It is well known that 23cm PAs using Russian GI7b / GI7bt tubes suffer from thermal drift and can oscillate under certain conditions. The majority of amateur radio designs using these tubes are based on 3/4 lambda cavities, the famous CT1DMK design being the prime example. These amplifiers commonly exhibit thermal instability and low efficiency. Instability may also cause the PA to oscillate or, at minimum, create bad intermodulation distortion on the output. This creates splatter which is especially problematic during contests, particularly if many stations are active.

The GI7b / GI7bt tube is not the best tube for linear amplification on the 1296MHz band, but because these tubes are relatively cheap and accessible, many of us tried to use them. Based on the characteristics of the tube, one should not expect PA output power higher than 200 - 250W for ordinary operation, with efficiency no higher than 35%, even with very well made and silver-plated cavity. Thermal instability is a constant problem.

Where is source of instability? One part of it is well described by KD5FZX here. The second part, which is the primary cause at lower output levels, is due to the positive feedback inside tube from parasitic energy transfer between anode and cathode. This feedback can increase gain of amplifier and lower attenuation of input reflections of the amplifier. The result is that input matching varies with anode cavity tuning. When anode cavity is detuned by thermal drift, feedback to the cathode decreases and rapidly lowers aplifier gain. This results in much lower output power even if input power is stable.

How to test your amplifier:

• Test your amplifier for highest gain with low output level (aprox. 1 W).   Amplifier should have adjusted ordinary BIAS and HV.  What is the maximum gain when you tune up?  If higher than 13 dB, something is probably wrong.
• Test your amplifier for input SWR with very low input (less than 1W).   Adjust cathode circuit capacitors for best SWR (you will probably need to use a good directional coupler and spectrum analyzer for accurate measurement).  Is input SWR better, than 20dB?  If not, try tune anode cavity out of frequency - for example 20 MHz down or up.  How high influence it has to the input SWR?  If the input SWR changes from previous level by more than a few dB, or input reflections are worse than 18 - 20 dB, something is wrong.
• Switch off HV but leave filament voltage applied.  Connect tracking generator of spectrum analyzer to the output of amplifier and input of analyzer to the input of amlifier.  What is the feedback attenuation?  For an unneutralized tube, it is usually around 13 to 16dB maximum (with adjustment of anode tuning to maximum).  As you can see, it is usually just few dB more than forward gain of amplifier.

... and the feedback creates many of the troubles described above.

Now you should to try neutralize your amplifier:

• Neutralization is accomplished by increasing inductance between the grid of the tube and ground.  It creates a resonant circuit with Cag & Cgc which couples anode voltage to the cathode shifted in phase; it is 180 degrees out of phase with parasitic transfer of energy through Cac.  This phase shifted transfer reduces positive feedback to the point where oscillation no longer occurs.
• Theoreticaly it is simple - because of small variations in construction of different cavities, however, there are differences in inductance and capacitance which can complicate the neutralization process.  With this in mind, VE4MA developed a reliable method which is shown here.   It causes a small increase of grid holder inductance by isolating part of the grid ring from the grid contacts.

I have tested many tubes in different cavity arrangements and can recommend the solution below which is illustrated in the pictures.

• Put a teflon (or paper for test purposes only) strip as shown, width aprox. 24mm, on the tube, install the tube in the cavity and measure feedback transfer as described in point four above.  Feedback should be 13 to 16 dB or a bit better (more dB).  If not, pull out the tube and change width of strip to 22mm and repeat.  For different tubes and grid holders is usually possible find proper strip width to increase feedback attenuation to at least 26 dB.
• Next, rotate the tube in steps of 15 degrees or so.  Remember to be careful of the HV on the anode!  Because grid inside the tube is not absolutely symmetric, you will probably be able find a feedback minimum - at least 30 dB.  Make this position of tube and teflon strip permanent.

• Connect a load to the output and switch the amplifier on.  Test the input SWR.  Although anode cavity tuning will still have an influence on the input SWR, attenuation of reflection on the input should be now at least 18-20dB for any adjustment of anode cavity tuning.  Do not readjust cathode from setting achieved above.
• Test gain of amplifier with low output now - maximum gain, after peaking output with anode cavity tuning and loading, should be between 10 and 12 dB.

Results:

• Your amplifier will have a bit lower gain than before, but will be much more stable. You will need to use higher driving power.
• Because of much lower feedback, you will be able probably setup lower loading of anode cavity and by this way obtain a bit better efficiency without danger of oscillations.

DO NOT FORGET to be careful with HV on the anode!!  Best regards and many successfull QSOs on 23cm.

Vladimir OK1VPZ

www.qsl.net/ok2kkw