FM Oscillator

1) Description

This package contains the schematic and a PCB-layout for a FM oscillator using the BF900 Dual Gate Mosfet, and can be used between 90 and 110 MHz. Specific things about this oscillator is that it has very low noise, and it is very stable. If you use proper buffering, a PLL shouldn't be necessary any more (although it is ofcourse possible to use one).

The schematic is here.

The frequency is first set by trimmer C4. Adjustments can be made with R8 (VFO range is about 4 MHz).

Some testing gave me next results:

(Frequency = 100 MHz; Output = 50 Ohm Dummy (20 * 1K/1W par; homebrew);
 Input = Sweeposcillator 0.5-55 kHz (Stereo Coder));
Power Supply    : 15 V DC ('no ripple');
Input Current   : 50 mA
RF Output       : 55 mVpp
Freq. Deviation : 0.005 MHz (after 10 hours; the deviation had taken place
                  during the first hour, after that no deviation)

2) Components/Hints

Use a good quality PCB, the other side is copper, make some connections between the two sides. Build the oscillator in a RF-proof casing, use good connectors and cable, and be aware of all the other RF-rules of building.

To mount the BF900 a hole (diameter about 6 mm) must be made where the casing can hang in. I myself had problems with the 'other side copper' thing, because my oscillator wouldn't turn on (not too enough noise?? just kidding). It was solved when I removed the copper under the BF900 (about 1 - 1.5 cm diameter). If someone has an explanation for it, let me know...

You could use a zener-stabilisation instead of the 7800-regulator. The only problem is that it could cause a DC-ripple on the oscillator, and that will be noticed as humm. A voltage regulator isn't so expensive, so don't cut down expenses on it.

Instead of the BF900, a BF905 or BF907 can be used without any problems. Maybe a BF981 would do also, but that I haven't tested.

Component list of the FM Oscillator with the BF900

Resistors:

(All resistors are metalfilm types except R1 which is a carbon resistor)

R1 :  47k, 1/8 W
R2 :  47k, 1/8 W
R3 :  47k, 1/8 W
R4 : 100E, 1/8 W
R5 : 220k, 1/8 W
R6 : 100k, 1/8 W
R7 : 150E, 1/8 W
R8 : 10k, lin. pot

Capacitors:

C1 : 100 nF Multilayer or ceramic
C2 : 4.7 pF ceramic
C3 : 100 nF ceramic (1 or 10 nF will do also)
C4 :  40 pf trimmer
C5 : 4.7 pF ceramic
C6 :  10 pF ceramic
C7 :  22 pF ceramic
C8 :   1 nF ceramic
C9 :   1 nF ceramic
C10: 100 nF ceramic
C11: 100 nF ceramic
C12: 100 nF ceramic
C13:  47 uF/25V
C14:   1 nF ceramic

Inductors:

L1 : 4 wdg, dia 8 mm, 1 mm CuAg, space 1 mm;
     Mounted 8 mm above PCB (Leads are 4 mm)
L2 : 6-hole ferroxcube wideband HF-choke (5 wdg)
L3 : 6-hole ferroxcube wideband HF-choke (5 wdg)

Others:

Q1 : BF900
D1 : BB405,BB102 or equal (most varicaps with C = 2-20 pF (approx) will do)
U1 : uA78L08, uA7809, uA7810 or equal (MC78LXX, UC78LXX etc)

3) Problems?

If you have certain problems, you have to try to solve them yourself first. But if no solutions can be found you can contact me, so we can try to work it out. This circuit is build by me and it works.

You can also contact me if you have some interesting schematics or material, use E-mail: ericg@stack.urc.tue.nl (at least untill October 1994).

Ofcourse you can use the same adress for other questions or comments on this schematic.

Eric van Gorkom

And much added work by G-Man