Grzegorz "gsmok" Makarewicz, This email address is being protected from spambots. You need JavaScript enabled to view it.  (written in 2006)

For some time I had the intention to build an amplifier based on the famous "devil" tube 6S33S (6C33C). I heard about this tube a lot of contradictory opinions. The fundamental flaw mentioned, is the time instability of it's parameters.  Unstable anode current during operation of 6S33S tube is a constant theme in almost every discussion board dedicated to electron tubes. With all its disadvantages 6S33S tube is mysteriously attracted for audio designers. Why? I do not know. Probably anyone who used this tube had their own reasons. As for me there are also several important reasons. I will mention one - it looks very cool, and these "devil's horns" - amazing!!!

Some time ago I heard from a friend, that his friend had seen with his friend a simple tube amplifier based on 6S33S. Apparently, sounded great. This friend of mine decided to take the same amplifier and to do this has acquired the schematic of the amplifier. Schematic remained unused for two years, until it got in my hands. The diagram of an amplifier does not have any revelations. A lot od such schemes can be found on the internet. This is not a disadvantage for someone who wants to build an amplifier according to a proven idea.

A bit under the pressure of  my friend (being already mentioned two times), I decided to listen personally to 6S33S amplifier. It took me some time to gather the necessary elements, but eventually I started to work.

This description begins with a moment when I managed to run a power amplifier designed as a prototype on a piece of plywood. So, there are chances that my project will not end up on the description of the struggle with an amplifier, but I  produce its final version, suitable for placing it on the shelf.

Along with the progress in the construction, the description will be replenished gradually . I will also modify the materials already available. For this reason, I recommend from time to time not only to read the new text fragments but also check whether the earlier texts have not been significantly changed.

All visitors interested in this article are invited to TRIODA's discussion board. There I've initiated the thread about this design. It is available at:

General comments on the prototype construction

I decided to put together a prototype system on plywood. Each amplifier channel requires two separate bases - one for power supply and one for the amplifier. Together, they provide a considerable number of four coasters made of plywood.

Although it is only a prototype, but I decided to assemble it as neatly as possible. Firstly, I'm going to evaluiate a lot of research and experiments based on the prototype. Secondly, in the system we are dealing with high voltages and for security purposes, amplifier has to be assembled in the right way. Details on the implementation of the mounting base are illustrated below with the relevant photos.


As a material for the mounting base, you can choose wood or plywood. I chose the plywood, because with the same thickness (about 10mm) as wood, plywood ensures better rigidity. It is not without significance, given the fact that really heavy items will be screwed to the base. In order the mounting base to be easily moved on the table, I decided to equip it with small wheels.


I used the cheapest available weels offered by "Praktiker". They are small but reliable. The only drawback is the fact that they protrude from the side, which are screwed on to the base. So you can not attach them directly to the plywood, because that restrain them and would not enable to rotate.


But from what are the tools. Using mini-drill equipped with a small roll of sandpaper I milled cylindrical cavities for the wheels.


This is the one of the fixed weels. Fixing is done by screwing the two screws. Here reveals for the first time the benefit of easy mounting of various components to the base of plywood. In case of error in setting the given item you just need to move it and tighten the screws in a different place. Quick and easy.


And here is the  base ready to mount the amplifier / power supply.


Now it's time to prepare mounting technology of electronic components. For this purpose generally available connectors can be used. I chose ring connectors and so-called. automotive connectors.


The main advantage of connectors and the method of mounting elements with screws is that, depending on your needs you can make as many soldering pads as you want.


It is also possible by screwing a single screw to create a connecting point consisted of various types of terminals. The photo on the left presents some examples.

 Power Supply

The diagram of the power supply of the amplifier is shown in Figure below. With regret I have to admit that the power supply uses some semiconductor components.

Schematic diagram of the power supply

As you can see, the power supply is, in terms of number of electronic elements used, the main component of the amplifier. Setting aside a functional considerations, you can even say that the amplifier is only a small supplement to an extended power supply. But now, seriously. Although this is a prototype  system, it is worth spending a bit of effort to design the items connecting the power supply to the mains voltage in such a way that they are fixed firmly and safely. This applies to the electrical outlet, fuse and mains switch. Indeed, I do not advise you to mount them carelessly. The ability to quickly turn off the amplifier, without nerve exploration is really invaluable.


Socket to connect a mains cord with integrated fuse [3] and the power switch [2] will be mounted on special aluminum bracket [1]. I used a 1mm thick plate with properly cut holes.



I bent a plate in a vise,  so as to form a bracket allow it to be screwed into the base.


And here is how looks ready to mount bracket with mains socket with fuse and switch.

Proper installation of the power supply I started from the power supply input elements: the power transformer and its 'environment'. These elements are shown in the diagram.

Power transformer and "security" elements.


In the power supply I used a toroidal transformer, which I fixed with a large screw. Here I used the standard (supplied with the transformer) fasteners in the form of rubber washers and metal plates tightened the transformer to the base.


Now, it is time on the so-called "soft start" system. Although the toroidal transformer has a relatively low power, but given that it will be frequently switched on and off, I decided to assemble this system just in case.


Time for another element. I screwed on to the base aluminium bracket with mains socket, fuse and switch described earlier.


Here is a view of the transformer and associated elements. Yuo can see the wires of a soft start. This also illustrates how the bracket with a switch and socket is securely fastened do the base (three screws with wide heads)


Time for wiring of mounted components. After connecting a switch and soft start circuit to the transformer I soldered all the leads of the transformer to terminals screwed on to the base. Because it is not known how long the winding terminals will be needed in the future I didn't cut them to the dimension, but bent all of them in the form of a loop (hence the nicely curved colored leads).
After that phase of wiring I made measurements of the voltages on the transformer secondary windings. They were (numbering according to the circuit diagram):

Winding I: 8.65 V
Winding II: 13.3 V
Winding III: 82V
Winding IV: 335V
Winding V: 196V.


Here are some details of wiring mains socket, switch and soft start circuit. I used a combination of slides, which also are protected by isolating shields. This way you can not accidentally touch the "hot" wire by hand.
Safety above all!


Here is the enlarged photo showing the details of how the transformer secondary windings were soldered to connectors.


Now it is time for the latest heavy element of power supply - the choke [DL1]. In this way, on the base I have all the parts needed fixing before the implementation of other time-consuming mounting activities.


Before assembling of individual sections of the power supply I mounted all bridge rectifiers. They are marked, according to the circuit diagram as M1, M2, M3 and M4.
Voltage measured at the outputs of bridges (no load and no-capacity filters) are:

M1: 10.3 V
M2: 73V
M3: 316V
M4: 186V.


This photo shows installation details of bridges M1 and M2.


And here a diodes forming bridges M3 and M4 are shown.

Time for assembling of individual sections of the power supply. We start from the power supply providing filament voltage for the tube in the input stage of the amplifier. Heater power supply diagram is shown in the figure below.

Heater power supply for 6N1P tube

Parts list:
   M1 - KBL04 bridge,
   U1 - 7806 regulator,
   C1 - 4700µF/16V
   C2 - 1µF/63V
   C3 - 4700µF/16V
   C4 - 1µF/63V

The photo shows all the elements giving the output AC filament voltage U1. Marking of elements is in line with the schematic diagram.

No-load voltage measurement showed that the voltage across the capacitor C1 has a value of 10.3 V and the voltage U1 is equal to 6.08 V.

Now, on to the next section, the heater supply of 6S33S. Filament voltage is taken directly from the winding number II of the power transformer. It has been brought to the output terminal U2 with double twisted insulated wire. Furthermore I fix this wire to the plywood at two points [P1] and [P2] with a glue gun.

Trzecia sekcja zasilająca to napięcie służące do wstępnej ujemnej polaryzacji siatki lampy 6S33S. Schemat zasilacza pokazany jest na rysunku.

The third section is a power supply voltage used for the biasing (initial negative grid polarity) of 6S33S tubes. The diagram of the circuit is the figure below.

Schematic diagram of the biasing 6S33S section

Parts list:
   M2 - any bridge rated to 200V/1A,
   T1 - BD244,
   DZ1 -90V Zener diode,
   R1 - 4K7,
   R2 - 4K7,
   R3 - 4K7,
   PR1 - 4K7,
   C5 - 100µF/160V,
   C6 - 10µF/160V,
   C7 - 10µF/100V,
   C8 - 1µF/100V.

This is how biasing section has been assembled. The individual elements are marked in the same manner as in the schematic diagram.

Time to relax ;-). This is the mounting base on which we have three voltages (U1, U2 and U3) necessary to power the amplifier circuit.


The fourth section is a 6N1P tube anode voltage supply. The circuit diagram is shown in the figure below.

Schematic diagram of the anode voltage supply of 6N1P input tube

Parts list:
   M3 - Four BY55 diodes bypassed by  200pF ceramic capacitors,
   T2 - MJE13005,
   D1 - 1N4007,
   DZ2...DZ5 - 100V/1,3W Zener diodes,
   R4 - 20/5W,
   R5 - 10K/2W,
   R6 - 4K7,
   C9 - 220µF/500V,
   C10 - 100nF/1000V,
   C11 - 220µF/500V,
   C12 - 100nF/1000V,
   C13 - 22µF/450V,
   C14...C18 - 100pF.

This is the anode voltage supply mounted on the base. Marking elements are compatible with circuit diagram.


Electrolytic capacitors C9 and C11 were fixed to the base with glue gun. The next photo showsl assembly details.

The fifth and final section is an anode voltage sipple of 6S33S output tube. The diagram is shown in the figure below.

Schematic diagram of the 6S33S anode power supply

Parts list:
   M3 - Four BY55 diodes bypassed by  200pF ceramic capacitors,
   Dl1 - 10H/300mA choke,
   C19 - 680µF/350V,
   C20 - 0,1µF/1000V,
   C21 - 680µF/350V,
   C22 - 0,1µF/1000V.

This is the 6S33S anode power supply. Capacitors C19 and C21 were made with four capacitors in parallel each.

After running all sections of the power supply, connect the grounds to one common point. This will allow you to easily connect a common ground point of the power supply of the ground point of the amplifier. I don't advise you to use of multiple grounds joined at various points (such an approach often leads to problems with noise interference). How to connect grounds of the power supply sections is shown in the photo. Grounds are conducted via the cable in the green insulator.

After tests conducted with the power supply I introduced three small changes to the initial design. They are not necessary, but I briefly describe them.


Modification no 1

Because I used quite a small heat sink to reduce the power losses in the 6N1P-EW heater voltage stabilizer, I put a resistor R with a value of 1 ohm between the bridge rectifier and stabilizer U1.


Modification no 2

On TRIODA discussion board there is a thread about the role of bypassing of high voltage capacitors in the power supply for security reasons. Although I knew about it, I needed a stern 'kick' to appreciate the fact that the voltage on the capacitor can be dangerous even after a long time after switching off the power supply. So I bypassed 6N1P-EW anode voltage power capacitors, with 200K resistor (item labeled as R)


Modification no 3

For the same reason as above I bypassed electrolytic capacitors in the 6S33S anode power supply (resistor R of 100K).


Diagram of the amplifier is shown in the figure below. This is a"minimalistic" design contains only two electron tubes per channel.

Schematic diagram of the amplifier

The table shows the components used. During the amplifier tests, some values have been changing and so is going up to now. There are not large differences, but I want the person building the amplifier to has this in mind during collecting of the elements.. I try whenever possible to show in the table below the most current element values.

With power resistors in brackets [] the approximate value of the voltage that builds up on their ends is written. This allows for easy selection of required power.

Designation Value Designation Value
C1 0,1uF/1000V (MKP) R4 470KΩ
C2 0,1uF/1000V (MKP) R5 560Ω
C3 0,1uF/1000V (MKP) R6 28KΩ (2x56KΩ in parallel) [155V]
C4 0,1uF/1000V (MKP) R7 1KΩ
C5 100uF/400V (electrolytic) R8 130KΩ
C6 0,1uF/1000V (MKP) R9 220KΩ
C7 100uF/400V (electrolytic) R10 not necessary
R1 470KΩ R11 110K [130V]
R2 2K2 RP 1Ω/5W non-inductive
R3 100KΩ [120V] P1 10KΩ (potentiometer)

The best way to start assembling of the amplifier is planning the layout and fitting of the largest components. In this case, I started with the output transformer [TG1], ceramic socket for 6S33S [V2] tube, and socket for 6N1P [V1] tube. In addition (as shown in the photo below) I mounted the input socket [WE], output socket [WY] resistor [Rp] for measuring the 6S33S anode current and solder terminals, which are connected to the amplifier supply voltage [U1] [U2] [U3] [U4] and [U5]


This is the base for the amplifier from the bottom side, after fitting sockets, transformer and tube sockets. Visible are also four wheels enabling it to move across the surface without lifting.


On the following photos  details of the aforementioned fastening elements are presented.

In addition to the input socket [WE] you can see  the tube socket [V1].


The ceramic socket of 6S33S  tube [V2] is leaned on four aluminum spacers.


Here's how output sockets [WY] and output transformer [TG1]  are screwed to the base.


After verifying that the 6S33S tube socket is properly mounted I disassembled it and soldered all elements "cooperating" with the tube. These are:


[A] - anode voltage supply wire

[Ż] - [Ż] - filament wires

[K] - cable to connect the cathode

[R7] - grid resistor



After soldering the items to the 6S33S socket I fixed it back to the base. Now it's possible to install the remaining elements.


A similar procedure was applied to the 6N1P tube. Before attaching the socket to the base I soldered all the elements and the connecting cables. The photo shows a view of the tube socket from the bottom. Markings of the elements are compatible with circuit diagram.

After screwing tube sockets I made final wiring of the amplifier. In the photo (below) all the details necessary for assembling the amplifier are given. In addition to the elements soldered ditectly to sockets' connections, parts forming the input stage anode voltage filter were placed (C4, C5, C6, C7, and R11). Global negative feedback resistir (R9) is not connected to the secondary winding of transformer. You can do this during the launch of the amplifier.


And so, here is the amplifier before inserting tubes into sockets ...

  ... and after inserting the tubes.

Before the "firing" of the amplifier remains only to combine it with power supply. Below is the fully assembled amplifier channel (the power supply on the left and amplifier on the right).

Examples of measurement results

I include here the results of successively complemented measurements of the amplifier. For now, these are merely the results without the detailed analysis. I'll try to add it at the end of experiments (read: probably never).

Amplitude and phase characteristics in the frequency range 0-25,6 kHz. Readings for the frequency of 20kHz. (amplifier with output transformer type I, the system without correction and feedback, the input signal - white noise of 10mV, the anode current of 6S33S tube Ia = 155mA)

Amplitude and phase characteristics in the frequency range 0-50Hz. Readings for the frequency of 13Hz. (amplifier with output transformer type I, the system without correction and feedback, the input signal - white noise of 10mV, the anode current of 6S33S tube Ia = 155mA)

Nyquist diagram (amplifier with output transformer type I, the system without correction and feedback, the input signal - white noise of 10mV, the anode current of 6S33S tube Ia = 155mA)

The input signal (upper graph) and output (bottom graph), (amplifier withoutput transformer type I, the system without correction and feedback, 1kHz sine wave input signal, P1 Closed / 0Kohms, 6S33S tube anode current Ia = 155mA)

The input signal (upper graph) and output (bottom graph), (amplifier with output transformer type I, the system without correction and feedback, 1kHz sine wave input signal, P1 around 10K, the anode current of 6S33S tube Ia = 155mA)

Harmonic distortion (amplifier with output transformer type I, the system without correction and feedback, 1kHz sine wave input signal of 10mV, 140mV output signal, P1 around 5K, the anode current of 6S33S tube Ia = 155mA)

Harmonic distortion (amplifier withoutput transformer type I, the system without correction and feedback, 1kHz sine wave input signal of 62mV, 700mV output signal, P1 around 5K, the anode current of 6S33S tube Ia = 155mA)

Harmonic distortion (amplifier with output transformertype I, the system without correction and feedback, 1kHz sinusoidal input of 128mV, the output signal 1.41 V, P1 around 5K, the anode current of 6S33S tube Ia = 155mA)

Harmonic distortion: 1 kHz input signal

The prototype on a steel chassis


Elaborated by Grzegorz "gsmok" Makarewicz,