface area is required and leakage will be virtually negligible.
Lower dielectric constant materials such as Teflon, despite its underlying high resistivity, may require a larger surface area. Then the leak may be caused by the slightest contamination or impurities. DC leakage is probably a good quality control, but it is not related to audio quality.
conclusions
1. Indeed, the larger the capacitance and the smaller the dimensions, the worse the linearity. Here is the dependence of distortion on capacitance for capacitors K10-17a, which have housings of almost the same size:
2. Small capacitors (less than 5 nF) have good linearity. Moreover, their distortions (within the limits of my measurement error) do not depend on the capacitance. Probably a different dielectric is used there?
3. Capacitors in larger packages are more linear. Compare 2-3 and 2-5 (they are shown broken down in the photo above). The volume of the case, and most importantly, the volume of the “crystal” is several times larger, and the distortions differ by more than an order of magnitude!
4. Different types of capacitors have different characteristics for the same capacitance. (Well, this is understandable, it’s not clear why they produce so many different ones?!)
5. I wonder what happens in SMD capacitors, which are even smaller in size?
6. The dependence “the better the TKE, the better the linearity” (and this is a widely held opinion) is generally confirmed, but not entirely unambiguously. Somewhere it’s like this, and somewhere it’s the other way around. Apparently, everything depends on the properties of the dielectric, and while TKE is standardized by manufacturers and specifications, linearity is not. But in order to thoroughly understand the issue, you need to conduct many experiments with capacitors of different TKE groups, and this is not yet possible.
7. The sound quality of an amplifier with high-capacity ceramic pass-through capacitors will be spoiled.
Unwanted parasitic components
Transistors, integrated circuits, and other active components have a significant impact on the quality of audio signals. They use power from current sources to change the characteristics of the signal. Unlike active components, ideal passive ones consume no power and should not alter signals.
In electronic circuits, resistors, capacitors and inductors actually behave like active components and consume energy. Because of these parasitic effects, they can significantly alter audio signals, and careful selection of components is required to improve quality. The ever-increasing demand for audio equipment with better sound quality is forcing CAP manufacturers to release devices with better performance. As a result, modern capacitors for use in audio applications have better performance and higher sound quality.
Parasitic CAP effects in an acoustic circuit consist of equivalent series resistance (ESR), equivalent series inductance (ESL), series voltage sources due to the Seebeck effect, and dielectric absorption (DA).
Typical aging, changes in operating conditions and specific characteristics make these unwanted parasitic components more complex. Each parasitic component affects the performance of an electronic circuit differently. To begin with, the resistance effect causes DC leakage. In amplifiers and other circuits containing active components, this leakage can lead to significant changes in offset voltage, which can affect various parameters, including quality factor (Q).
The capacitor's ability to handle ripple and pass high frequency signals depends on the ESR component. A small voltage is created at the point where two dissimilar metals are bonded due to a phenomenon known as the Seebeck effect. Small batteries due to these parasitic thermocouples can significantly affect the performance of the circuit. Some dielectric materials are piezoelectric, and the noise they add to the capacitor comes from the small battery inside the component. Additionally, electrolytic CAPs have parasitic diodes that can cause changes in offset or signal characteristics.
Application nuances
To choose a capacitor for yourself, you should pay special attention to the parameters that meet your needs, as this will ultimately affect how pleasant the audio will sound.
Electrolytic ones can be chosen if the quality of the reproduced sound is not very important. If you choose this model, then a capacitor at an average level will complete the task. Such a device will not be expensive and will not require any complicated installation of a sound card. As you would expect, the type of capacitors considered will be installed in models of the lowest price segment and high Hi-Fi results cannot be expected.
It's a different matter when it comes to devices based on film and paper. When making sound amplification equipment, first of all, you should give preference only to this type of capacitors, but there are still some things worth taking into account.
Film capacitors often suffer from interference, so not all models need to be used. This is partly caused by parts that may indicate non-linear distortion, particularly at specific frequencies. It is advisable to use this kind of device for recharging in not the most important circuits on the board. It is better when film capacitors are responsible for the main work; they will be able to accumulate current.
Parameters affecting the signal path
In electronic circuits, passive components are used to sense gain, establish DC blocking, suppress power supply noise, and provide bias. Inexpensive components with small dimensions are commonly used in portable audio systems.
The performance of actual polypropylene audio capacitors differs from that of ideal components in terms of ESR, ESL, dielectric absorption, leakage current, piezoelectric properties, temperature coefficient, tolerance and voltage coefficient. While it is important to consider these parameters when designing a CAP for use in an audio signal path, the two that have the greatest impact on the signal path are called voltage coefficient and inverse piezoelectric effect.
Both capacitors and resistors exhibit changes in physical characteristics as the applied voltage changes. This phenomenon is commonly referred to as stress factor and varies depending on the chemistry, design, and type of CAP.
The inverse piezoelectric effect affects the electrical rating of capacitors for an audio amplifier. In audio amplifiers, this change in the electrical value of a component results in a change in gain depending on the signal. This non-linear effect causes sound distortion. The inverse piezoelectric effect causes significant audio distortion at lower frequencies and is the main source of stress coefficient in Class II ceramic CAPs.
The voltage applied to the CAP affects its performance. In the case of Class II ceramic CAPs, the capacitance of the component decreases as an increasing positive DC voltage is applied. If high AC voltage is applied to it, the component's capacitance is reduced in a similar manner. However, when low AC voltage is applied, the component's capacitance tends to increase. These changes in capacitance can significantly affect the quality of audio signals.
All sorts of non-electrolytes
It all started with the fact that I didn’t like the sound of one of my amplifiers, and I had long suspected that the capacitor connected to its input was introducing nonlinear distortion. After, when examining an amplifier based on the TDA7294 chip, I discovered an increase in distortion at low frequencies, and as the capacitance increased, the distortion decreased (everything is clear here - the larger the capacitance, the lower the resistance of the capacitor, and the less its influence on the signal, and therefore distortions), my suspicions turned into confidence. And I decided to measure what kind of distortion the capacitors introduce. And compare several of the most common types. After all, capacitors have a great influence on the sound quality of amplifiers !
I must immediately warn you that this is not a completely correct comparison - I used the capacitors that I had. They had different capacitances, so I worked with them at different frequencies and the voltages applied to them were not exactly the same. On the good side, it was necessary to carry out measurements under absolutely identical conditions: both frequency and voltage should be the same. And it was necessary to measure at several frequencies and with different voltages. Yes, and I had to take several pieces of identical capacitors - suddenly I came across one of them that was slightly defective. That is, the measurement results are not the “ultimate truth” when comparing capacitors. If the results differ greatly, then we can confidently say that one of the capacitors is better than the other. But if the difference is small, then it is quite possible that the one, which in my case was a little better, will work a little worse at a different frequency.
And then, because I only measured the harmonic coefficient, and did not measure the other quality parameters!!! Although from the point of view of the effect of pass-through capacitors on the sound, the quality of capacitors largely depends on their linearity. Agree that if after the capacitor there is a resistor of tens of kilo-ohms, then there is no difference between a capacitor with ESR = 0.01 Ohm and a capacitor with ESR = 0.001 Ohm! These fractions of an ohm will be lost against the background of the resistance of the leads, soldering and tracks! But if half the amplifier Kg consists of a capacitor Kg, then this is not good.
However, I would call the results stunning. There are good and bad capacitors, and there are absolutely terrible ones!!! I knew that ceramic capacitors with a dielectric having poor TKE are nonlinear, but I didn’t think it was that much!
All measurements were carried out accurately, correctly and correctly, without methodological errors. The measurement scheme is shown in Figure 1.
Rice. 1.
A sinusoidal voltage of maximum amplitude (2V rms) was supplied from the sound card, the resistor was selected so that the voltage on the capacitor was within 2...2.5 V amplitude (i.e. approximately 1.5 volts rms) value. In addition to the voltage on the capacitor, the output voltage of the sound card was also measured to control its distortion. From the measurements it is clear that the distortions of the map itself are much smaller and do not affect the accuracy (distortions of the map were subtracted from the results, the subtraction was absolutely correct: the square root of the difference in the squares of the amplitudes of the corresponding harmonic).
In order to show the accuracy of the measurements, I will give two spectra of the capacitor current (and in this way I measure the current). These spectra will be further processed for greater clarity. In the calculations, only harmonics were taken into account; interference, if there was any (look for interference in the pictures!), was not taken into account.
Rice. 2.
Rice. 3.
Another important point is the calculation of the harmonic coefficient Kg. In addition to the usual method (Fig. 4 a), I used one normalized to the harmonica number (Fig. 4.b).
Rice. 4.
This method of rationing was invented by engineers from the laboratory of the British Air Force company in the 50s of the twentieth century. And this method, when the harmonic voltage is multiplied by the square of its number, allows us to take into account the width of the harmonic spectrum. Why is this necessary? And then, the higher the order of nonlinearity and the wider the spectrum of harmonics, the worse the sound. Here is an example in Figure 5:
Rice. 5.
All three variants of the distortion spectrum give the same Kr = 0.1%. But the green spectrum contains only two harmonics, which means that such distortions are less noticeable to the ear. The red spectrum contains harmonics up to the 10th, and is the worst to hear. And Kg is the same for all of them and does not allow these spectra to be distinguished. And the normalized K'r will give the following values for these spectra: 0.12%; 0.18% and 0.33%. Feel the difference!
I want to say that this is not “The Next Newest Great and Accurate Distortion Measuring Method”! This is simply a modification (and quite legal) of the conventional method, but more advanced: if the traditional Kg allows one to take into account only the average value of the nonlinearity of the transfer characteristic (this is like the average temperature throughout the entire hospital, including the morgue), then the normalized one allows one to take into account the order of this nonlinearity. And, despite the fact that it is very far from perfect and does not correspond very well to auditory sensations, it is still better than simple Kg. That is, you can look at it from the other side: regular Kg correlates even less with subjective sensations than normalized Kg. The coefficient is normalized to the second harmonic and its physical meaning is to show the average nonlinearity, taking into account how much higher harmonics are worse than the second.
And this approach was beneficial. Further it will be seen that the capacitor EPKOS and K73-16 Kg is the same and equal to 0.0017%. Does this mean that the capacitors are the same? It may very well not be the case. But if you look at the normalized coefficients, then for EPKOS K'g = 0.0053%, and for K73-16 K'g = 0.0091%. Those. domestic lavsan capacitor has a wider range of harmonics and sounds worse than imported polypropylene. But in order not to deprive readers of familiar reference points, I also give the usual Kg.
It’s time to move on from the protracted introduction to the point and introduce today’s participants in the “Mr. Capacitor” competition (Fig. 6).
Rice. 6.
Ceramic capacitors K10-17a and KM-5 (most likely this is an imported analogue of our K10-17b or K10-17v; I recently saw exactly the same domestic capacitor type K10-73, but according to the text I will leave the name KM-5, i.e. because they all originated from KM-5), lavsan film K73-16 and K73-17, fluoroplastic FT1 and polypropylene domestic K78-2, K78-19 and imported EPCOS. I don’t know the brand of the capacitor located in the center of the top row. I suspect it's film, but what kind? This is most likely imported (the kind found in multimedia speakers, for example), it is actually dark green in color (it didn’t turn out in the photo), so I will call it “green”. When I find out the type, I will write it here.
So, let's go! In the spectrograms, the red spectrum is the capacitor current, the blue spectrum is the audio output (since connecting a capacitor as a nonlinear load leads to distortion; I already wrote above that these distortions were taken into account when calculating the harmonic coefficients).
1. Ceramic K10-17a
Kg = 0.83%, K'g = 2.2%
Scary? Me too. I loved these capacitors for their good TKE (temperature coefficient of capacitance), but was not interested in distortion (I rarely used them for sound). And how bad it is. Moreover, the spectrum of harmonics is very wide.
Conclusion: do not use for sound!
2. Ceramic KM-5 [K10-73] (class N90)
Kg = 2.1%, K'g = 6.1%
This is actually some kind of nightmare! I suspected that these were bad capacitors, I thought that their distortion was so high that it could even be half a percent. But it turned out that everything was much, much worse! And if you consider that their capacity depends very much on temperature...
Please note that connecting this capacitor to the audio output immediately creates a heap of harmonics for it! Those. and the output voltage is distorted due to this capacitor!
Conclusion: keep away from sound circuits, preferably in another closet and in another room! Also not recommended in power supply circuits for audio devices.
| Important Note |
| In my opinion, we have a completely stupid designation system for ceramic capacitors in our country. The fact is that they use completely different ceramics: if the container is small, then the ceramics are of quite high quality, with good linearity and temperature stability. When you need to get a high capacitance with small dimensions, they use simply disgusting ceramics - the linearity is very poor, and there is no thermal stability (when heated by 20 degrees, the capacitance can change 2...3 times!), and there is also a ferroelectric effect - the capacitor works both as a piezo speaker and as a piezo microphone! Moreover, the infectious manufacturers do not tell anyone which capacitor contains which ceramics. Kind of figure it out for yourself. If I were them, I wouldn’t lump everything into one pile, but would give different types depending on the type of dielectric. Then everything would be clear - capacitors of this type have a small capacitance, but the stability and linearity are good, while capacitors of another type have a high capacitance, but at the expense of quality. But no! They're deliberately confusing it, probably, so that the spies won't guess! |
Why did I love K10-7a capacitors before? They have a large body compared to KM-5 (K10-73) and good TKE. So I thought that this large case was filled with a lot of quality ceramics. But it turned out that the ceramics there, although better than those of KM-5, are still crap. Just for fun, I broke apart a couple of capacitors (each 0.1 µF) to see what was inside:
A heartbreaking sight: such a small crystal in such a large case! Now it’s clear why the linearity is bad - I thought that the walls of the case were thin, and the inside was all guts. But no... But my assumption that a larger capacitor (with the same capacity) can have a higher operating voltage seems to be confirmed - the crystal there is larger, probably due to the greater thickness of the dielectric. But a microscope will give the exact answer, and I don’t have one.
I will definitely find and measure a capacitor of this type, but a small capacitance with a good dielectric! To compare...
3. Film K73-16 (lavsan)
Kg = 0.0017%, K'g = 0.0091%
Well, that's a completely different matter! If it weren’t for this “tail” of harmonics of a rather high order...
Conclusion: Use for your health.
4. Film K73-17 (lavsan)
Kg = 0.0019%, K'g = 0.0074%
This is where it’s interesting: his normal Kg is higher than the previous one, and his normalized one is less. This is because its 3rd, 4th and 5th harmonics are slightly higher, but the 11th is not there at all! And the “bad” 8th and 9th are noticeably smaller.
Conclusion: it seems that the “people’s” capacitor is slightly better than the K73-16, despite the fact that the K73-16 is a military one (5th acceptance). But maybe this is an accident - the difference is small...
5. Fluoroplastic FT-1
Kg = 0.0023%, K'g = 0.0098%
Overall a good capacitor. Fluoroplastic has a number of advantages (for example, maximum transmitted reactive power at high frequencies), but they are maximized in other places, for example in speaker filters.
Conclusion: normal.
6. Film K78-2 (polypropylene)
Kg = 0.0022%, K'g = 0.0064%
The lowest normalized harmonic distortion so far. As usual, Kg is inferior to the K73-16 capacitor, but after comparing the spectra, you understand that it is better to use the normalized coefficient K'g to evaluate linearity! The maximum that was found was the 5th harmonic. There are no higher ones.
Conclusion: a very linear capacitor.
7. Film K78-19 (polypropylene)
Kg = 0.0015%, K'g = 0.0049%
Same picture, only a little better!
Conclusion: the most linear capacitor in the review! It will “sound”!...
8. Film EPCOS (polypropylene)
Kg = 0.0017%, K'g = 0.0053%
Ours turned out even better! True, this is at the limit of accuracy, and at the same frequency. I don’t know where the 11th voltage harmonic came from, and why there is no corresponding 11th current harmonic. Maybe some tricky feature of the capacitor. I tried it several times under different conditions - the result was the same.
Conclusion: it’s not for nothing that they charge so much money for it. But it would be nice to take a closer look at our K78-19 - it looks like it is not inferior to the bourgeois one (and according to these measurements - even better)! And it's cheaper.
9. Film green
Kg = 0.0025%, K'g = 0.024%
In principle, not bad, if it weren’t for the “separate” 12th, 14th and 17th harmonics that came from incomprehensibly. Even though they are small, they are there. They were immediately caught by K'g, sensitive to such outrages, who immediately grew 10 times larger because of them (does anyone still doubt its usefulness?).
Conclusion: can be used for power supply and non-critical circuits. For example, in the same multimedia acoustics (in an amplifier).
10. Imported “K73”
Compared to “regular” K73-17 capacitors, these (apparently) imported ones (I don’t know their brand yet) have smaller dimensions and are sold for voltages of 100 volts and above. I have never seen one with a voltage less than 100 volts. Moreover, more and more of them have appeared over the past year or two. Let's see what kind of bird it is.
Kg = 0.0027%, K'g = 0.012%
Linearity is slightly worse than that of K73-16 and K73-17. This is probably a price to pay for the smaller dimensions. But in principle it’s not bad.
Conclusion: you can use it, but our K73-17 is better. But in power supply circuits, these capacitors turn out to be more profitable - at voltages above 50 volts, the K73-17 at 63 volts is no longer worth using. But these will easily fit and will be smaller in size (which means you can put a larger container in the same place!).
Winner's reward ceremony
Let's put the capacitors in their places, taking into account that we have two evaluation coefficients, and the table of records also turns out to be double (interestingly, in the right half, all the first places are taken by polypropylene capacitors, which, according to subjective assessments, are always put in first place. Does this mean that normalized K'r is closer to subjective sensations?..)
| Place | Type | “Regular” Kg, % | Place | Type | Normalized K'g, % |
| 1 | K78-19 | 0,0015 | 1 | K78-19 | 0,0049 |
| 2 | EPCOS | 0,0017 | 2 | EPCOS | 0,0053 |
| 3 | K73-16 | 0,0017 | 3 | K78-2 | 0,0064 |
| 4 | K73-17 | 0,0019 | 4 | K73-17 | 0,0074 |
| 5 | K78-2 | 0,0022 | 5 | K73-16 | 0,0091 |
| 6 | FT-1 | 0,0023 | 6 | FT-1 | 0,0098 |
| 7 | "Green" | 0,0025 | 7 | Imported "K73" | 0,012 |
| 8 | Imported "K73" | 0,0027 | 8 | "Green" | 0,024 |
| 9 | K10-17a | 0,83 | 9 | K10-17a | 2,2 |
| 10 | KM-5 | 2,1 | 10 | KM-5 | 6,1 |
I think comments are unnecessary.
General characteristics of harmonic distortion THD
The THD of capacitors for audio depends on the dielectric material of the component. Some of them can produce impressive THD performance, while others can seriously degrade it. Polyester capacitors and aluminum electrolytic capacitors are among the CAPs that give the lowest THD value. For Class II dielectric materials, the X7R offers the best THD performance.
CAPs for use in audio equipment are generally classified according to the application for which they are used. Three applications: signal path, functional tasks and voltage support applications. Ensuring that the optimal MKT capacitor is used for audio in these three areas helps improve output tone and reduce audio distortion. Polypropylene has a low dissipation coefficient and is suitable for all three areas. While all CAPs used in an audio system affect sound quality, the components in the signal path have the greatest impact.
Using high quality audio grade capacitors helps to significantly reduce audio quality degradation. Because of their excellent linearity, film capacitors are commonly used in the audio circuit. These non-polarized audio capacitors are ideal for premium audio applications. Dielectrics commonly used in audio quality film capacitor designs for use in the signal path include polyester, polypropylene, polystyrene, and polyphenylene sulfide.
CAPs for use in preamplifiers, digital-to-analog converters, analog-to-digital converters, and similar applications are collectively classified as functional reference capacitors. Although these non-polarized audio capacitors are not in the signal path, they too can significantly degrade the quality of the audio signal.
Capacitors, which are used to maintain voltage in audio equipment, have minimal impact on the audio signal. Regardless, care is required when selecting CAPs that support voltage for high-end equipment. Using components optimized for audio applications helps improve the performance of the audio circuitry.
Best Electrolytic Capacitors for Audio
Electrolytic devices are characterized by significant capacity and long service life. They are reliable, inexpensive, can operate at constant voltages, but are not suitable for high-quality equipment that requires detailed sound.
Elna Silmic II
The leader among electrolytic components are audio amplification capacitors from Elna Silmic II. These are budget devices with good technical characteristics. Used to play high quality music.
The case is made of aluminum, which reliably protects the inside from external influences. Inside there are silk fiber and oxygen-free thread, providing decent sound at all frequencies without distortion. The sizes are small, so the product can be used together with different equipment.
Advantages:
- Stable operation at low, medium, high frequencies;
- Pleasant natural sound without distortion;
- Low cost;
- Sold in almost all electronic components stores.
Flaws:
- Low operating voltage.
Mundorf E-Cap AC Plain
A non-polar capacitor for sound from a German manufacturer is made with smooth plates. The use of such technology significantly improves sound quality with minimal losses.
But the use of smooth facings caused an increase in the dimensions of the final product, which must be taken into account when choosing. The charge storage device is able to operate stably over a long period.
Advantages:
- High-quality assembly;
- Long work;
- Decent sound characteristics;
- Minimum losses;
- Low cost.
Flaws:
- Large size.
Polystyrene plate-dielectric block
Polystyrene capacitors are made by winding a plate-dielectric block, similar to an electrolytic, or by laying them in successive layers, such as a book (folded foil film). They are mainly used as dielectrics of various plastics such as polypropylene (MKP), polyester/mylar (MKT), polystyrene, polycarbonate (MKC) or Teflon. High purity aluminum is used for the plates.
Depending on the type of dielectric used, capacitors are produced in different sizes and capacities with operating voltage. The high dielectric strength of polyester makes it possible to produce the best electrolytic capacitors for audio in a small size and at relatively low cost for everyday use where special qualities are not required. Capacitances from 1,000 pF to 4.7 microfarads are possible at operating voltages up to 1,000 V.
The dielectric loss factor in polyester is relatively high. For audio, polypropylene or polystyrene can significantly reduce dielectric loss, but it should be noted here that they are much more expensive. Polystyrene ones are used in filters/crossovers. One disadvantage of polystyrene capacitors is the low melting point of the dielectric. This is why polypropylene audio capacitors are usually different because the dielectric is protected by separating the solder leads from the capacitor body.
Rating of capacitors for sound
Electronic circuits consist of many components that are responsible for the performance of the entire system. Such products include capacitors. This is a device consisting of two plates and insulation, which is capable of storing different amounts of energy. Capacitors are classified according to many criteria, but the main one is the dielectric material. Highlight:
- Electrolytic or oxide. This is a type of capacitor, when connecting which it is important to observe the polarity, otherwise the circuit will not work. They have an oxide layer that appears on the aluminum or tantalum anode. A liquid or gaseous electrolyte is used as a cathode. Used in budget musical equipment. Most products are not suitable for audio equipment due to low strength, but there are some designs of good quality.
- Film. The dielectric can be made of polystyrene, polyethylene and other film materials. It has a high insulation resistance and has a self-healing effect in case of breakdown. Despite the capacity, which is less than that of electrolytic ones, they show fast operation, filter the supply voltage, and perform a separation-transition function.
- Paper. The insulator is made of paper, which can be dry or impregnated with substances. This is one of the highest quality capacitors, so their price is high. Produces excellent sound with the best audio performance. They are used primarily in premium audio equipment that requires high detail of each sound.
For a specific project, the capacitor is selected individually. It is possible to combine several types, for example, electrolytic and film to increase the total capacity. It is difficult for an inexperienced user to understand all the nuances when choosing a suitable component, so in order to make the task easier, we studied the main characteristics of capacitors, determined where they are used, in order to create our own rating of the best. When selecting nominees, the following were taken into account:
- Dielectric material;
- Capacitive characteristics;
- Permissible deviation from the nominal value;
- Sound quality;
- Availability.
All devices were divided by the type of dielectric used so that the user could select a product from the desired category. The rating was developed based on the specified passport data of the charge storage device, reviews from real customers, and expert reviews.
Is a capacitor needed in the tweeter circuit?
#1 OFFLINE rema
- From: Riga, Latvia
#2 OFFLINE Feanor
- From: Moscow
Dear colleagues in the ward, please tell me whether a capacitor is needed in the tweeter circuit, the system from the head sends the signal to a crossover (Clarion MCD 700x), then channel by channel to amplifiers and then to speakers, thanks in advance.
#3 OFFLINE Black_ru
- From: Omsk
In my opinion, just install an incandescent light bulb and that’s enough.
1st order is like casual sex without a condom - cool, but risky. © Nitro
HDD player+P50x(tweak) – hand-made interconnect – Nakamichi pa200(tweak)+SS9700 – Scan SS800(tweak)+SPX-Z15 – EOS 620(tweak)+Ciare HW251
#4 OFFLINE rema
- From: Riga, Latvia
#5 OFFLINE Feanor
- From: Moscow
Why a light bulb after the crossover?
As far as I understand, in this case there is an active crossover, and after it there is a crossover. And the light bulb is needed so that at high power the tweeter does not burn out, because... bulbs in a cold state have very little resistance, but with increasing “power” they become excited and the resistance increases. Because the lamp is switched on in series with the tweeter; after energizing the lamp filament and increasing the resistance, the tweeter does not burn out.
PS: passive crosses have built-in lamps; in these cases, additional lamps are not required.
In principle, there was a good article on the bluesmobile about this, I’ll look for it now, otherwise I’m retelling other people’s words.
The selection of lamps for tweeters is even described here: https://wap.hiend.bor. 0000021-000-0-0
#6 OFFLINE rema
- From: Riga, Latvia
