Audio amp as power supply?

[brainwash]

I haven't seen any design that discusses this, not that it would be extremely useful. Very few audio amp datasheets provide DC characteristics, a few of them supply a bandwidth. I thought in an amp design gain decreases (almost linearly) with bandwidth but at least for TDA2030 this seems not true. Is it oscillation at low frequency a reason for this?
http://www.st.com/web/catalog/sense_power/FM125/CL1503/SC979/PF65115

I also have this one in the drawer: http://offer-shem.narod.ru/data/tda/tda1011a.pdf and it does provide DC data and internal schematics but incomplete gain graphs.

Obviously, I have a few of these laying around and was thinking of converting them to controlled PSUs for stress testing. They have really poor precision (0.1%-10%) especially when driven hard, but they provide shortcircuit and thermal protection.

[Paul Price]

Audio amps are not quite suitable for this purpose because most become unstable and oscillate with high capacitive loads, such as a filter capacitor on the "OUTPUT" of your audio power supply-amplifiers. This idea might work with small loads if you decouple the output with a 4 to  8-ohm resistor in series with the output, but the regulation might be rather poor except at low current loads.

[brainwash]

Good explanation, I didn't realize that they were designed for inductive loads.

[hlavac]

Also, most audio amps have protection against DC current on output bacause that burns (literally) speaker coils.