Whilst I'm aware that this is not its intended purpose I'm using the
TPS65316 as a bipolar supply for a low-power audio stage in
high-fidelity portable application where noise performance is very
important.
The datasheet suggests that if the current draw on either of the
outputs differs by greater than 30%, the device will go out of
regulation.
Under high output levels, the audio drive current
will be the dominant draw on the device and I believe this will indeed
lead to a greater than 30% difference between the positive and negative
rails. This mismatch will be following the audio signal (e.g. reverse at
a frequency of no less than 20Hz) and the overall current draw will be
considerably less than the total ability of the device.
Does the AC nature of the current draw mitigate the current mismatch
between the rails?
Is 30% still a relevant figure for the device
going out of regulation if the total current draw is less than, say,
100mA.
Do you have any suggestions as to how prevent the device going
out of regulation in this application, or do you feel the part is
entirely unsuitable - if so, can you recommend another solution with
similar output noise characteristics and minimal external component
count?
Regards,
Sam.