nos-dac to gainclone coupling capacitor question

[sharpi31]

I have a nos-dac (with passive I/V conversion) and LM3875 gainclone (20K pot as passive pre). As I have built both I have only one coupling capacitor in the chain, currently between the pot and LM3875.

This means that DC from the TDA1543 can flow through the 20K pot to ground, and no DC can get from the pot to the LM3875.

Is this stupid? Would I be better to put the coupling cap between the TDA1543 and pot preventing all DC from the dac? Had I thought of this before I'd probably have gone down this route.

Any experience would be much appreciated. The voltage at the output of the TDA1543 measures 4.5V, and this would lead to 0.225mA flowing through the 20K pot. It's not much, but is it worth avoiding?

Thanks for all the help so far.

[sharpi31]

Well, I changed configuration last night and put the input capacitor before the pot, and this was a definite improvement.

Two possible reasons I reckon: (1) No DC route to ground from TDA1543 outputs, so no current drain from DAC (2) voltage accross the input cap is always at full line level, rather than post pot attenuated level.

This also removes the tendancy for the speakers drivers to move in and out as you turn the volume up or down.

Conclusion: (in my system to my ears) the input coupling capacitor is better placed before the potentiometer than after. Gains include greater resolution, dynamics and smoothness. Also, the battery seems to last longer on my dac due to lack of DC on outputs.

I guess my current system is equivalent to a having no input coupling cap on the amp, and only output coupling caps on the TDA1543 dac. I'd have located the coupling caps in the dac if they weren't so big! (4uF russian paper in oil from Ebay)