I built a 13.56MHZ RF power amp (4W) however the PA was oscillating and noticed that the AC signal was shorting to ground. However my prof told me to took off the 0.1uF capacitor (C55, grid C3) and and put 1 600pF capacitor(in parallel, to GND) just outside the gate. Note that R58 was not placed in the PA.
What is the rational behind it? (referring to the capacitor)
If R58 is not placed, there is no reason for entire Manual Bias Adjustment circuit (R56 and others).
I would recommend from the beginning to disable the Auto Bias Adjust circuit (remove R42) and use only the Manual Bias Adjust, populating R58.
In this way, eliminate some feedback that is possible to appear in the circuit.
If R58 is not placed, there is no reason for entire Manual Bias Adjustment circuit (R56 and others).
I would recommend from the beginning to disable the Auto Bias Adjust circuit (remove R42) and use only the Manual Bias Adjust, populating R58.
In this way, eliminate some feedback that is possible to appear in the circuit.
so what is the use of putting a Capacitor of lower value outside the gate of the transistor? Some how the whole amplifier becomes stable after adding the 600pF cap
Doesn't make sense.
If R58 is not populated, the value of C55 is not such important.
If it works when change C55 from 0.1uF to 600pF, for sure R58 is in place.
When you say R58 is not placed, do you mean it's shorted or open? Because if it's open then the value of C55 should make absolutely no difference in the amplifier behavior.
Also I'm not seeing how the auto biasing could work effectively... you'll set the bias current in Q2, but its drain and gate are tied together, unlike Q1 whose drain bias voltage is fixed at 12V.
DNP (Do Not Populate) means there is nothing, no resistor at all (open space).
So if is DNP, C55 doesn't matter the value. But you already mentioned that with R58 not populated, it matter.
Regarding the Auto Bias you are right, is more likely a fix voltage bias (not auto).
Obviously, the idea of the "auto-bias" is to generate a Vgs value for a defined Id, assuming both transistors share the same Vgs-Id characteristic and temperature dependancy, and the Vds influence can be ignored. In fact, you shouldn't expect too much from this simple method, which is probably not more than just an idea.
When you say R58 is not placed, do you mean it's shorted or open? Because if it's open then the value of C55 should make absolutely no difference in the amplifier behavior.
Also I'm not seeing how the auto biasing could work effectively... you'll set the bias current in Q2, but its drain and gate are tied together, unlike Q1 whose drain bias voltage is fixed at 12V.