|
|
|
|
|
designers. If nothing else, having a bipolar power supply almost instantly lowers the PSRR of a circuit, whether it is transistor or tube based, as the power supply noise on a split rail power supply tends to be out of phase with itself and thus cancels at the output.)
A tube based Op-Amp can become very complex, usually requiring at least 3 triodes and a negative power supply, something which is not always available. Complexity does not necessarily equal bad sound, just as simple does not always equal good sound. Still, fixing bad sounding simple circuits is usually much easier than fixing the complex. Furthermore, few are willing to build a circuit that eludes understanding. Consequently, although a tube based OP-Amp can be built, an example is shown below, a simpler approach would probably prove more useful.
The tube based Op-Amp shown above is simple enough. A Differential amplifier receives the input signal and its output is cascaded into the output stage. The output stage has a gain of a little less than unity. Consequently, all the gain derives from the first stage and feedback would only subtract from the total gain. Still, this circuit would need much feedback, as it is very linear. Whereas transistor based Op-Amp require all the feedback that can be safely applied to them to bring down the distortion.
|
|
|
|
|
|
|
|
|
|
is 0.00001. Here the tube's low gain is definitely a disadvantage, but not an insurmountable one for our use. The tube Op-Amp would certainly require some tweaking of resistor values and the end result would disappoint the oscilloscope, but please the ears. Nonetheless, a simpler way that uses less circuitry is possible.
Half an Op-Amp
Actually, all we really need to do is match the IC's functionality, not its topology. It turns out that just 2 triodes are all that is needed in this situation. Basically, we have taken only the push-pull output stage of a tube Op-Amp. (In this issue we have an article o push-pull amplifiers.)
At first glance, the output stage of the tube Op-Amp seems to consist of a Cathode Follower on top of a Grounded Cathode amplifier. But if we look closer, we see that yes the top triode defines a Cathode Follower but the bottom triode defines a circuit something like a Plate Follower.
Any externally applied voltage to the output will be reflected back to the grid of the bottom tube through the capacitor C1 and the plate resistor. But as the signal presented to the grid will be inverted at its plate, the plate voltage's movement will be in opposition to the externally applied voltage. Thus C1 and the plate resistor define a feedback loop from plate to grid. Feedback both reduces the gain of the bottom triode and lowers the output impedance at its plate. In other words, both the top and bottom tube work as followers that buffer the output without providing any gain.
Note that the input signal must be balanced, plus and minus. Since the top half of our output stage is not phase inverting, it should receive the positive input signal. The bottom tube is phase inverting, thus it should receive the negative input signal.
|
|
|
|
|
|
|
|
|
|
If this tube Op-Amp is so good, why no use it in place of a IC Op-Amp? The Op-Amp provides more functionality than just low distortion gain; its super high gain allows the ignoring of the gain in the use of the Op-Amp. In the previous example of how differential signals are amplified and common mode signals are ignored, the actual results are off by basically the inverse of the open loop gain of the Op-Amplifier, which could easily 100,000, the inverse of which
|
|
|
|
|
|
|
