On Fri, 30 Jun 2006 01:20:52 GMT, David
wrote:

I am currently experimenting with construction of my first Dual-Gate
Mosfet Amp.

Below is an attempt to explain the circuit typology as best I can

Freq = 148 MHz to 152 MHz

Power supply is 8V, Bias on G2 is 4V. Source is grounded

Drain has Variable inductor to Vcc and Cap to ground to form first stage
of a double-tuned circuit. This then couples through a small cap to the
second half of the DTC.

RF input to Single tuned circuit with tapped capacitors attached to G1
of BF998.

C1 to ground is 56pF, C2 to G1 is 15pF, L from G1 to ground is variable
with nominal inductance of 85nH (tunes 75nH to 110nH), Q approx. 100

The Drain has the same type of inductor to Vcc (decoupled Vcc end).
C3 is 18pF from drain to ground (This forms part of the double-tuned
circuit on the output).
Coupling to second half of DTC is 1pF. Second half of DTC C4 = 18pF to
ground and L is same as previous inductors (85nH nom.).

This will later go to G1 of a Dual-Gate Mosfet mixer but for testing I
replaced the 15pF on second DTC with series 18pF and 56pF to form a
capacitive divider down to 50 Ohms for my spectrum analyzer input.

With this configuration the circuit will oscillate on its own when the
input is removed. If I replace the Capo from drain to ground with a trim
cap I can adjust it so the self-oscillation stops but as soon as I try
to change the tuning of the inductor from drain to Vcc, it starts to
oscillate on its own.

If I remove the cap from drain to ground altogether, the oscillations
stop but I almost the ability to tune the Inductor, it becomes extremely
broad tuning.(The input inductor and output DTC inductor tune fine though).

I have tried a series 33R between the Drain of the mosfet and the
Inductor/Capacitor and this did not help.

The measured gain with the capacitor removed was 17dB. Drain current was
4.8mA. 3dB bandwidth was 3.2MHz.

The circuit is laid out "dead-bug" style on a flat bare PCB. I'm not
sure if it is a layout problem or typology issue.
I have seen damping used across tuned circuits in output stages of
rf amps and may try this next. But, I would like to know what is
happening because if I manage to stop the oscillation, I am not sure if
it is marginal and will reappear with changes in temperature or drive
level etc.

Any help much appreciated.

Regards

David

Usually DGFET amps are stable. But they are also very high
gain.

The likely culprit is the input can see the output (incidental
coupling). I always build this with a shield partition right across
the device and also a ferrite bead in the G2 lead near the deivce
as possible (or for SMD in the Source leg bypass cap). These
devices will take off at uhf due to gain and less then careful
bypassing. That includes making sure the G2 bias resistor
(Gate to DC source) does not couple RF back. As part of this
also insure the input and output leads to the connector are short
as possible or even coax.


Allison