Hello, I have little experience in the design of cfa amplifiers and I want to know if anyone could give me some advice about the following schematic with which I am working. The main problem I have is the high distortion in low power. Thanks in advance.
View attachment CFA.asc
View attachment CFA.asc
I'd rather not derail the thread right at the start but suffice to say that D. Self tells us that for precision and cost reasons, servos are increasingly used to set 0V with no requirement for adjustment. Yes, you can do that without a servo in standard VFA designs, if careful with the design and fine tuning, such that the offset remains stable over the full temperature range. However, that still requires a potentiometer and set-up operation for each amp.
I would read Doug's short chapter on DC servo design in his 6th ed. Audio Power Amplifier Design book. If you don't have access to a text, there are several online discussion forums that touch on the subject periodoically. Here's the first I saw that gives some reasons why: op amp - Integrator as DC servo - Electrical Engineering Stack Exchange
I would read Doug's short chapter on DC servo design in his 6th ed. Audio Power Amplifier Design book. If you don't have access to a text, there are several online discussion forums that touch on the subject periodoically. Here's the first I saw that gives some reasons why: op amp - Integrator as DC servo - Electrical Engineering Stack Exchange
How much bias flows in the output stage transistors (each of them, and all 4)?
I like servoes being attached further away from the input, eg. right where the voltage multiplier sits.
On a side note, I'd rather avoid that CFP output stage and make it a dual or triple EF, as per Leach amp. There is the chance of spurious oscillation when they switch on and off. At least make sure you check for correct operation with a fast enough oscilloscope as soon as you have built the circuit.
Otherwise, it looks fairly good (I have not checked operating points and such, however)
I like servoes being attached further away from the input, eg. right where the voltage multiplier sits.
On a side note, I'd rather avoid that CFP output stage and make it a dual or triple EF, as per Leach amp. There is the chance of spurious oscillation when they switch on and off. At least make sure you check for correct operation with a fast enough oscilloscope as soon as you have built the circuit.
Otherwise, it looks fairly good (I have not checked operating points and such, however)
Yes, but in parallel with a simpler resistor R6 ? is the point I am making, not denying the correct use of a DC servo.
the DC servo introduces a needed time constant, but is then in parallel with R6 that I am suggesting is undoing that needed delay.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.