virtual ground in multiple output amplifiers

[kokodin]

I post it here because i am just embaresed that i don't know it

situation is simple:
single power suply, ground at 0V and positive rail at 24V
i was thinking about making 4 tda2030 output stages for 4 separate 4 ohm speakers ~10W each

i made first power stage, tested it, and i begin to think. can i use virtual ground from this power stage to bias other 3 or should i make 4 separate resistor deviders to bias each tda chip biasing as its own thing?
not that i want save 2 resistors per chip, but i was wondering how much would single common virtual ground drift and distort inputs offor instance silent at this moment  channel?

[paulca]

Take a look at my recent thread "Circuit Review?" before committing to the resistor only based v-ground.  It has serious draw backs.  There is a link to literally dozens of designs for v-ground circuits for when you need much more power handling.

If the v-ground is stable enough I would expect it would be fine for biasing multiple outputs.  Mine does two.

A few tests I would do.... connect your circuit to a noisy power source, or power something noisy and digital from the same power source as your amp.  I found I got a lot of noise pick up with the resistor only divider.  The other is to measure the v-ground with a scope in AC coupling mode to see how much drift you have at high power outputs.   Personally I take that as a percentage of the full output and subjectively determine if it's good enough.  So 10mVpp drift outputing 10Vpp is 0.1% and fine.  100mV is 1% and probably too much.