It is not only that electron tubes gently and progressively clip the large parts of the signal while leaving the magic of the average part intact. Much of this can be predicted by examining the characteristic curves of triodes, load resistor lines as the signal peaks approach the anode supply rail etc.
Electron tubes also change their gain acting as compressors or limiters (automatic gain control).
For example the Neumann U47 VF14 circuit.
This was observed by chance around 2000.
By carefully unplugging the capsule assembly it was surprising that there was almost no hiss or noise coming from the VF14 prepreamp output. Low noise and of good quality, of not very high frequency content. This can possibly be explained by the reduction of secondary emission effects by the low anode voltage and also by similar and other effects because of the underheated VF14 cathode. It was also understood that a high grid resistor can contribute very low noise. Many of these principles influenced and inspired many of the Pleiades V series electron tube battery prepreamps.
Coming back to the large signals topic, once the capsule assembly was removed it was time for more experiments.
How does the VF14 prepreamp inside the U47 overload if a high enough sinewave voltage is applied at the capsule input terminals?
A top quality Wein bridge Hewlett Packard 205AG electron tube sinewave generator was used. It's output stage is a pair of 6L6 tubes! And after the output transformer is are 2 brilliant rotary switch constant impedance attenuators in steps of 10 or 1 dB.
Once the VF14 output was connected to a similar quality HP 120B electron tube oscilloscope and input levels adjusted it was time to start increasing the input level to see what will happen to the beautiful crystal clear sinewave at the screen.
As the input was increased to perhaps no matter how high levels the sinewave was there undistorted.
If suddenly the signal was attenuated by 10dB, the sinewave would reduce in amplitude but after a few seconds it would have grown again to the full size of the screen!
This is gain control and if memory is correct the X Y connection mode was also used on the oscilloscope. On the X axis the input signal and on the Y axis the output signal. As the signal was increased the brilliant straight line did I not curve but only its slope was reduced. The more the signal input (after a thresshold) the less the slope.
An explanation of the brilliant effect is the following.
As the signal is increased beyond a threshold voltage the grid starts to be driven positive and draws a tiny current. At the negative part of the input signal current is not drawn. This is a diode effect between cathode and grid.
The rectification of the large input signal (we may now call it the side chain signal) has the effect of charging the grid more negative. When electron tubes operate with more negative bias, anode current drops, transconductance drops and therefore voltage gain drops.
This lasts for a few seconds. If the input is reduced the grid is charged back again to its normal bias. Discharge takes place for those few seconds which is the release time. Thid may be governed by the bias resistor, the grid capacitance etc.
The grid resistor is many megohms.
Would circuits with smaller grid resistor still exhibit this effect but at a much quicker attack and release time? Is this another factor involved in explaining the electron tube transient overload characteristics?
How would a JFET behave if a diode is connected between source and gate to simulate this effect?
The rectification effect of electron tubes is brilliantly explained on how the envelope of the signal changes. In fact in radio engineering the signal is the RF signsl. And the slowly varying side chain signal is the audio signal itself derived from the modulated radio transmiter carrier.
Possibly the most important and clear paper ever written on how electron tubes operste is by the inventor of FM, Edwin Armstrong himself.
This brilliant paper played an important part in explaining how electron tubes are negatively biased internaly by nature. Even before any outside source is applied.
The Pleiades V series preamplifiers can operate at anode voltage of only 3 volts by leaving out any outside negative bias. And by supplying a positive bias high enough to counteract the by nature internal bias while still keeping the grid negative. Still negative but less so with respect to the cathode.
References:
Neumann U47 schematic
The use of multi grid tubes as electrometers - J.R. Prescott - The review of scientific instruments
Open-grid tubes in Low-Level Amplifiers - Robert J. Meyer - Electronics magazine oct. 1944 - Electronics for Engineers - Markus and Zeluff
Operating features of the Audion - Edwin Armstrong
On preserving the transconuctance of electron tubes at anode potential as low as 3 volts - euroelectron blog spot
Tubes vs Transistors, is there an audible difference? - Russel O. Hamm - JAES
No comments:
Post a Comment