back to

To get more information contact me at:

Solid State Amplifier 2022 


After the good experience made on hybrid amplifiers with circlotron configuration (see the various version of the my Amplifier End) I thought about making a completely solid state amplifier of this type.
I have analyzed various schemes published online mainly those on (pages) but then I have decided for a design vey near to the my DC Circlotron
It is the result of many months of simulations, instrumental testing and listening test.
This design is inspired to the Circlotron US patent n. 4229706 by James W. Bongiorno and to the Thorens TEM 3200 by Frank Blöhbaum.
I have used the LTspice free simulation tool to design and optimize this circuit.
This is not a common solid state amplifier with lot of global feedback, there are only 6dB to reduce output impedance and to linearize the voltage stage.
The my aim was to make a solid state amplifier that sounds like a tube without feedback so a correction factor of only 6dB was added.
Now my Amplifier End has a real rival and in fact this amplifier with complex signals it is better and even if it is slightly less soft it is more neutral, recommended with analog sources or hi-resolution DSC DSD. 
The distortion at maximum power largely depends on the limits imposed by the power supply voltage chosen for the final stage so you can increase it without any changes to the circuit.
Like all my amplifiers also this can and should be driven directly by a CD, DAC or Phono pre-ampl.

The main benefit of a Circlotron design is an high power amplifier with no switching output stage like a pure class A but higher efficiency, with only 1A bias current still 1mA with 2.3A on 8 load.
You can increase the output power to 60W on 8ohm changing the chassie and increasing the power supply voltage of the output stage.


Pout = 50W on 8ohm with 2 x 33VDC and Vin=0.42Vrms Thd=0.6%
Pout = 90W on 4ohm with 2 x 33VDC and Vin=0.42Vrms Thd=1.3%
Pout = 18W on 8ohm no-switching with 1A bias (see simulation)
Pd = output stage at 35VDC = 70W each channel with 1A bias

Zout = 178mohm with the Exicon (see simulation)
Ft(-3dB) = DC - 2MHz at -3dB
Global feedback = 6dB
Output DC offset about 50-90mVDC (because there is no active DC serve)

Please do not touch with multimeter probe the mosfet gate after the power on because this operation will destroy the mosfet.




These are the main characteristics:

 - dc coupled of all stages
 - no capacitors on the signal path
high slew-rate
 - good damping factor near to a solid state amplifiers
with high feedback
 - no switching output stage like a pure class A
 - high efficiency because only 70W of consumption per channel
 - only single ended stages
 - only a single output device

only a transconductance amplifier with I/V
 - global feedback easy to modify without high frequency compensation
 - full balanced circuit but it accept also un-balanced input signal

As my other Circlotron amplifiers the ECW20N20 mosfet are the first choice to get a good damping factor, a low distortion, a high output power also without global feedback and it is not necessary a thermal compensation of Vgs.


This is the last pcb version


J1,J2,J3        J113D74Z
T1,T2           BC546ABU
T3,T4,T5,T6     BC556B
T7,T8           MJE340
U$1,U$2         ECW20N20-S or ECW20N20 or 2SK1058

R1,R2           100Kohm 1% 1/4W    Vishay Dale RN55
R3,R4,R7          1Kohm 1% 1/4W   Vishay Dale RN55
R8               500ohm 1% 1/4W   Vishay Dale RN55
R5,R6             47ohm 1% 1/4W   Vishay Dale RN55
R13,R14          470ohm 1% 1/4W   Vishay Dale RN55
R9,R10,R11,R12   15Kohm 1% 1/4W   Vishay Dale RN55
R17,R18         3900ohm 1% 1/4W   Vishay Dale RN55
R15,R16          10Kohm 1% 1/2W   Vishay Dale RN65
R19,R20         2700ohm 5% 2W     Vishay PR02
R21,R22         2000ohm 1% 1/4W   Vishay Dale RN55
R23,R24   120 or 240ohm 1% 1/4W   RN55 bias  120ohm with ECW20N20  
R28,R29          900ohm 1% 1/4W   Vishay Dale RN55
R25,R31           1Kohm multi turn trimmer Bourns 3296W-1-501LF             
RGND             100ohm 1% 1/4W    Vishay Dale RN60 (not necessary for dual mono conf.)
RGROUND           10ohm 5% 2W    Vishay PR02 (see photo)         

C1,C2            100uF 50V UFG Nichicon
C3,C4              1uF MKP or MKT
C5,C6                  not used
CIN                1uF MKP or MKT to use in parallel to R2 if only un-bal input is used

F1,F3             FUSE 5A FAST
KK1,KK4        HEATSINK SK104 13.4C/W  513002B02500G (cod. MOuser 532-513002B25G)

The connections are 63862-1 (CUT STRIP) by TE Connectivity / AMP (cod. Mouser  571-63862-1-CT, cod. RS 718-7987)

CZ                 0.1uF        Vishay  F339X1
RZ                10ohm 2W   Vishay PR02

The driver stage need a regulated power supply module to have a stable +36VDC and -36VDC.
The normal linear regulator does not support 40V input so I have design a shunt regulator following the application note in the datasheet of TL431 and I have optimized this circuit with LTspice.
In the pcb there are the pins to connect all the secondaries of the r-core transformer and also to get the DC protection power supply.

D1,D2,D3,D4             UF5407
C1,C2                   4700uF 50V +105°C Nichicon 20 x 40mm
C3,C6                     10pF
C4,C5                      1nF
R1,R2,R3,R4              47ohm 1W  (use only R1 and R4 for dual mono configuration)
U$1,U$2                  TL431
R5,R9                    1Kohm 1% 1/4W
R8,R11                  15Kohm 1% 1/4W
R6,R10                 1100ohm 1% 1/4W
R7,R12                  100ohm 1% 1/4W
Q1,Q4                    BC556
Q2,Q3                   MJE350
C7,C8                     22uF 50V UFG Nichicon

KK1,KK4       HEATSINK SK104 11C/W  513102B02500G (cod. MOuser 532-513102B25)




The main problem of this project is the fet selection necessary to get the max performances.
This design using a little feedback cannot compensate the large differences between two fet J113 therefore it is necessary to buy more pieces and make a selection for a matching.
Below I report 2 measurements made with a different selection of these components.

At this time I am testing the Peak DCA75 Pro semiconductor analyser to check the J113 differences but I will give a procedure for making a selection without purchasing this special tools.

Here the plot of measurements of 3 different J113 fet.

 Here the plot of measurements of matched J113 fets.

An alternative cheap method to match the fet is this procedure:

  • before solder output mosfet

  • add a jump on R21 and R22 (see on photo yellow wire)

  • set R8+R25 = R7 = 1Kohm

  • power on driver stage

  • set R31 to have 0volt on R19

  • using Arta or other software generate from teh pc a 1KHz sine wave with a level about 0.42Vrms

  • measure the driver output on R19 and R20

  •  change J113 until both the output give 10Vrms



  1. 1. Verify the power supply voltage on driver power module +36VDC and -36VDC
  2. 2. Set both trimmers in the middle, both the end of the path can be identified by listening to the slight click
  3. 3. Connect a multimeter at the connection point between R23 - R28 and ground
  4. 4. Give the power supply only to the driver stage
    5. Rotate the R31 trimmer until you read a voltage near the 0v
  5. 6. Connect a multimeter at the connection point between R29 - R24 and ground
  6. 7. Rotate the R25 trimmer until you read a voltage near the 0v
  7. 8. Repeat again point 5 and 7 until you read about 0v on both these points
  8. 9. add a 0.1ohm in series to the mosfet power supply so on the connection before one fuse to very the bias current of output
    9. Connect the power supply to the output stage
  9. 10. Turn the trimmer R31 to read 0V on terminal OUT1
  10. 11. Turn the trimmer R25 to read 0V on terminals OUT1 and OUT2
  11. 12. Check again both these points and modify the trimmer position if necessary
  12. 13. Verify the bias current measuring the voltage across the 0.1ohm resistance



Frequecy response with Exicon mosfet about Ft(-3dB) = 2.5MHz


Follows the simulation of the distortion at 23Vrms on 8ohm 66W with 2 x 36VDC, about 0.8%.


Simulated noise on power supply module output.



ECW20N20 version with driver +36VDC and output stage at 2x 33VDC and 6dB global feedback

Follows some measurements using un-balanced input signal.



2SK1058 version with driver +-35VDC and output stage at  2 x 36VDC, load 8ohm and 6dB global feedback.

Follow the distorsion at 17Vrms on 8ohm = 36W, simulated Zout about 0.5ohm.


It is possible to use any trasformers from 20V to 30V with a power about 200-300VA but in any case 2 separated secondaries are required for each channel.

The my choice is a RS Pro Encapsulated Transformers 2 x 25V  225VA (cod. RS RS123-4028)

The transformer for the driver stage is the LITE R26-10 ( an r-core 30VA with 2 x 9V 3A and 2 x 18V 0.5A



Following the experince of my last amplifiers I have used Schottky Fast Soft recovery diodes in any section of the power supply but you can use normal 36A diode bridge like the IRF 36MB60.

To create a bridge with these TO220 diodes I have created a simple pcb.

Noise on power supply module output.



In order to dissipate all the heat generated by this amplifier in my case I chose this container by HiFi 2000.

Dissipante 03/300B 3U 10mm SILVER or BLACK
Product Code: 1NPD03300B or 1NPD03300N

Used for the 3U 300mm deep chassis temperature coefficient 0,40 C°/W

2 x 33V  1A  = 66W      66W * 0.4C°/W = 13.3C° + 25C° = 38C°

Inner baseplate for Dissipante 300mm
Product Code: 1BASEPD300



Any serious solid state amplifier need a protection circuit because a fault on output transistors or mosfet can destroy the loudspeakers.

The Cirlclotron design never will produce an high dc output but I suggest to use a dc protection circuit.

I have decided to use 2 x AIYIMA 2.0 Digital Power Amplifier Speaker Protection Board Delay Relay Speaker Protection available on Alixpress online shop. 

This module use 2 optoisolator
PC817 for each input and are necessary only some little changes to increase the start-up time.

It is not necessary modify this module with 220Kohm and diode.




I am using this module got on Alixpress because it incluse a soft-start and a temperature protection at 75°.

1NO1NC Momentary switch from Alixpress


The first photos  are relative to the test environment.

Here some photos of the test environment to compare the switching power supply to traditional.

Here some photos of the final environment



Qty model brand unit price (euro) total price (euro) shop
4 ECW20N20 Excicon 12 48  
4 10000uF 50V Nichicon 6 24
2 Toroidal transformers   60 120  
1 R-core transformer   40 40
16 Diodes for the bridge On-semi 2 32  
1 Chassie Hi-Fi2000 170 170
2 DC protection module   10 20
1 Soft start   10 10
1 connectors   36 36  
1 other items   50 50  
      total 550