My 6SN7 Based Preamp

Always wanted to hear how the big bottles sound. Recently I build my version of 6SN7 preamp. It actually took many months in tuning.  Finally a CCS was added to the cathode follower to replace the cathode resistor. Although no significant improvement in measurement but great improvement in listening. It results much deeper and richer sound in mid and lower frequencies range. Here is the circuit topology 

Here is the LTspice FFT chart: 

THD and THD+N:

Proof of concept, components are point to point soldered  

As the cathode voltage of the cathode follower has reached around 170V, there is a need to lift the filament voltage to around 150V,  therefore two separate filament supplies are required. As in picture, 2 sets of filament supply circuits are around the transformers. They are not regulated but heavily filtered to supply the required DC for the filaments of three tubes. Both AC supplies are snubbed before bridge rectifier. 

HV from transformer is also snubbed before full ware rectifier and then feed to a R-C-R-C filter , then to the shunt regulator board. +300V of regulated DC output with 45ma current is adjusted to power the preamp. Actual measurement of residual noise is about 0.25mv.  That ends the proof of concept build. 

PCB design is in progress. 

Update: 

It has been quite a while since. Well,  I did not stop this project at all. Finally the sample of production version has been made. There are some minor updates in the final PCB layout. 

PCB with 2oz copper and 2mm thickness

Components populated

Back side

Finished 

Tubes Used: 

1X 6SN7 GTB  Sylvania  coin base  NOS

2X JJ 6SN7 new production 

Finally, some real time measurements done by REW and Focusrite solo 2i2

                                                                              

As can see, the frequency response is about 35hz - 24khz @-3db

THD= 0.061%,  THD+N=0.062%  @ 0.775V output ,  second harmonic is dominating.

(LTspice result was THD=0.042% and THD+N=0.052% at the same output level. ) 

Also, as can see,  the 50hz component is about -106db which equals to around 10uv. Also measured the output residual noise using an analog millivoltmeter, the reading is 0.04mv RMS 

Have to say, the PCB version measures much better than the "point to point "soldered version. 

Listening test setup : 

Source:  Singxer  SU6 USB interface + Rockna Wavedream R2R DAC

Power Amplifier:  Audio-GD C1 

Speakers: Sonus Faber Amati Homage 

This speaker has 92db sensitivity, but when stick my ears very close to the speakers, it's very very quiet ! :) 

Tube Pre-amp CT-9H

I have gone though 2 upgrades since the release of the CT-11H preamp. However, I finally decided to create a new model based on the CH-11H with all the upgrades implemented and some changes in the actual circuit design. It's amazing because a very low level of distortion is achieved by LTspice simulation.

 

In the new circuit, the cathode follower now has a cathode resistor connected directly to ground and feedback is taken from the grid of the cathode follower with an added capacitor in the loop. Also made change is the output from the second stage which is now taken from the cathode of the top tube ,  making the second stage a typical mu-follower. The top and bottom tubes for the mu-follower are now separated in two triodes by the new PCB design. An ECC88 is employed for the cathode follower instead of 6BQ7A,  running at around 8ma current. In the chats, THD is 0.000061%, and second, third harmonic distortion are all below -120db.

Although simulation can hardly reflect exact the same thing in reality, but simulation results for the original circuit were never so good. Listening experience also tells me that it's absolutely worth the change. The sound is much more refined and I can feel more timber/body like with piano. Bass is also fuller and richer.

The main PCB has been re-designed in order to reflect all the changes and improvement.

UPDATE: 

Some reality measurements using ARTA software and Focusrite Scarlett 2i2 USB sound card.

Frequency Response

FFT Chart

2nd/3rd Harmonic Distortion 

At -3db output, second harmonic distortion is -80db, the third is about -100db. then the rest of  are all less than -100db. This could explain the preamp sounds very natural, no harsh across entire audible range. 

----------------------------------------------------------------------------------------------

Tube used 

3X  JJ802S new production

1X  Beijing 6n11  NOS

1X  6X4 Sylvania NOS

Filament Arrangement:

12.6V regulated for 2 X  JJ802S

6.3V   regulated for 6n11 and JJ802S

Output Noise level

0.1mv~0.12mv when input shorted to ground

Gain:

21.5GB

Power Input:

AC: 230V~240V

Connectivity:

3 set of RCA for input signals.

1 set of RCA for output (output signal can be muted by a mute switch in the front panel and automatic delayed for appox. 100 seconds on start-up)

For more info please contact peter@consoletronics.com









The Valve Phono Stage PH-11

It has been a long time that I've been thinking to make a nice sounding phono stage. The challenge in design and building of it is how to handle the power supply ripple and layout of PCB to minimise output noise level.  I auditioned a few brands of SS and tube phono stages and finally I concluded that my ultimate favour for the SQ goes to tube design.

 

Design like Ear 843P that has 3 stages with RIAA network in the negative feedback loop sounded a little muddy to me. Circuit with no global feedback but consists of passive RIAA network probably is the better way to go. After some searches,  I found a promised looking circuit designed by Bruce Heranat at DIY Audio Projects site. The input stage is a symmetrical SRPP which feeds a passive RIAA network. The second stage is almost the same as the first one but it is DC coupled to the cathode follower that is the output stage. The gain is about 45db, this is sufficient for most of the MM cartridges. Using a pair of Lundahl LL1678 step up transformer adds extra 24db gain to suit lower output MC cartridges. The circuit has no global feedback.

Power supply is quite critical for the sound quality as well as the final output noise level of the phono stage. Salas SSHV2 shunt regulator has been my favourite in the past for its clean and natural sound signature. Its low impedance and ripple voltage across a very wide frequency range makes it very suitable for this project. In order to achieve even lower noise level, the power transformer is placed in a separate chassis and the source of filament supply is connected from the transformer chassis to the main chassis. It's then rectified and pre-filtered in the main chassis before feeding to the actual power board.  I found this has significantly avoided hum radiation during its journey to filament. It's again filtered by 3 Nichicon caps in parallel with total 10K uF of capacitance, and finally gets to the LT1085 for voltage regulation of filament.


Switching between MM /MC input by relay could cause huge popping sound due to high gain amplification. Output must be automatically muted before actual switching occurs. A new logical control board is therefore designed to achieve that. Also added on the board are delay /mute and Mono/Stereo buttons. Mono button is useful when listening to mono LPs.

 

 

Main Board

 

 

SHHV2 with Filament Supply Regulator

 

 

Back Side

 

 

Control Board

 

Mono/Stereo Relay & Connector for Control Board

 

 

Assembling

 

 

Assembled

 

 

Another View

 

 

I/O Board with Lundalh SUT

Front Appearance

                                          

Actual RIAA Curve Messured

 

 

 

 

   MM Gain : 45db

     MC Gain :  24db

Noise Level : 0.2mv

Tube Used: 1X JJ ECC81, 4X JJ ECC803S Gold Pins

 

Onboard Loading Selection via DIP Switch for MC Cartridges

40, 60, 100, 150, 200

Extra Set of RCA input is provided for adding more loadings when using external SUT

 

                         

                             

For more info or an arrangement of trial, please contact peter@consoletronics.com 



 



 



Tube Preamp CT-11H based on CAT SL-1 ( updated )

After the successful JM-70H preamp, my next project is a bit challenge one -- CT-11H.
It was a fellow audiophile friend of mine who brought his beloved preamp for me to repair one day, which had a burnt 6X4 rectifier. The preamp was bought from eBay and it's actually a shrink clone of  the famous CAT SL-1. He had all the coupling caps upgraded and valve tubes replaced with NOS ones. I had it in my system and listened for a few days after having it repaired. What it amazed me was its  unique sound to my ears, that makes me very attempting to build a version of mine.

After some search on Google, I found members over forums at DiyAudio.com had discussed the original circuit intensively and they pointed out that there were actually some mistakes with the circuit design, but not known if these faults were left on purpose or ignorant. Anyway, the computer simulation does show improved distortion figure after the faults being corrected.

The design of my version on power supply part is emphasized using two dedicated power  transformers, one for high voltage supply only and the other for the filament supplies.  Starting with Sylvania 6X4 rectifier, then followed by the SALAS SSHV2 shunt regulator, the filament and high voltage circuits are able to fit in the same board as intended:

Using the same chassis as CT-11H used, the whole layout looks like this:

A close-up look at the main board:

Finished wiring :  



 Front view :



Back View :



UPDATE : 2/1/2019

MK1 version is developed due to gain changing. I found 26db gain is a bit too much for some of the power amplifiers. Due to DC coupling from second stage to output stage, and the first stage's bias is directly determined by the feedback resistor, I ended up re-biasing all the stages, changed the cathode capacitor value for the lower tube of the u-follower, added 47P silver mica capacitor across the feedback resistor.  Finally, the gain is about 21.5db. 

UPDATE: 20/02/2019

Can I call it MK2 now?  As a result of continuously tuning of this preamp, I found 12AX7 tube is not the best for the lower section tube of u-follower, because tubes with high u do not work well at low anode voltage. Tubes like 12AU7, 6DJ8 are way better for this task. In order to replace 12AX7 with say 12AU7, both Rk for upper and lower tubes as well as Rl need to be re-calculated. The input stage is also optimised again to achieve better linearity. Lower plate resistor in the first stage improves high frequency response when connected to the second stage. New operating point is  Ia=1.76ma, Ua=160V, Ug=-8.5V, Rl=51k. Pictures below are LTspice simulation results. After completion of all the changes , the preamp sounds much richer , clearer and more detailed to my ears.

 

Distortion at 1.7 Vp-p /4khz output

FFT chart at 4khz

Frequency response

Tube used:

3X  JJ 802S   New production
1X  6BQ7A  Toshiba NOS 
1X  6X4 Sylvania black plate  NOS

Filament Arrangement :

12.6V regulated DC for 3X JJ802S
6.3V regulated DC for 6BQ7A

Output Noise Level:

0.1 mV when input shorted to ground

Gain :  21.5db

Power Input:

AC: 230V -240V


Power Consumption: 45W

For more info or an arrangement of trial, please contact peter@consoletronics.com

JM-70H The Valve Preamp with Shunt Regulator

Started a year ago, following the first successful project JP-200 clone valve preamp,  another valve preamp JM-70H has been created. It took more than 10 months in the final tuning session. During the period, quite a number of experienced audiophiles were invited for serious listening and many suggestions were received in regards to the sound quality of this preamp.

The design is based on the legendary Marantz 7 valve preamp with completely new designed power supply section. JM-70H employed two EI transformers supplying more than enough AC power to two separate SSHV2 shunt regulators, one in each channel for high voltage supply. As a result, the preamp sounds much more transparent, detailed and energetic than any of those with CRC/CLC or serial regulator designs. JM-70H is a preamp that has achieved a level of performance that can be favorable for many experienced audiophiles for the cost of an entry level preamp of big names.

From PCB design to selection of components, JM-70H is not compromised at all. Quite a number of expensive components are used , such as Amrg, Kiwame carbon film resistors, and Mundorf Evo oil , Wima MKP capacitors. All PCB used are heavy duty 2.0mm thickness with 2OZ copper FR4 boards. The chassis is fully made of aluminum with the front panel thickness being 8mm. All panels are completed by CNC machine.

Basic Specification

Tubes Used: V1:  Tungsol 12AX7 Golden Pins,  V2/V3:   JJ 5751
Filament Arrangement : 6.3V AC
Input:  3 sets of RCA for CD/DAC/AUX signals
Output: 1 set of RCA with automatic delay on start up or muting manually by push button on front panel.

Gain: 20db
Output Noise Level:  0.08mv (when input shorted to ground)

Power:  230V/240V
Power Consumption: 40W

For more info or an arrangement of trial,  please contact sales@consoletronics.com

We are Sydney based audiophiles in electronics business.


SSHV2 Shunt Regulator:

 

 

Main Board

 

 

 

  Input Module

 

 

Top view of JM-70H

 

 

 Back View of JM-70H

 

 

 Front View of JM-70H

 

Packaging