who can help me,the problem of the Power amplifier self-osc

[springf2000]

dear all:

i am design a power amplifier,but now the power amplifier transistor is self-oscillated,who can give me some experience to solve this problem of the power amplifier!
at first ,thanks

 

[wky]

Re: who can help me,the problem of the Power amplifier self-

It could be a circuit problem or else layout problem.

 

[sergio mariotti]

Re: who can help me,the problem of the Power amplifier self-

I assume that your design involved the "stabilizzation" technique.
If the oscillation is in low freq. (MHz, KHz), the 1st step is place shunt capacitors on the DC feed lines.

At the oscilling frequency, the Capacitors should have enough low reactance compared to the DC source internal resistance.
The capacitance may be realized as close parallel of 2 capacitor (C+0.1C) in "koala" connection.

Also the "path" from Active terminals (e.g. gate, drain) should be short enough because it's an inductance.

E.g for a 10 MHz self oscillation a a close parallel of 3.3nF+330pF sholuld be placed at no more of 5 mm from the drain.


In any case the that path inductance must resonate, with the shunt capacitor, at frequency far from the natural oscillation frequency.

To give an idea of the path inductance, consider 1 nH each millimeter.

Also, take in care the wiring connecting the PA to the power supply.
The wire carring Id should be twisted with its "return" wire (GND), the same for the gate line, the two return wire should be shorted at the PA not at the power suply. Totally 2 twisted pairs.

 

[springf2000]

Re: who can help me,the problem of the Power amplifier self-

thank y and i will test this way today,and i hope y help me more if i meet problem again

 

[springf2000]

Re: who can help me,the problem of the Power amplifier self-

i use only one power supply ,27V,then use the 7812,7805 to transform the 27v to the required voltage of the chip required,then how to twist

 

[goodboy_pl]

Re: who can help me,the problem of the Power amplifier self-

use a twisted wire, one wire is for low voltage (e.g. gate bias) and other for ground/return path; one end of the wire is connected to ground plane near 7805 and other end is connected to ground plane near gate!
With this method there is no loop in gate bias path, thus no voltage will be induced in it due to magnetic coupling to produce positive feedback and oscillation.

BEST!

 

[myicejade]

Adding a series resistor(15 ohm) in Grid and Drain separately

 

[DJ]

Re: who can help me,the problem of the Power amplifier self-

Hi springf2000,
There are several reasons for oscillations:
1. If your output stage consists of two or more transistors in parallel then you have to look at the "odd Mode" oscillation. you can solve it by adding a resistor between the two gates or drains so there is no effect on thr even mode but the odd.
2. Whenever you design an amplifier, look at the stability factor input to output all over frequency (DC to high frequency). Also look at the K factor from DC gate to DC drain . This way one can find low frequency instability.
3. Also simulate the gain from DC gate to DC drain. Best simulation is by using no resistanse source at the gate and loading the drain by series of inductor, capacitor and inductor. Then looc at Vout/Vin for different values of L,C and R load. Try to reduce the gain to the minimum at low frequency. Normaly in package there is no coupling between gate to drain and there is no oscillation when you use high capacitor at the drain and some times at the gate. Many times 50 ohm resistor at the gate helps. The oscillation normaly happens when you measure the MMIC on wafer. There is coupling in thr Probe itself and you have to use high capacitor as close as possible to the gate and drain pads.

D.J

 

[rainer]

Re: who can help me,the problem of the Power amplifier self-

You have to stabilize not only the input and output matching networks but even the supply network.

 

[vfone]

Re: who can help me,the problem of the Power amplifier self-

springf2000
From your post I don’t see any working frequencies and output power.
I understand that you are using 27V supplied transistor(s). A 27Vcc transistor has at least double gain in dB compare to a 12Vcc transistor. This will make the design to be more sensitive to any positive feedback that can appear. In high power amplifier design one of the main reason of the positive feedback (potential unstable) is the position of the input and output match inductors.
For low frequency oscillations usually a series RC filter helps (ex. 10ohms series with 100nF).
Good luck