SCREEN GRID MODIFICATION
 

My toroidal output transformers for valve amplifiers have a very wide –3dB frequency range (3 Hz to 200 kHz) and little linear and non linear distortions. The wide frequency range ensures excellent reproduction of the precious wide bandwidth audio signals from for instance the SACD. These features are recognized worldwide, as is proven by the many compliments which I receive.

But, this wide bandwidth can cause problems under certain circumstances. The amplifier can start to oscillate and becomes a radio transmitter. This never was my intention and therefore I have spent a lot of time to investigate this problem.

If you don’t have an oscilloscope (which clearly shows any kind of oscillation), then you can recognize the oscillation by humming from the amplifier. The high frequency oscillation (around 640 kHz, far above our hearing range) uses all the energy inside the amplifiers power supply and the result is a deep 50 or 60 Hz humming coming out of your loudspeakers. You might think that you made a mistake when building your amp. But that certainly is not the case. You built the amp in the right way. The humming is caused by oscillation, which is caused by a not damped coupling between the output transformer and the screen grids of the power vacuum tubes.

This oscillation sometimes occurs when the amplifier is in the Ultra Linear Mode. It does not occur in triode or pentode mode. In the Ultra Linear mode the screen grids, with series resistors of 150 of 180 Ohm, are connected to the UL-taps on the primary winding of the output transformer. The function of these screen grid series resistors is to prevent high currents through the screen grids and to decouple the output transformer from the screen grids at very high frequencies. Because my toroidal OPT’s have such a wide bandwidth, the standard decoupling of these screen grid resistors is too little, which can cause the oscillation.

What to do if your amp oscillates? Change the screen grid resistors into 1 kOhm (1 Watt). The larger value of these resistances is now enough to decouple the screen grids at very high frequencies, and the oscillation is suppressed and stops.


The listing below refers to my UL40-S2 tube amp kit and to the figures in my book “Modern High-End Valve Amplifiers …”. The resistors to be changed are mentioned.

UL40-S2 Construction manual R23, R24
Book fig.11.1/4/5/6/7 R14, R15
Book fig 14.2 R14, R15
Book fig.16.1 R13, R17
Book fig.18.1 R14, R17, R20, R23
Book fig.19.2 R14, R15

With many thanks to Ian L. Miller, Senior Electronic Engineer (Tenix LADS Corporation Ltd, Australia). He immediately recognized the cause of the oscillation and showed me where to look and how to solve the problem, which I checked and he was correct. Now you see the power of world wide cooperation!

The Netherlands
Zwolle: 29-12-2004
Ir. Menno van der Veen