DIR9001 to CS8412 Converter BD

October 11th, 2007

This converter board is to replace CS8412 in my DAC project D1.  This is designed to be used in the D1V2, D1V3 and D1V33 versions.  There are other DIYer who is interested to use this to replace the CS8412 in their DAC.  I have tested that it can be used to drive SM5842, SM5847, PMD100 and TDA1541 etc…  Take note that even through it is working fine but the output logic for DIR9001 is only 3.3V while all the old ICs are at 5V logic.  ACT logic should be used if 3.3V logic is needed to buffer up to 5V.

Another important note is the EMPH (pin 3) output in DIR9001 Converter BD is High when the data has pre-emphasis data added.  In a CS8412 or CS8414 chip, the pin 3 is LOW when the data has pre-emphasis data added.

The major improvement in DIR9001 is the lower jitter of 50ps (200ps for CS8412/14).  The improvement is very obvious in sonic performance of any DAC.

Sorry all sold out!

DIR9001 converter board Manual

DIR9001 converter bd BOM

9001 to 8412 V3 Circuit Diagram

F4 – Amplifier

August 8th, 2007

This is my F4 amplifier completed in August 2007:

  1. I designed the PCB and Casing myself and get China factory to custom make for me.  The casing size is 350mm (W) x 420mm (D) x 200mm (H) with front plate 10mm larger than the casing.
  2. R3 and R4 still 22 ohm, and Q1 and Q2 are using 2SK170BL and 2SJ74BL for matched Idss to within 9.8 to 10mA.
  3. C1 to C4 use 330uF 35V Panasonic FC, gate resistors R10 to R15 are 100 ohm, Source resistors R16 to R21 are 0.51R 3W.
  4. IRFP240 and IRFP9240 are matched Vgs within 5mV which give variation of bias current within 5% (measured voltage at 0.51R 5% resistor) among all the mosfets.
  5. Use two 18V x 2 250VA transformers and 4 birdge rectifiers to parallel the secondary voltages to about +/-21.8Vdc with 1.45A bias for each channel.
  6. Filtering capacitors are Nichicon 68,000uF 63V x 2 + 0.47R//4 pcs + Mallory 33,000uF 35V x 4.
  7. Heat sink size is 420mm x 200mm x 50mm for each channel (side).  Temperature rise is about 17 deg C (i.e. 47 degC for room temp of 30 degC).
  8. The DC offset is stable to be within 10mV after warm up for 1 hour.
  9. Measured distortion is 0.02% at 1W output for both Channels.  Noise is mainly at low frequency band from 25Hz to 1kHz.
  10. The F4 amplifier has been playing for 4-5 hours a day in the pass week and the sound is already stable.  The sonic is skid to soft type bass (due to limited power) and clean and clear mid and high frequency.
  11. Overall the sound is very good and satisfying.

Some photos below:

FFT measurement at 1W and 20W:

F4 FFT2 1W 8ohm

F4 FFT 20W 8ohm

Jfet IV Board

June 8th, 2007

* * Sold OUT in 2009 and no more selling * *

This Jfet IV board was released in 2007.  The purpose of it is to modify a DAC or CD player IV converter.  It can be used for most current output DAC chips like PCM63, TDA1541, AD1862, PCM1702, PCM1704 etc…  NOS DAC is not recommend to use this Jfet IV due to excessive high frequency noises.

During the time of release, 2SK170 is used as the active device for IV.  It is now recommended to use 2SK369BL as the active device due to its higher gain around 40mS for better result.

Jfet IV Circuit


Ipod LM1876 amp

May 27th, 2007

This is a photo type for a Ipod amplifier.  It use 6DJ8 with supply at about +44V as input buffer and then drive a LM1876 IC power amplifier.


All Jfet Op-amp – not completed

May 16th, 2007

This is another Borbely Erno design op-amp using all jfets.  This is the most complicated one with high gain and thus a servo is required to maintain the output at near 0Vdc.  I have tried this op-amp but I have not make it up to the best sound.  As of 2009, I think the all jfet buffer line amp sound the best.

All Jfet op-amp Circuit