Adcom GFA-5400

The GFA-5400 is a two-channel MOSFET power amplifier from Adcom. It came in two distinct versions, and that matters if you're servicing one. Version 1 shipped without Adcom's output protection circuit. Version 2 added it, built around a pair of NJM555 timer ICs (IC711/IC712) and a dedicated standby supply. That supply is a small transformer (TT002), rectifier D701, and a 7812-type regulator (IC701) on a board at the right front of the chassis. The power transformer, speaker binding posts, and input jacks are also unique to each version, though the amplifier circuit itself is nearly identical between them. Adcom published schematics for both in the service manual. The parts list covers version 2, with an addendum for version 1-only parts.

Rated output is 125 watts per channel into 8 ohms and 200 watts into 4 ohms. Both are continuous average power per channel from 20Hz to 20kHz with all channels driven and THD under 0.18%. The output stage uses IRFP240 and IRFP9240 MOSFETs, and the manual is emphatic about one thing: those devices must be replaced in matched sets of three per channel. Mixing unmatched parts causes current hogging and makes proper bias alignment impossible. Bias is set for 35mV across R675/R676, with DC offset trimmed to 0mV at the speaker terminals. The manual warns that correct bias is critical to output power, distortion, and heat dissipation, and to the amp's longevity. The main supply uses four 12,000µF/63V electrolytics, so there's real energy stored in there. The service literature is blunt about discharging before opening it up.

Adcom's own advertising placed the GFA-5400 in a family alongside the GFA-5200, GFA-5300, and GFA-5500. It described the line as the result of new technology built on the company's engineering foundation, and noted single-ended Class A front-end circuitry in the GFA-5500. Audio (October 1997, p.56) framed the power amplifier as the foundation of a high-performance system and the product category on which Adcom built its reputation for performance and value. Treat those as period marketing context, not measured review data. The surviving chunks are promotional copy, not bench tests.