EEVblog Electronics Community Forum

Electronics => Projects, Designs, and Technical Stuff => Topic started by: German_EE on July 12, 2015, 11:47:29 am

Title: Audio Distribution Amplifier
Post by: German_EE on July 12, 2015, 11:47:29 am
I want to build this circuit but with two inputs rather than one:

http://www.seekic.com/circuit_diagram/Amplifier_Circuit/AUDIO_DISTRIBUTION_AMPLIFIER.html (http://www.seekic.com/circuit_diagram/Amplifier_Circuit/AUDIO_DISTRIBUTION_AMPLIFIER.html)

What bothers me about this design is the direct connection between an opamp output and the three non-inverting inputs. What will happen when I connect two opamp outputs and three inputs together?
Title: Re: Audio Distribution Amplifier
Post by: SeanB on July 12, 2015, 12:06:33 pm
Put a 10k resistor after the feedback point on the first opamp (R2), and simply duplicate the input again with the added resistor, and connect the resistors to the 3 opamp inputs you are using. You can also share the R4, R5 bias network but reduce the resistors to 10k, and leave C2 at 100uF.

Otherwise, just common the inputs with 10k series resistors and you will be fine.
Title: Re: Audio Distribution Amplifier
Post by: German_EE on July 12, 2015, 02:34:30 pm
OK, thanks, in my head I tried to image one opamp at +9v output and the second at -9v, the results were not pretty and those series resistors fix the problem.
Title: Re: Audio Distribution Amplifier
Post by: Mark Hennessy on July 12, 2015, 07:10:19 pm
It would also be wise to add series resistors for each output. Something like 47 or 51 ohms would be enough. Otherwise, the op-amp might be unhappy at having to directly drive the capacitive load presented by the cables.
Title: Re: Audio Distribution Amplifier
Post by: T3sl4co1l on July 12, 2015, 08:06:29 pm
The differential input range is fine, even if the output is shorted and the first amp is driving full scale.

TL082 isn't much of a driver, you may want something with higher output current.  And then, add a series output resistor (maybe 100 ohms, so the output doesn't directly drive line impedance, which might be capacitive or worse), and for perfect LF/DC fidelity, you can also add a feedback loop: cap from OUT to -IN (maybe 100pF), resistor from -IN to line connector (maybe 1-10k).

A common sight for DC accuracy is using equal series resistance (or Thevenin equivalent) to the input terminals, but since TL082 is a JFET type, its input bias current is low enough to ignore.  This is a useful tip for best DC offset on bipolar types (e.g. LM833?).

Tim