december 2005

by-passing the active output stage

unbalanced version

introduction

This modification is the easiest to perform on the DCX2496 for a beginning DYIer.

Let's first see the purpose of it, it's advantages and it's issues. Before touching anything in the Behringer you should first read what I read myself on a french forum - Audax Industries - in October 2005 :

"Hello! The experiments of JJP show well that the audio DAC are very sensitive to their output stage interface. One has the right to wonder whether the mistakes of output stages using tubes (expensive absurdity with ICs working on 5V) with the only purpose of a coloured sound ! There is a part of explanation in the structure of the output stage. The IC is subjected to square wave signal. It over-oscillates at each attack, the width of this over-oscillation depending on the height of the step (difference between two consecutive datas) and on the compensation of the opamp (Q in end of band). The integrator (low-pass filter) which follows will add to the signal the same integration of the over-oscillations if they are except band. It is a considerable interfering signal. It is composed of carrying in Ft the operational amplifier modulated by Fdac and in amplitude by the difference between two values. A pretty nonlinear signal that the integrator brings back in the audio spectrum (the secrecy of the quality of the DAC without filter of exit?) . One transports a part of dynamic qualities of the amplifier of exit in the form of distortions in the audio spectrum. It is important to have a stable operational amplifier (minimum of over-oscillation) with a time of minimal establishment duration of the over-oscillation). For this reason you will often find a capacitor at the output of a DAC, just to soften the clock front. It tries to limit the over-oscillation of the following amplifier, even if it's presence looks like a stupidity"

I immediately submitted this to my friend Jean Marc Plantefève who helped me so much to make my system evolve toward active multi-amplification.

He answered : "I had never thought of this point of view. This would explain why opamps with high slew-rate behave better than traditional operational amplifiers. The passive solution does not have any limitation of this kind indeed." However if you look at what's inside the DCX2496 you won't find any high slew-rate opamp : the thing is stuffed with those poor little plastic cans called JRC4580 - twenty four of them, not less. If you have been reading 'till there, congratulations, you are brave and patient. All of a sudden it was self-evident that some surgery in the DCX2496 would help it a lot to let the Music flow. Therefor the tweak I am proposing to you, requesting for "Level 1" skill.

Before you go further

if you don't feel like playing with your soldering iron or if you think you don't have the requested skill level you can buy a ready to fit input-ouput board from this company. It is a true plug&play kit, extremely well engineered, with audiophile grade components and it keeps all of the DCX original functionnalities.

 

next

 

See also the balanced version of this modification