"Bridging" a LM4881 amplifier

[poiureza]

Hello,
I have a small PCB with a LM4881 amplifier chip installed and prewired exactly as in the TI datasheet here on page 2 :
https://www.ti.com/lit/ds/symlink/lm4881.pdf?ts=1691316121503

Now I'll be using only one channel for both in and out, thus mono operation.
Would there be any advantage of using both amplifiers still ?  I hear that available current is usually the limiting factor for this kind of amplification.
In case I suppose it's not as easy as simply shorting the L&R inputs and the L&R outputs ? Or is it ?
From the sheet I see that both amplifiers are inverting FWIW

Another question I have is about volume.  I suppose I should put the potentiometer at the input, not at the output.
If I'm not mistaken this chip always runs full tilt without any internal negative feedback ?

[PRR]

I have a small PCB with a LM4881 amplifier chip installed and prewired exactly as in the TI datasheet
If I'm not mistaken this chip always runs full tilt without any internal negative feedback ?

The datasheet clearly shows the external negative feedback; the chip won't run right without it.


Paralleling transistor amp outputs is a Bad Idea. They fight.

It is a 5.5V chip. It has been scaled-down to suit a commercial niche: headphone jacks in laptop PCs. If you want POWER there are better choices.

[Rob Strand]

You haven't said what the load is.

For a headphone amp:  Each channel of the chip will go down to 8 ohm load.  Generally the chip will have an easier time if you just parallel the inputs then use a separate output to each earphone.  Volume control at the input.

If you are building some sort of overdrive pedal then use only the first channel.  For the second channel wire in the Rf resistor as shown and leave off Ci, Ri, Co.   There's no guidance on nullifying a channel but what I've suggested should work.   You can put the volume on the input or output, provided the load impedance isn't too low, say less than 220 ohm or so.

What you have suggested isn't bridging.   That's a whole different thing - both outputs are used and are floating without a ground.   No guidance is given in the docs but it could work.    Min load impedance would be 16 ohm.

[poiureza]

You haven't said what the load is.

For a headphone amp:  Each channel of the chip will go down to 8 ohm load.  Generally the chip will have an easier time if you just parallel the inputs then use a separate output to each earphone.  Volume control at the input.

What you have suggested isn't bridging.   That's a whole different thing - both outputs are used and are floating without a ground.   No guidance is given in the docs but it could work.    Min load impedance would be 16 ohm.

Yes, a mono headphone amp it is.
I like the idea of using both outputs.  At least I get max power in each ear.  SO I just short L&R inputs then, and put volume control before that.

How about sending the channel#1 out towards the second channel in ? In series ? That would probably work out no ?

[poiureza]

I have a small PCB with a LM4881 amplifier chip installed and prewired exactly as in the TI datasheet
If I'm not mistaken this chip always runs full tilt without any internal negative feedback ?

The datasheet clearly shows the external negative feedback; the chip won't run right without it.


Paralleling transistor amp outputs is a Bad Idea. They fight.

It is a 5.5V chip. It has been scaled-down to suit a commercial niche: headphone jacks in laptop PCs. If you want POWER there are better choices.

Yes I somehow completely skipped the external wiring 
Typical power from a laptop headphone jack would be enough for my needs.

I'll give it a try

[Rob Strand]

Yes, a mono headphone amp it is.
I like the idea of using both outputs.  At least I get max power in each ear.  SO I just short L&R inputs then, and put volume control before that.

Separate amplifier must feed separate headphone channels.   Shorting (ie. paralleling) outputs is bad, as PRR already mentioned.

How about sending the channel#1 out towards the second channel in ? In series ? That would probably work out no ?

There's nothing to be gained with that configuration when you can parallel the inputs.   In fact one of the outputs will be the opposite phase of the other - you might get away with it with headphones but it will cause hard to explain phase cancelling behaviour if someone plugs the output into something else.

[poiureza]

OK thanks