[balou]

Yeah, the thread title got a bit long

Wouldn't it be possible to bias the opamps in your source (if the output is unbuffered) to class A by adding just a CCS between the input jacket of your headphone amp and the opamp? If there are any output caps, they might be in the way, but you could just remove them from your source (ok, that would add a bit of modifying). of course you would then put suitable dc blocking caps in your headamp.

Have I just missed a good reason why this wouldn't work and why nobody thought of it before? Or could this be a viable option if the opamps in your source aren't already into class a (I'm thinking of portable devices)?

 

[AudioCats]

to do such bias you need to have access to the opamp's power supply rail, + or - depends on how you want to bias it, usually the - rail. The bias is to use a CCS to cause constand DC drain from opamp output to one of its supply rail. I don't see how this can be done without modifying the source.

 

[balou]

I somehow had in mind that you have to tie the CCS to ground... what result would that give if you tied it to gnd instead of to V-?

 

[AudioCats]

check your PPA schematic, bias draines to the - (which will load the NPN).

 

[Bjornboy81]

Quote:

Originally Posted by balou /img/forum/go_quote.gif I somehow had in mind that you have to tie the CCS to ground... what result would that give if you tied it to gnd instead of to V-?

If you went to ground instead of -V you'd just be loading it, not biasing it.

 

[vixr]

would this mod have an effect on battery life?

 

[amb]

Many sources have output coupling caps, so putting a CCS after it doesn't do anything because there won't be a DC path from the source's internal opamp output to it.

Also, the CCS needs to go to the opamp's supply rail (V- if you use a current sink, V+ if you use a current source, mostly commonly it's done with V- because this causes the opamp's internal NPN output transistor to be biased on all the time, and disables the PNP complement). If the CCS is tied to ground, it will not disable the PNP transistor and won't accomplish the task of making the opamp go into class A.

 

[balou]

yeah ok... so this is a non-idea

For a second I thought I had a great idea