|
|
|
|
|
|
Another mistake often made is to over-estimate the driving ability of a Cathode Follower. For example, don't expect a 12AX7 based Cathode Follower to drive a 600 ohm load to 10 volt peaks just because it has a 600 ohm output impedance. Let us assume that half of the 1 mA of idle current through this Cathode Follower can be delivered to the 600 ohm load. The resulting peak voltage is disappointing:
.0005 x 600 = .3 volts.
An idle current of 33 mA would be needed to make the 10 volts peak figure easily possible. The key point here is not to make the mistake of thinking solely in terms of voltage--current is equally important in a buffer's design
Sometimes too high a current draw leads to a compressed sound. Actually, this fault has nothing to do with the design or function of the Cathode Follower, but with sloppy power supply practice. For example, let us imagine a common tube lineup: a 12AX7 configured as a Grounded Cathode amplifier with a 150k plate resistor that then cascades into a 12AU7 based Cathode Follower with a 15k cathode resistor. Makes sense. Does it not? A high gain, low current amplifier at the front, followed by a high current, low output impedance buffer.
|
|
|
|
|
|
|
|
|
|
Everything looks good until we see that the both stages tie together at a power supply connection with relatively high series output impedance. So when the 12AX7 tries to pull its plate voltage down, the 12AU7's cathode will follow, but as the 15k resistor is ten times smaller in value than the 150k resistor, the change in current it produces in response to the input signal will tend to swamp out the change in current the 150k produces in response to the same signal.
The key point here is that while the Cathode Follower works in voltage phase with its input signal, it works in negative current phase to the Grounded Cathode amplifier. So while the 12AX7 tries to pull down its plate voltage by the increased current conduction, the 12AU7 lets go of ten times more current, which allows the power supply connection to drift upwards, which subtracts from the output.
Conversely, when the first stage tries to push up its plate voltage by the decreased current conduction, the Cathode Follower conducts ten time more current, which pulls the power supply connection downwards. Of course, if the power supply were perfect, i.e. zero impedance, this cancellation effect could take place.
The obvious strategy is to use a regulated power supply. A more subtle move is to match the Cathode Follower cathode resistor to the value of the plate resistor of the first stage so as to cancel any net variation in current presented to the power supply.
The final setback to the Cathode Follower is that because it is intrinsically a single ended circuit, it can only aggressively drive the output in one direction: up. If presented with a difficult load, i.e. low impedance, the Cathode Follower can be driven into twice or three times its idle current, but it can only decrease its idle current to zero conduction. This makes for an asymmetrical output drive potential. A popular myth is that this problem can be overcome by using a negative power supply. But a negative power supply only allows for a larger valued
|
|
|
|
|
