Naim "One Vbe" Constant Current Sink Modifications

[nobeone]

While discussing NAXO ALWSR wiring with @Andrew L Weekes he pointed me at an article from Walt Jung on "One Vbe" constant current source circuit modifications. He suggested I look to implement this in the NAXO to improve the underlying Naim constant current sink part of each filter circuit. Having hunted about on pfm I believe @martin clark has provided a bit more detail on how to do this in this thread on CD3 modifications.

If I understand correctly and relate it to the NAXO schematics Neil McBride provides, say the SBL low pass for example, then for each of the 47k resistors feeding a pair of ZTX384 from the +24V rail, I split the 47K resistor into two 22k resistors and stick a 10uF capacitor from the 0v rail to the middle of the two 22k resistors.

Have I got this correct?

Suggestions for cap? Obvious things I have hanging around include 22uF 50V Panasonic NGH aluminium electrolytic from ALWSR build. That do?

 

[PigletsDad]

That would be fine. The cap effectively sees two 22Ks in //, roughly like 11K, so 22uF would give a time constant of 240ms, or a corner frequency well below 1Hz. This won't give too long a startup, but will attenuate anything coming in from the rail - even at 20Hz you will 30dB of rolloff, and at the main hum frequencies (100Hz and multiples) it will be more than 40dB.

 

[nobeone]

Thanks PD, good to go then!

 

[nobeone]

Looking at that Walt Jung article I linked to above, in figure 1 he has two 49k9 with 10uF in his constant current source, so much the same time constant as two 22k with 22uF. If it is good enough for Walt, I think it is safe to say it will do me. I appreciate these things are not critical, just didn't want to be an order of magnitude away from sensible.