Fully differential 100W amplifier
This circuit was intended not primarily as a hifi amp but as a supply for driving a transformer as a load device. The most important requirement was that there is no offset voltage.
But I wonder if this architecture represents another possibility for hifi. It is a fully directly coupled, differential input-differential output circuit, and basically eliminates offset voltages through balanced circuitry which ordinary circuits are unable to match.
The circuit is essentially a "double standard" as it has a differential input stage, Darlington VAS as suggested by Self, only two with independent current sources, and twin output stages operating in balanced push-pull. The differential loading transistors of the input stage to VAS bases are balanced through a common mode bias circuit, after Bob Cordell and others, and the output is maintained at zero voltage common mode by a second differential bias circuit feeding the current sources to the VAS stage.
Parallel pairs of 2N3055/MJ2955 are used in each output side as the whole amp needs only +/- 25 to 30V rails, and these can dump several amps into an 8 ohm load.
I offer this not as a fully proven circuit but one which has good performance as simulated, and by way of example gives .02 % at 100W and 20 kHz, and .004% at 2 kHz.
Will let you know how the real circuit behaves when I have built it!
Cheers
John
This circuit was intended not primarily as a hifi amp but as a supply for driving a transformer as a load device. The most important requirement was that there is no offset voltage.
But I wonder if this architecture represents another possibility for hifi. It is a fully directly coupled, differential input-differential output circuit, and basically eliminates offset voltages through balanced circuitry which ordinary circuits are unable to match.
The circuit is essentially a "double standard" as it has a differential input stage, Darlington VAS as suggested by Self, only two with independent current sources, and twin output stages operating in balanced push-pull. The differential loading transistors of the input stage to VAS bases are balanced through a common mode bias circuit, after Bob Cordell and others, and the output is maintained at zero voltage common mode by a second differential bias circuit feeding the current sources to the VAS stage.
Parallel pairs of 2N3055/MJ2955 are used in each output side as the whole amp needs only +/- 25 to 30V rails, and these can dump several amps into an 8 ohm load.
I offer this not as a fully proven circuit but one which has good performance as simulated, and by way of example gives .02 % at 100W and 20 kHz, and .004% at 2 kHz.
Will let you know how the real circuit behaves when I have built it!
Cheers
John
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.