Current feedback amplifiers are less often used in audio applications. The main advantage of CFAs is extremely high speed, the main disadvantage is lower open loop gain and poor power supply rejection ratio. I have seen many designers build current feedback amplifiers that perform very well and like to try the CFA topology, too. In this article I present my CFA front end and explain its features in detail.
- Good power supply rejection
- Power supply invariant operation with wide operating range
- Input over-voltage protection
- Natural anti-saturation features
- Negative feedback loop stays intact during heavy clipping
- Adaptive clipping by sensing the output stage power supply
The power supply features are the same like for the Cordell inspired amplifier front end that I developed earlier. The diamond buffer at the input of the CFA front end allows to run from a lower power supply voltage and this is why I added two very simple series regulators. The supply voltages generated by the series regulators may be used by other circuitry like a DC servo.
The CCS design is based on my previous investigation of constant current sources for audio applications.
The input filter offers many options: C3 is a huge film capacitor, which is my preference. As an alternative, a bipolar electrolytic capacitor (C4) can be installed with a small film capacitor (C5) in parallel optionally. Resistor R3 ensures that no DC builds up. R2 and C2 may be used to properly terminate the audio input cable. They should only be installed in case a long cable connects to the amplifier input directly and be omitted inc case any pre-amp or audio input transformer is used. R4, R5 and C6, C7 form a second order low pass filter with a corner frequency of 600kHz. Most amplifiers use a first order filter here, but I find that a second order filter has better performance. Resistors R4+R5+R6 sets the input impedance. I chose 30kΩ as a middle course between 10kΩ often used for professional audio interfaces and 50kΩ for consumer audio electronics. R1 and C1 are the usual measures to break ground loops and may be replaced by a direct connection instead.
The amplifier input protection I developed earlier, is implemented here. This is supposed to be a robust amplifier building block. In case no protection is required or desired, the components can be left away easily.
The design does not depend on special components and all transistors may be substituted for ones with similar performance.
Shunt compensation is the most simple way to compensate a CFA. Only a R-C shunt network is required at the output.
Alexander compensation is an advanced compensation technique, where the R-C shunt is not grounded, but returned to the inverting input instead. I consider this somewhat experimental.
I haven't designed a PCB for this design yet. The form factor and layout of some shared schematic blocks will be the same like the Cordell inspired VFA front end.
Simulation is a convenient way to set up a circuit and toy around with it. With all the limitations, simplifications and unrealistic assumptions of circuit simulation in mind, this can be very useful.
CFAs usually don't have high PSRR and this is why I use active filters to improve PSRR.
There are three different modes of clipping. I kept the feedback Baker clamps from my CFA front end. The CFA front end has some anti-saturation diodes added and I also added resistor R50 as an experimental special effect. Here is how the clipped signal looks like with each of the clipping options:
Clipping the excess signal by feedback Baker clamps results in a very clean clipping of the waveform. The clipped part of the signal is returned to the inverting input of the amplifier.
In case a constant relation between front end and output stage power supply can be maintained, the anti-saturation diodes in the gain stage may be helpful. This results in a somewhat sharp onset of clipping and more gradual clipping beyond this point.
This is how the anti-saturation circuitry works: During normal operation, diodes D7 and D8 do not conduct because the cathode potential is held higher than the anode. Once the voltage at the collector of Q23 or Q24 approach the supply rail, diodes D7 / D8 turn on and this collapses the gain of this stage. Diodes D1 and D2 set the trip point of this clamp.
The addition of resistor R50 is an experimental fun feature. This resistor starves the current through the amplification stage made up of Q13 and Q14 with higher signal amplitude. The higher the signal amplitude, the lower the voltage across resistors R24 and R25 and the lower the current trough Q13 and Q14. This rises distortion even at low signal levels, but the effect gets very dramatic at high signal voltage. Basically this is a form of dynamic compression, but implemented entirely differently than a normal compressor or limiter. Modern music is compressed in the studio until instruments cannot be distinguished from each other anymore and adding more compression is certainly not a good idea at all. I guess this feature may instead be interesting for musicians seeking distortion to shape the sound of their instrument.
This modules has not been built yet.
The CFA module development is in a very early stage. Currently I experiment with the circuit in LTSpice.
There is a very short, but helpful article on Wikipedia about current feedback operational amplifiers.
Chris Russel from Hifisonix writes about VFA versus CFA designs.
Should this not be clear enough, there is another article from Chris Russel about VFA versus CFA designs.
And Chris Russel also wrote about the CFA input diamond buffer in detail.
The Hifisonix KX2 amplifier is a good example for a CFA design.