Bridge driver

[digi01]

I have a application layer design at year 2004 LINK talk about unbalance to balance convert.

here is a new project talk about ture bridge driver unit.

the attached is a simple hint.
Zang

[Fenris]

This circuit won't work. R1 is not needed. OP 2 is set up as an inverting amp, but its gain is 0. Put R1 in the feedback loop of OP 2 and make R1=R2. This will give you a basic unity gain differential driver. Use a dual op amp like an OPA2132 and it shouldn't be much different from a DRV134. You can also change the values of the feedback loops to give you additional gain.

[digi01]

so like this?

[Fenris]

Yes. Now you have an inverting amplifier and a voltage follower (non-inverting amp with gain=1).

http://en.wikipedia.org/wiki/Operati...r_applications

[Fenris]

Here's how you could lay it out on a small stand-alone board with a simple dual op amp. The red things are resistors, the black lines are traces, the big round things are capacitors and the small round things are connection points. This is of course a very basic design. There's alot more you could do with crossfeed, increased gain, bandwidth limiting, filtering, etc.

[theChris]

actually, the originally posted circuit is useless.

the way to look at opamps, at least in a linear context like amplifiers, is to look at two issues:

1.) the opamp will force the +in and -in terminals to the same voltage. the output will be whatever is required to make it so. (ideal negative feedback)
2.) for rule #1 to apply, if a small voltage difference were to occur between +in and -in, the output would take the same polarity as this difference. this needs to force the voltage difference to 0.

in the original circuit, the buffer has a gain of 1. (the output is equal to the -in, and if the input raises, the opamp will raise its outputs, which will move in- toward in+).

the incorrect "inverting" buffer has issues. Again, it would seem that the output is equal to the input (making it a noninverting buffer), BUT a small change in input voltage will cause the output to drive +in away from -in, forming a latch.

hopes this helps explain things.

see also sound.westhost.com for more info on balenced transmission projects.