|
Ultra high fidelity, and can run a 2ohm speaker at 46.3V. This has a frequency response of <1Hz to 200kHz at max power. It also has a THD of 40mV at max power giving a THD of 0.0864%. It also has a gain of -233mdB which increases to almost exactly 0dB as the output load resistance increases to infinity. It also has a power consumption of 818W at full power for a dissipation of 282W at full power and it has an idle power consumption (and thus dissipation) of 32W. The maximum input voltage without output clipping is 47V.
Edit: I added a discrete OP-amp to improve the THD to practically nothing, and as a result under all conditions except overload it has unity gain and its upper frequency limit has been lowered to 50kHz at full power because of that 100pF upper frequency limit limiting capacitor which is to remove self-oscillation. Even without it, it oscillates at 200MHz, so it would be inaudible but wires within the amp would act like miniature antennas and it could jam or deteriorate the signal quality of radios operating in that frequency range. Also, I haven’t tested it, but the self-oscillation could affect the behavior of the amp even in the audio frequency range. That’s why I put in that 100pF capacitor. Also, the relays are to remove startup issues involving the “Cannot find solution” message. In real life, just connect positive supply directly to the positive rail. Also, that current source at top that has two outputs isn’t actually because they need to be the same current, but because it removes extra components and keeps things less cluttered.
Edit2: I measured the efficiency to be 65.5%.
Edit3: Probably should have made some measurements before getting out the calculator. Power consumption is 742W (and my newly predicted consumption was 736W, and the difference should be able to be explained by the idle current). That makes for an efficiency of 72.2%, very close to the theoretical maximum of 78.5% (pi/4).
Edit4: I’ve added some current protection which reduces the maximum output voltage to 45V (from 46.3V) leading to a reduced efficiency of 70%. For best protection, a fuse should be added to each of the power supplies, but you have to take into consideration it’s resistance which will limit the maximum output voltage further reducing efficiency.
Edit5: I’ve removed the diodes that biased the output stage and replaced them with a 2 resistor 1 transistor biasing setup. The transistor in that setup should be in thermal contact with the main heatsink so that it heats up in conjunction with the main power transistors and it’s gain will therefore always match them for a constant bias that remains virtually unchanged with temperature. Also, I’ve made the contact resistance of the relays that help with startup 50mOhms to simulate the resistance of an electronic breaker on both power supplies to prevent failure of the main power transistors.
|
