There is a particular kind of audio purchase that appeals to the tinkerer in me: the clone. The FM Acoustics FM300A is a legend from the eighties, an amplifier that in original form costs more than most cars, and the Chinese replica market has decided that the rest of us should get a taste. Depending on the seller and the level of cosmetic excess you choose, these land somewhere between $450 and $900, with most of the decent standard versions sitting around the $500 mark. That is a remarkable price for a 25 lb chassis full of real transformer, real heatsink, and real discrete circuitry.
The topology, briefly
The FM300A layout is a fully complementary, fully discrete bipolar design — no op-amps anywhere in the signal path. The input is a differential pair feeding a voltage amplification stage, which in turn drives a complementary emitter-follower output section with paralleled bipolar devices per rail. Bias is set by a conventional Vbe multiplier sitting on the output heatsink, so it tracks temperature, and there are trimmers for both idle current and DC offset on each channel. Global feedback is kept modest, which is a large part of why the original has the reputation it does — and also why this design is unusually sensitive to how it is set up. Get the bias wrong and you are not listening to an FM300A anymore, you are listening to a crossover-distortion generator with a nice faceplate.
Specs worth knowing
80W per channel into 8 ohms, 150W per channel into 4 ohms
Frequency response roughly 10 Hz to 30 kHz
THD under 0.005% at 1 kHz
S/N in the 115 dB region
RCA inputs only, standard 430mm chassis, about 11 kg
Numbers are not the story here, and I am not going to pretend otherwise. What matters is what happened when I plugged it in.
The first listen, and the screwdriver
It sounded sharp. Not bright in a pleasant, airy way — sharp, glassy, fatiguing within a track or two. So before writing a word about it I did what anyone should do with a clone: pulled the lid and got the multimeter out.
DC offset was sitting around 7 mV on one channel and 10 mV on the other. Not dangerous, but not right either, so I trimmed both down to essentially zero.
Bias was completely off — about 1 mV. That is effectively nothing, which put the amplifier into pure class AB operation across the entire range, which is not what this circuit is designed to do. Target is 15–30 mV per channel, which gives you roughly the first watt in class A and AB from there up.
Setting bias properly smoothed the sound considerably — immediately and obviously. Then it took another 100 hours or so of playing time for the amp to fully settle, shed the remaining edge, and open up from a transparency standpoint. Patience is part of the deal.
How it actually sounds
Even sorted and run in, I still hear a slightly metallic character that I cannot quite talk myself out of. The midrange, though, is genuinely good — well defined, smooth, and free of the plasticky quality that plagues cheap solid state. Bass is impressive for a box this size, tuneful and controlled, though it will not go toe to toe with a properly big amplifier when things get demanding. It also runs remarkably cool, about 32°C (90°F) on the sinks, which my office appreciates during a Florida summer.
My first workaround for the sharpness was to have my PC convert everything to PDM (DSD), which pushed things in a more analog direction. Then I inserted my DIY 12AU7 tube buffer, running an Amperex for a little added sweetness — and the whole picture changed. The soundstage expanded, the remaining hardness simply disappeared, and the PDM conversion became unnecessary. With the buffer in front of it, this is a very good sounding amplifier.
Final verdict
The FM300A clone is a lot of amplifier for the money, but it is not a plug-and-play proposition. Mine arrived mis-biased badly enough to misrepresent the entire design, and I suspect that explains a good number of the lukewarm impressions floating around. Set the offset, set the bias, give it 100 hours, and ideally feed it from something with a bit of warmth — a tube buffer or tube preamp is the perfect partner. Do that, and you have a smooth, well-organized amplifier that punches far above its price. Skip it, and you will wonder what the fuss was about. Recommended — with a screwdriver and a multimeter in hand.
There is a particular kind of audio purchase that appeals to the tinkerer in me: the clone. The FM Acoustics FM300A is a legend from the eighties, an amplifier that in original form costs more than most cars, and the Chinese replica market has decided that the rest of us should get a taste. Depending on the seller and the level of cosmetic excess you choose, these land somewhere between $450 and $900, with most of the decent standard versions sitting around the $500 mark. That is a remarkable price for a 25 lb chassis full of real transformer, real heatsink, and real discrete circuitry.
The topology, briefly
The FM300A layout is a fully complementary, fully discrete bipolar design — no op-amps anywhere in the signal path. The input is a differential pair feeding a voltage amplification stage, which in turn drives a complementary emitter-follower output section with paralleled bipolar devices per rail. Bias is set by a conventional Vbe multiplier sitting on the output heatsink, so it tracks temperature, and there are trimmers for both idle current and DC offset on each channel. Global feedback is kept modest, which is a large part of why the original has the reputation it does — and also why this design is unusually sensitive to how it is set up. Get the bias wrong and you are not listening to an FM300A anymore, you are listening to a crossover-distortion generator with a nice faceplate.
Specs worth knowing
Numbers are not the story here, and I am not going to pretend otherwise. What matters is what happened when I plugged it in.
The first listen, and the screwdriver
It sounded sharp. Not bright in a pleasant, airy way — sharp, glassy, fatiguing within a track or two. So before writing a word about it I did what anyone should do with a clone: pulled the lid and got the multimeter out.
Setting bias properly smoothed the sound considerably — immediately and obviously. Then it took another 100 hours or so of playing time for the amp to fully settle, shed the remaining edge, and open up from a transparency standpoint. Patience is part of the deal.
How it actually sounds
Even sorted and run in, I still hear a slightly metallic character that I cannot quite talk myself out of. The midrange, though, is genuinely good — well defined, smooth, and free of the plasticky quality that plagues cheap solid state. Bass is impressive for a box this size, tuneful and controlled, though it will not go toe to toe with a properly big amplifier when things get demanding. It also runs remarkably cool, about 32°C (90°F) on the sinks, which my office appreciates during a Florida summer.
My first workaround for the sharpness was to have my PC convert everything to PDM (DSD), which pushed things in a more analog direction. Then I inserted my DIY 12AU7 tube buffer, running an Amperex for a little added sweetness — and the whole picture changed. The soundstage expanded, the remaining hardness simply disappeared, and the PDM conversion became unnecessary. With the buffer in front of it, this is a very good sounding amplifier.
Final verdict
The FM300A clone is a lot of amplifier for the money, but it is not a plug-and-play proposition. Mine arrived mis-biased badly enough to misrepresent the entire design, and I suspect that explains a good number of the lukewarm impressions floating around. Set the offset, set the bias, give it 100 hours, and ideally feed it from something with a bit of warmth — a tube buffer or tube preamp is the perfect partner. Do that, and you have a smooth, well-organized amplifier that punches far above its price. Skip it, and you will wonder what the fuss was about. Recommended — with a screwdriver and a multimeter in hand.
How to Bias and Set DC Offset on an FM300A Clone Power Amplifier: https://tubemaze.biz/fm300a_clone_settings/