So the whole thing weighs a meager 11 pounds. With the "toroidal power supply" (as indicated in their literature) that's not much of a transformer for the power rating. Even if it had a switching supply, that's pretty light weight. Anyone think that they might just be pushing the thing a little too hard? Smoke would be nearly inevitable.
I have a similar, low-grade power amplifier supposedly rated at "8Ω, 20-20,000 Hz. 2 channels, 250W/channel" and yet the output devices (2SA1941s) are rated for 100W maximum, and the manufacturer specifies them for use in a
70W audio amplifier. I'm afraid that all the misrepresentation, outright lies, and Madison Avenue drivel has permeated the audio market once again.
If the AudioSource sales literature states "50 watts RMS per channel @ 8 ohms" (as it does) for the AMP100, it's probably truly only capable of a continuous 20-25W without going full-incendiary. I'll guess that the actual transformer in this doorstop is probably only 0.200KvA, or maybe 250 on a cold day. You can perform the rest of the math, but if you won't, consider 80V P-P on the main rails, the RMS factor, and then Ohm's Law. That transformer needs to be at least 0.560 KvA not counting ANY overhead or reserve. A bare-minimum 560VA transformer alone (even a toroidal at 20-25% less than an E-core unit) would weigh over 10 pounds:
https://www.parts-express.com/avel-lindberg-y236803-500va-35v-35v-toroidal-transformer--122-665
Tell me again how this is a 50WPC (continuous) amplifier. A TRUE 50WPC stereo amplifier like an SA-8500 weighs almost 30 pounds, not eleven.
Failing with smoke? Yes, Mildred, that would be expected.