Triode amplifier. Single-ended triode tube amplifier

The Musical Paradise MP-301 MK3 tube amplifier was designed in Canada by Harry Huang and produced in China by a small factory.

The first version - Musical Paradise MP-301, was released in 2008.

Actually, in Canada this amplifier has the largest number of fans. Read them, and there are a lot of interesting things there.

Pay attention to the photo from the forum - a man relaxingly enjoying himself with his dogs to the sounds of this amplifier.


This is exactly the effect that tube sound has on the body of all mammals. So don't be surprised if, after turning on the amplifier, you want to lie down, close your eyes, relax and put all thoughts out of your head. (Remember to leave the door open so your dogs, cats, wives and children can come hang out with you).

“Tube sound,” by the way, is a good cure for depression, melancholy, melancholy and headaches.

The third version of the amplifier is much different from the second, and the first is not at all similar to the third, neither in design nor in circuitry.

This is the first version:




This is the second version:






This is the third one:


Inside, everything is much better, and completely different. Pay attention to high-quality components: “military” resistors Vishay Dale, capacitors RIFA 450, Rubycon, Philips BC and Nichicon:


The transformers are wound with oxygen-free copper wire, the cores are Japanese transformer iron Z11 and M6. The volume control is made using ALPS technology.

The main feature of this amplifier is that it is universal - its output tubes can be selected at will (and then changed independently): 350B


350C, KT66, KT77, KT88, 6L6, EL34, 5881, Russian 6P3S and 6P3S-E.

You can install G807 lamps (Russian G807) through a special adapter:



Input tubes can be either 6J8P, 6SJ7, 6SH7 (or Russian “glass” 6Zh8P and “metal” 6Zh8).

The amplifier has a chip for automatic bias control, so no matter how bad your tubes are, it will automatically set them to the optimal operating mode. Despite this, the manufacturer still recommends installing selected pairs of lamps in the device.

When ordering an amplifier, you will be given the opportunity to choose the type of tubes for free. I ordered the weekend 6L6GC (included lights up like a Christmas tree):






and input 6J8P:


All of them turned out to be produced by the Chinese company Shuguang.

For branded ones you will have to pay extra when ordering. All lamps arrived new, in packaging, matched in pairs according to parameters.

I compared Chinese Shuguang 6L6GC lamps with used lamps purchased on Ebay: American Sylvania 6L6GB, Japanese Toshiba 6L6GC. I also listened to 6P3S from the seventies, and the new 6P3S-E from the eighties, and did not notice any particular difference between all of them, although the Japanese tubes looked more carefully made.

Canadians also noticed that O A greater effect is obtained by replacing input lamps (RCA with a metal bulb) than replacing output lamps. But still, it is recommended to install Czech JJ (Tesla) EL34 or KT88, or RTF EL34 “on the way out”.

Now I have replaced all the Chinese lamps. At the entrance I put “mixed-caliber” used Sylvania 5SJ7GT:


Fine.

I ordered old metal radio tubes 6Zh8 for fun:


New ones arrived, each in a cardboard box, made in my native Novosibirsk, year of manufacture 1968. They are at my entrance and will remain.

P.S. No, they won't stay. The sound of the new 6Zh8 is worse than that of a used Sylvania 5SJ7GT.

At the exit they are now (bought at):

I paid 1600 rubles for two lamps and delivery (700+700+200) (to Novosibirsk from St. Petersburg).

I put the same on the output (bought on):

I paid 1200 rubles for two lamps and delivery (500+500+200).

(You could also order:

although they are much more expensive - 2000 rubles each. Besides, because of the “golden” paint, they look somehow vulgar).

The sound of the Musical Paradise MP-301 MK3 amplifier is wonderful.

In an audio salon in my city, I compared it with the VINCENT SV-237 hybrid:


for a crazy 162 thousand rubles, and (in my opinion) mine sounded better.

I had heard tube amplifiers before, a long time ago; my first was the legendary Priboy 50 UM 204S in the nineties. Even then I remade it (although I had never held a soldering iron before in my life) according to the recommendation in the article by A.M. Likhnitsky in the Audio Store magazine, No. 1, 1996. He sounded very good.

Then I gave it to specialists to convert it into a single-ended circuit using 6C4C tubes. After the conversion, only the output transformers, the top cover, the chassis, and the bottom cover remained from the original Priboy.

This amp sounded good, but it didn't have much highs or lows, and it didn't play well at Black Metal, which is what I mostly listen to. That's why I sold it in the two thousandths.

After studying the forums, I came to the conclusion that for metal I need an amplifier using 6L6 or 6P3S tubes.

And finally, ten years later, I bought myself just such an amplifier.

Having dragged this amplifier, which is not particularly heavy, into the audio salon, and listened to different acoustics on it, I discovered that the expensive Bowers & Wilkins 685 bookshelf speakers are better than the cheap floorstanding speakers. I had to run over the toad.

The amplifier has a headphone output (2 watts) on the front panel, the quality of which was a pleasant surprise for me. The sound through it is simply amazing, it is powerful, assertive, and in some ways even epic. Moreover, this is clearly audible even on headphones not in the highest price category.

So if everyone in your house is asleep and you can’t stop listening to music, connect headphones to the MP-301 MK3.

This is important: if your source has a headphone output, do not use it, but connect your headphones through this amplifier. It will improve the sound of any recording.

Pros and cons of the purchase:

There are a lot of advantages.

Minuses:
1) Power cord with Canadian three-prong plug.

2) The glossy piano lacquer finish easily shows fingerprints and dust.

3) The amplifier gets quite hot during operation, especially the output tubes outside and the choke in the power supply inside. The ventilation holes are located only at the bottom.

4) The headphone output may have a small background (with 6L6 tubes it will be minimal).

5) The output connectors for acoustics are located very close to each other, which makes it impossible to use spade type connectors for connecting speakers due to the risk of short circuit:


6) Harry Huang does not give the circuit diagram of this amplifier to anyone, so you will upgrade it (replacing resistors and capacitors, for example) at your own risk.

This review will be updated, please write comments.

I'm planning to buy +35 Add to favorites I liked the review +53 +132

Tube Amplifiers

The article offers two options for tube audio power amplifiers. A feature of the presented designs is the galvanic connection between the cascades. The authors attempted to find the optimal combination of simplicity, quality and repeatability of an UMZCH with a single-ended output stage with a power of up to 8 W per channel.

Often, when evaluating a sound reproduction system, the listener consciously or unconsciously focuses on subjective sensations that determine the sound quality. In this case, such characteristics as naturalness, “transparency”, “softness” of sound, “speed” (distinctness) of bass, detail of the reproduced composition, etc. are used. Of course, with a certain degree of convention, these characteristics can be associated with the objective parameters of this system - amplitude -frequency response (AFC), harmonic and intermodulation distortion coefficients, noise and background levels, speaker damping coefficient, etc. Each of the tube UMZCHs offered here, from our point of view, can be considered as a combination of acceptable sound quality, good technical parameters and comparative simplicity circuit solution.

The first amplifier is single-ended, based on a G-807 tube (in Fig. 1 there is a diagram of one of the channels of a stereo amplifier). It is a modernization of the "Profundo" amplifier. Here, an additional cathode follower is used, assembled on the triode part of the VL1 (6F1P) combined radio tube. This inclusion makes it possible to coordinate the operation of the input and output stages in order to eliminate the drop in frequency response in the HF region and reduce nonlinear distortions mainly in the LF region that arise in such a circuit when directly connecting the pentode anode and the G-807 control grid.
As in the first version of "Profundo", all stages of the amplifier are covered by a chain of local feedbacks that follow one another. Local positive feedback (PLF) is necessary not only to exclude the oxide capacitor from the VL1.1 cathode circuit, but also to improve the reproduction of low frequencies (“fast” bass). In its circuit, a voltage divider R7R5 is formed, to which the tetrode screen grid is connected. Capacitor C1 is not required, but can be used to eliminate possible noise when moving the resistor R1 slider. The output stage is assembled using an ultra-linear circuit, which reduces its nonlinear distortion and output impedance.

Power supply for lamp UMZCH

The UMZCH I power supply uses a unified transformer TS-180 (from old TVs). The rectifier is made on semiconductor diodes VD1, VD2 according to a symmetrical voltage doubling circuit. The shallow depth of the general feedback does not provide significant suppression of the background from anode voltage ripples, therefore, U-shaped filters with chokes are used in the power supply.
The installation of the UMZCH is carried out either according to the method described in the article [1], or (in the absence of instruments) by adjusting the resistor R4 until the maximum audibly undistorted signal is achieved. The quiescent current of the anode of the G-807 lamp, equal to 70 mA, can be adjusted by selecting resistor R8. The offset on the control grid of the output lamp relative to the cathode is about -20 V.
This UMZCH allows the use of an output transformer with a relatively small magnetic core without loss of low frequencies. A broadband highly sensitive (90... 100 dB/W/m) dynamic head can be used as BA1.


In Fig. Figure 2 shows a diagram of a single-ended UMZCH II using 6S41C triodes in the output stage (one of the two channels of a stereo amplifier). The first stage is amplified by tetrode VL1 (6E5P), from the anode of which the signal is supplied to the grid of the output lamp VL2 (6S41C). The signal from the middle of the secondary winding of the output transformer T1 through capacitor C2 enters the screen grid VL1, forming a PIC loop. It additionally increases the amplification of the low-frequency signal using circuit LC2 (where L is the inductance of half of the secondary winding of transformer T1), i.e., it performs a corrective function in the region of lower frequencies of the audio range. In this case, the resonant frequency of the circuit can be estimated as fres = 1/(2π√LC2). The OOS is formed by resistor R6 on the screen grid VL1. OOS reduces nonlinear distortion and prevents self-excitation of the amplifier at low frequencies.
In the power supply of this amplifier, a rectifier based on semiconductor diodes (using a bridge circuit) is used for the output stage, and for the first stage (on tetrode VL1) a half-wave rectifier is used on diodes VD5, VD6 with capacitors SZ, C5. As a mains transformer in the power supply for both UMZCHs, you can use the TS-180 transformer (or its modifications, for example, TS-180-2) with sufficient power reserve, appropriately connecting the secondary windings to obtain the required alternating voltage (63+63+ 42 V).
The amplifier is adjusted by setting the quiescent current for VL2 to 10 = 120 mA by selecting resistor R3. In this case, the bias voltage on the grid of the output lamp relative to the cathode should be about -75 V.
The magnetic cores of the mains and output transformers should be placed mutually perpendicular in the housing to minimize magnetic coupling through the stray field.


The parameters of all UMZCH are given in table. 1. They were measured using an S-107 oscilloscope, a B3-38 millivoltmeter, a GZ-118 generator and a notch filter included in its kit.


In Fig. Figure 3 shows the frequency response of the two proposed amplifiers. For UMZCH I, the frequency response was measured at its rated power Pnom = 5 W (hereinafter - at a frequency f = 1 kHz), for UMZCH II - at a power Pnom = 6 W.


In table 2 shows the parameters of the output transformers for lamps used in UMZCH I and II.
To extend the service life of radio tubes, it is advisable to install a switch (toggle switch), through which voltage is supplied to the anodes of the lamps approximately 20 s after their filament is turned on.
Chokes L1 and L2 in Fig. 1 and fig. 2. can be replaced by standardized D31-5-0.14. If they are not available, you can use chokes Dr-1.2-0.16 and the like, however, in UMZCH II the capacitance of capacitors C4, C6 and C7 should be increased to 300 µF.
In both UMZCH designs, variable resistors R1 with a type B control curve are used. The remaining resistors are MLT or imported. Powerful resistor R8 (2.4 kOhm) in the diagram in Fig. 1 - for example, PEV-10 or imported one with higher power. The tolerance for variation in resistor values ​​is ±10%. Trimmer resistor - SP-2-2-0.5, SP-3-9, etc., preferably with an axis stopper.
Oxide capacitors - for example, K50-12, K50-17, K50-31 and similar (or imported). The capacitor at the input of the UMZCH can be selected from film (for example, K73-9 series) or paper (K40U-9 series), although its effect on the sound is less noticeable than interstage (in both amplifiers the connection between stages is direct, without capacitors).

When assembling and debugging amplifiers, extreme care and caution should be observed (high voltage). The issues of eliminating AC background are well outlined in. Let us add that the UMZCH chassis can be made of aluminum or steel with a thickness of 1.5 and 0.5...0.8 mm, respectively. Input connectors are RCA ("tulips"), output terminals are threaded. It is advisable to place the trimmer resistor in the cathode circuit as close as possible to the input lamp. Its body is connected to a common wire or shielded. The wires of the incandescent circuits are twisted together.
The psychoacoustic characteristics of each of the described UMZCHs have their own characteristics. In our opinion, the first UMZCH is characterized by detail and transparency of the sound palette, the second - a combination of the softness of the bass register with the clarity of high-frequency sound components. A common characteristic feature of both designs is the “warmth” of sound, as is commonly said about sound with tube amplifiers.
We wish you success!

S. AKHMATOV, D. SANNIKOV, Ulyanovsk

LITERATURE
1. Akhmatov S., Sannikov D. "Profundo" - tube audio frequency amplifier. - Radio, 2012, No. 5, p. 16, 17.
2. Adamenko M.V. Secrets of low-frequency tube amplifiers. - M: NT Press. 2007.
3. Simulkin S. Secrets of High-End tube technology. - Radiohobby, 1999. No. 4, pp. 49-52.

The design of a single-ended tube amplifier using triodes is proposed by Vitaly Brusnikin, Petrozavodsk. Despite the simplicity of the design (and in many ways because of it), its sound is distinguished by good musicality. The amplifier can be recommended for the first experience of those interested in “tube” sound with a modest budget. Once you assemble this circuit, you will be amazed by the sound of your favorite recordings!

The design of the device goes back to the so-called Loftin-White circuit - a two-stage amplifier with direct coupling. The absence of an interstage capacitor (or transformer) has a positive effect on the linearity, broadband and phase response of the amplifier.

Attention. You should not start manufacturing this design without having experience working with high-voltage electrical circuits (up to 1000 V). Safety - first of all, be careful and don't put yourself at risk!

Its sensitivity is about 1 V, which is quite enough for most CD players - the most widespread source of music signals today. An audibly noticeable limitation appears at an output power level of about 8 W. But it is so “gentle” that you can continue listening by lowering the volume a little. At a power of 4 W, the harmonic coefficient does not exceed 0.4%, at 1 W - no more than 0.1%. The main product of distortion is the 2nd harmonic of the signal.

Rice. 3.54. Loftin-White “budget” scheme.

In the first stage (Fig. 3.54) one of the triodes of the 6N8S lamp operates. The lamp was chosen after listening to several types in this circuit in various modes and, in the opinion (ear?) of the author, has several advantages. Its sound signature is natural, and its stereo panorama is built with precision and depth.

If the amplifier is manufactured in the form of monoblocks, it is reasonable to use single 6C2C triodes - even with slightly better results. 6C5 (6C5C) may also be suitable.

Note. Perhaps someone will like the sound of the 6N1P lamp, which also works well in this circuit, but the author considers its sound to be harsh here.

The small automatic bias resistor is not bypassed by the capacitor, which creates a small local feedback loop with a linearizing effect. In addition, the absence of a capacitor eliminates the problem of its effect on sound, which is almost never positive.

For the output stage, also after testing, an EL34 lamp was selected in three-one switching.

Advice. Try to find it, it will delight you with elastic bass, harmonious tonal balance - a wonderful device!

There was also a domestic analogue, although not very widespread - 6P27S. 6PZS (in this case 6PZS-E and GU-50 are preferable) will be somewhat worse. With the appropriate recalculation of the output stage, you can try 6N5S (6N13S), a pair of EL84 (6P14P), 6P1P, 6P6S (all in triode connection) or 6S19P connected in parallel.

And finally, those who have 6SZZS or 6S41S lamps can experiment with them by increasing the power supply and lowering the anode voltage. Although, strictly speaking, the power of the first stage in this circuit is not enough to “drive” them.

The diagram (Fig. 3.54) shows a general negative feedback circuit (resistor R7), introduced to reduce intermodulation distortion. However, in the author's version the OS is not included.

The circuit contains a very small number of elements, but the quality of each is very important. Here we will outline some desirable “excesses”, although even in the minimal version the sound of the amplifier will not disappoint you.

Let's start with the volume control. If you have access to a high-quality discrete switch, it is better to use it by assembling an L-attenuator circuit. In practice, only 5-6 positions are enough.

Resistors R2—R5 type MLT-2. R2 performs a protective function in case of possible contact failure in the potentiometer Rl motor. R6 can be assembled from four MLT-2 in a mixed connection or use a vitrified wire with a power of 10 W. The tolerance for the values ​​of all resistors is 5%.

It is better to purchase an imported capacitor for the cathode circuit of the output lamp - only K50-35 is suitable from our parts. But the best solution would be to use polypropylene or paper (MBGCh or MBGP) capacitors with a capacity of at least 50-60 μF.

The output transformer deserves a separate discussion. If you focus on making it yourself, the cross-section of the iron should be about 10 cm2 (a little more is possible). Sectional winding - alternating 5 layers of “primary” (2500 turns, 00.25 mm) and 4 layers of “secondary” (300 turns, 00.6 mm). Make taps from the 240th and 270th turns to adjust for specific acoustics.

The author prefers the sound of transformers based on W-iron, but a strip core can also be used. Much depends on the quality of the steel - it is best to use a core from some kind of audio transformer (cinema equipment, tube radio units, etc.). The thickness of the non-magnetic gasket during assembly is about 0.15 mm.

If you use interwinding insulation made of capacitor paper, it is better to impregnate the assembled transformer with paraffin to increase the electrical strength and fix the turns. The author uses thin plumbing tape “FUM” (white) for insulation. It fixes the turns well and has the properties of an excellent dielectric. And if the transformer is wound carefully (the turns at the edges do not fall through, etc.), it does not need to be impregnated.

A few words about the installation features. It is better if all connections are as short as possible - fastening components “from point to point”. If the length of the connecting wires is less than 10 cm, their effect on the sound will not be so noticeable. Mount the cathode circuits as compactly as possible, the grounding point of all components should be close to the input connector. Connect the common anode power wire here.

Note. The author is of the opinion that only powering the amplifier channels from separate power transformers (Fig. 3.55) fully realizes the advantages of a single-cycle tube amplifier in the “delicacy” of stereo panorama transmission.

Rice. 3.55. Power supply circuit for a stereo amplifier channel from a separate power transformer.

Rice. 3.56. Option for using a separate filament transformer.

Don't skimp on the power source and save money! As a last resort, use one transformer with sufficient power reserve and separate rectifiers. It also seems reasonable to use a separate filament transformer (for example, serial voltage transformers) - Fig. 3.56. This will simplify the winding of anode transformers and solve the problem of preheating the lamp filaments before turning on the high voltage.

Powering the filaments with rectified voltage is preferable in terms of background level and sound quality. The optimal anode rectifier is full-wave (not bridge!). Such a circuit (Fig. 3.55) suppresses common-mode interference from the network, and the amplifier turns out to be very “quiet”, without unpleasant background or rumble.

For rectification, it is better to use power valves for currents above 5 A with a large crystal area. Conventional diodes, when switching, create broad-spectrum interference that spoils the sound. You can see this for yourself!

Chokes - from tube color TVs, type Dr-5-0.08. If you wish, use kenotrons in the source (5TsZS are good), MBGCh or KBG-MN paper capacitors in combination with large chokes of 10-15 H.

If you don’t mind the dimensions, this will be a truly “uncontested” source, which may be useful to you in the future when you want to try elite output tubes (6S4S, 300V and others).

As a compromise, it is proposed to bypass capacitors C1, C2 and C4 (it’s good if they are imported) with paper-oil capacitors (MBGCh, KBG) and a capacity of at least 10 μF.

Advice. In order to increase the output power to 10-12 W, we can recommend introducing a fixed bias amplifier of 2.5-3 V in the first stage.

Low amplifier power requires sensitive acoustics. Among domestic loudspeakers of previous years, small-sized acoustics with a power of 15-30 W are preferable to S-90 clones. Nowadays you can still find old cinema loudspeakers with 4A-32 drivers, they will work well.

An option with a 2A-12 and a horn HF link based on a 1A-20 head or similar - this is already serious acoustics! After a little modification, we get speakers with good dynamics, beautiful bass and high sensitivity. If your capabilities are completely limited, collect several old 4-watt heads with a paper diffuser suspension and place them in large open boxes, adding, if available, “paper” “tweeters”. 6GD-2 are good for bass. I wish you success and enjoyment of the sound.

Sukhov N. E. - The best designs of ULF and subwoofers with your own hands.

Completed work on building a tube amplifier of my own design.
Single-ended without OOS with directly heated triodes 2A3 in the output stage.
Amplifier power 2.5 Watt/channel.

This is my first project, as they say now, implemented from start to finish:
starting from the calculation of operating modes and selection of lamp connection diagrams to the organization of the engine compartment, selection of types of parts and direct installation and soldering

Background

It all started with the fact that I decided to change the EL84 connection in my STA45 tube kit to a pseudo-triode one. For such a connection, a different load was required and I began to choose a 3.5K output transformer. I chose the Daburu DA-35X, despite its Taiwanese origin, the transformer is wound on high-quality double c-core Z11 hardware. This is a special silicon steel (Silicon Grain-oriented Steel), designed for the manufacture of transformer cores with a thickness of 0.35mm, manufactured by the Japanese Nippon Steel Corp. According to manufacturers, such steel has a higher resistivity, which has a positive effect on the properties of the steel used for the manufacture of cores (reduces the occurrence of eddy currents, etc., etc.). However, the respected community already knows all this.

No sooner said than done. The transformers arrived (7 kg, not 3 per pair as indicated on the seller’s website) and then I had to state an extremely disappointing fact: the size of the transformers is too large to fit on the small steel chassis of my STA45. It was the presence of large and high-quality transformers that pushed me to the idea of ​​not remaking a small pentode amplifier, but of building a new amplifier using a real triode in the output stage.

Lamp selection and diagram

When choosing a lamp for the output stage I had few requirements: a linear lamp, capable of operating at a low anode voltage (up to 350V) and definitely a directly heated lamp. I excluded 300B from the list right away because of its high cost. In my humble opinion, a lamp is a consumable, not an investment in old age.
After some research, I quickly came to the decision that the most suitable triode for my design would be the 2A3. A small microphone effect (a big enemy of directly heated triodes), decent linearity and the ability to obtain a power of about 3.5 Watts per channel - all this spoke in favor when choosing a lamp for the output stage. The non-standard filament voltage of 2.5V did not scare me, since I initially decided to order a universal Hammond power transformer with several secondary windings designed for filament voltages of 6.3V, 2.5V and 5V. There was no question of choosing a bias for me; having chosen an auto-bias with a powerful resistor in the cathode, I moved on to the most important thing - choosing the operating point and operating mode of the output stage.

I tried several modes, including an interesting sound one 272V anode-cathode and 51V mixing(convex and dynamic sound), but due to the limiting value of dissipation at the anode more than 15 Watt I settled on the gentle mode 224V anode-cathode and 40V bias. In this mode, about 11W is dissipated at the anode and at the output it gives about 2.5W per channel, which is quite enough to sound a room of 17 sq.m. Warm, balanced, I would even say delicate sound - what else is needed from a low-power triode amplifier.

When choosing a lamp for the driver there was nothing left to do but follow the traditions of classical tube construction and choose a low-current triode with a high gain, for example 6SN7 or 6SL7. At first I didn’t want to invent something special and decided to take J.C. Morrison’s circuit from his “Fi Primer” as a driver. These are two stages on halves of a 6SN7 triode with direct coupling, it would seem that the optimal solution is one less capacitor, but after prototyping I got a high level of distortion. It should be noted that even before assembly, I recalculated the cascade modes and I did not get the values ​​that J.C.Morrison suggests in his example, but nevertheless I decided to try. Later, on English-language forums, I found reports of an error in this circuit, also known as Sun Audio SV-2A3.

As a result, I decided to remake the driver stage, turning it into an SRPP stage on the halves of a 6SL7 triode (better known in our country as 6H9C). After studying the options for various connections, I chose SRPP with a load connected to the upper terminal of the cathode resistor of the upper triode (according to the circuit, the lower point of R5). It is in this connection that the cascade works as a real push-pull amplifier. The highlight of the cascade is, at first glance, the extra resistor R6. Its purpose is to equalize the currents passing through the upper and lower halves of the triode. I took the formula for calculating its value from the TubeCad magazine website.

It is this implementation of SRPP that makes it possible to obtain maximum signal amplification at the cascade output with the lowest nonlinear distortion factor; it is precisely this inclusion that allows you to obtain a driver cascade with a low output resistance.
Next we should say a few words about the power supply.

power unit

I took a conservative approach when designing the power supply.
A rectifier based on a 5U4G kenotron and a smoothing P-filter (with an input capacitance). Since the signal also passes through the filter circuits, I also tried to get rid of electrolytes here. The first after the inductor is a 2.2 microfarad paper-oil capacitor. The quality of this capacitor greatly affects the sound and that is why I used a Jensen “copper” one here. The other two capacitors (a separate capacitance for each stage) are a film Ansar 100 uF and a large 220 uF Elna Cerafine, which is only formally electrolytic - its properties are so precise and linear. It is not for nothing that Hiroyashi Kondo used these capacitors, along with Black Gate WKZ, in the power supplies of his designs.

Design

To accommodate the amplifier components, I chose a spacious chassis.
All inductors (choke, power and output transformers) and lamps are located on top, on the top cover. Since the cover is made of steel, and I don’t have much experience in processing it, I entrusted the drilling of holes for the lamp sockets to the chassis manufacturer, since such an option was possible. Which I never regretted later.

I placed all the other parts on the back of the cover using hinged mounting. I used the “point to point” principle; petal bakelite panels helped me with this.
When choosing options for the layout and placement of stages, I sacrificed aesthetic principles in favor of pragmatic ones, placing the input stage at the back and the output stage at the front (usually they do the opposite). This placement allowed me to use short wires from the input connectors to the potentiometer, which is located next to the input stage, using unshielded Kimber mono core. I extended the potentiometer handle with a rod so that the volume control was placed traditionally on the front panel of the amplifier. The power supply and output transformers are located as far apart as possible. I tried to separate all AC and DC circuits from each other, and in case of their intersection, place them at an angle of 90 degrees. The filament circuits are made of naturally twisted wire.

Grounding

As you know, organizing the “ground” is the most labor-intensive task when designing an audio amplifier. After reviewing many solutions and reading the experience of numerous developers, I decided to use the earth bus method.

As a bus, I used a copper conductor without insulation with a cross-section of 2 mm.
This is quite enough. With a diameter of a larger cross-section, problems will arise with soldering, since it will be difficult to warm up (without overheating) such a conductor.

The bus runs through the entire amplifier from the power supply to the input connectors, where it is connected to the chassis (at one point). The “grounds” of each of the cascades are assembled mainly into mini-stars and only then connected to the bus.
There are three such points:

Special thanks members of this community, viz.



 

It might be useful to read: