Special requirement output buffer

[Svart]

Makes me wonder what is happening in my system where I run analog audio through "digital" cable.

digital
110ohms
.18uh/ft
12pf/ft conductor to conductor
26pf conductor to shield
76% velocity of prop
23.7 ohm/ft DCR

VS.

analog
60ohms
.18uh
26pf/ft cond to cond.
47.5 pf/ft cond to shield
66% vel of prop
25 ohms/ft

 

[JohnRoberts]

20+ ohms per foot ?

JR

 

[Guest]

Nichrome cables? :shock:

 

[Svart]

:shock:

sorry I was trying to type fast before lunch.

ohms/1000ft is better..


:green:

 

[mikep]

[quote author="JohnRoberts"][
Putting a class A buffer "after" the feedback loop is just adding a source of distortion. [/quote]

agreed. why whould you ever do this? IMO line output stages are already the weak link as far as linearity in alot of designs. This is where I personally want the most feedback! Im surprised to hear this is a trend.

I have one thing to add to what others have said. limit the bandwidth of the NFB loop around your output buffer to only what is needed for your application. this might help you achieve stability. another way to think of it is to use feedback to lower the output Z of the buffer, but only over the frequency range of interest. let it run open loop from say 35kHz on up.

then when you have your stable, ultra low-Z output stage, put in a control to add output resistance in steps, and lable it "cable focus". the audiophiles will love that.

if you are going to exhaustivly research this, maybe also look at receiver topologies that actively drive the cable shield.