Fax

Number 23 on your Feedback card

Abo ve and beyond

VHF and Above Operation

C,L, Houghton W&61GP San Diego Microwave Group 6345 Badger Lake San Diego CA 92119

The 100 MHz Overtone Oscillator

Last month I covered a temperature control circuit that could be used for a crystal oven control circuit. This month the topic is a crystal oscillator circuit I used In conjunction with the temperature control circuit. The circuits were built to supply a reference frequency for control of a 6 GHz microwave brick oscillator. The bricks normally require an oscillator in the 90 to 108 MHz rangeh with some tolerances for off-frequency operation. The oscillator circuit shown here will work from 90 to 110 MHz, and you can use it with any brick that requires an external oscillator.

Brick oscillators normally come with an internal crystal circuit, but recent surplus items do not have this feature, t picked up quite a few brand-new 6 GHz bricks, and presume others did as well. Additionally, I have just received some information aboul a large quantity of surplus high power (10W) 6 GHz equipment, and I've written for details. VII pass them on as soon as I receive them. It s possible thai the circuits might be adaptable toh or even describe part of. the system.

The schematic diagram of a standard internal oscillator was developed by reverse engineering. While other circuits are available that would workh we want one circuit and one set of specifications for a standard crystal. Crystals could then be ordered from multiple sources, making reproducibility and frequency tolerances as close as possible. By eliminating the wobble in the main wheel, we can hold frequency tolerances and stability to an acceptable level. The International Crystal Co, specification is #585132 for a MS-54XOL (Frequency West) type of brick oscillator.

Temperature Control

The oscillator is constructed on a postage-stamp size PC board. Consulting the parts layout, I drilled and reamed all holes for direct connection. I didn't have time to put a circuit board together for this project. See Figure 1, the suggested layout. The small size of the circuit was not done to get you lo work with microminiature parts, but to minimize the effects of temperature. The smaller the mass of the box, the easier it is to control the temperature, I searched for a suitable small container and could not find anything I deemed acceptable until l noticed lhat a short section of waveguide could enclose the entire oscillator circuit.

Each end of the waveguide is closed off with a plate of scrap brass. On one end, I soldered the plate to the waveguide, closing it off. Carefully holding it in position, i attached the other end to the PC board by the component leads. See the construct lion details in Figure 2,1 slid the oscillator into the waveguide, with a small piece of Mylar™ as insulation to prevent shorts. I didn't solder the power feed end of the waveguide; the circuit itself makes a close fit- Also, covering the entire assembly with a layer of styro-foam holds the unit together and gives excellent thermal Insulation to the heater circuit- Three connections are mounted on the end plate: t, DC power, 2. capacitance adjustment, and 3. coax RF out. Leads run out through the foam insulation.

The crystal oscillator, being small, is supported quite well, even rigidly, by the component leads. Usually the circuits I construct are quite large, to accommodate surplus and junk box components. This oscillator is the exception. But don't get me wrong; this is not the only possible arrangement of parts. The prototype was much larger and it worked well, You can change the layout if you wish. Just keep the crystal leads reasonably short, and the modifications should be just fine.

Transistors, Resistors, and Coils

The transistors I used were 2N930 (NPN), but you can use any good UHF type. If you plan to enclose the oscillator in a waveguide oven, you will need a TO-18 case to fit the PC board. This is a small metal-cased transistor about 3^ 16-inch in diameter. The other components are standard Vi-walt resistors and a mixture of mica and CK-05 ceramic capacitors The RF chokes were home-brewed using 1 megohm, Va-watt resistors for the RFC forms. Any high value will do They showed a good Q, about 35as I recall.

Don't use the inexpensive, imported resistors for the RFCs, since they den t have a flat tubular design They wilt work, but winding the coil over acurved surface is a little tough. Use standard IV watt symmetrical tubular resistors. See Figure 3 for details.

The three inductors: 0,1, 0.39, and

Frequency Accuracy

Test the circuit with the adjustable capacitor set to about mid-value so the oscillator can drop out of oscillation. This is normal since the circuit works only over a narrow range of adjustment. If you use the circuit in a non-oven crystal, set the frequency with the capacitor, and that's it. If you use an oven-type crystal, its frequency will be quite high before the oven cycling. Oven type crystals pull low in frequency as the oven is cycling (heating) lo its preset temperature. When this is translated, considering multiplication to our 5.7 GHz band, errors in the order of 60 kHz are common.

That equals about 1 kHz at the crystal frequency. You may get tired of hearing about frequency accuracy, but when you relate it to the low bands, remember that most rigs can give a reliable readout to 100 hertz. Why, then, should we not expect the same tolerances for microwave converters? If we are going to use two tin cans and a piece of string, a crystal detector is just fine. Using modern, stable transceivers in this microwave converter application is not only a good use of expensive equipment, it's cost effective, too.

By the way, CK-05 capacitors are really chip capacitors in disguise These

i*. luflNi +y± ni;;n-1 ¡¿a1______ in tumnS p Js It itirir'i c

Figure 3 RFC construction. The standard Allan Bradley type "RC" fiat body is great for coil-winding, modern capacitor types are enclosed in a square epoxy case. They are very stable and display a higher Qthan the older disc ceramic types, which you can also use in this circuit. I just happened to be out of some values of the disc types.

A note: With diligence and a pair of wire cutters, you can free the chip capacitor from the epoxy case and remove the chip cap for PC board use. You will destroy several caps until you get the hang of how lo chip away the epoxy at the edges without fracturing the chip cap inside. I caution you to wear a pair of safety glasses. SAFETY ALWAYS COMES FIRST!

New Products

I have received several requests for information on just where to purchase waveguide for construction projects, and how to determine frequency on the 10 GHz band Emcom Industries {Ed

2n930 Dipmeter

tsSJS m.

Figure 1. Parts placement. Larger scale (2.5 size) for test setup; 100 MHz oscillator PC board is 1.25 H x 0.9"Wt to fit inside the waveguide in Figure 2.

fihA

jOMAh&Qri VA*:-Ü3-F CAPACITOR lOpP

0.47uH PF chokes, are wound with #36 enameled wire. The 0.1 uH choke required 6 turns spaced about half a wire-diameter apart. The remaining coils were wound unspaced, with tight turns. The 0.39 uH required 15 turns, and the 0.47 uH choke, 18 turns. I adjusted the turn spacing slightly while using a Q meter to set the inductance on the mark. But this step is not essential, as the values came out quite close to my first measurements.

I coated the finished RFCs with a coat of coil "Q" dope. You could also use shellac or clear fingernail polish. If you use the latter, be sure to use dear polish, since some colored types contain metal flek that might cause trouble. The purpose of the coating is to hold the coil, keeping the turns in place.

A positive ground or negative DC power feed makes the oscillator compatible with the brick oscillator. If you wish, you can convert this to negative ground The second stage of the oscillator unit is a buffer amplifier. This stage isolates the oscillator and load to improve stability and prevent loading of the crystal oscillator.

tsSJS m.

Figure 1. Parts placement. Larger scale (2.5 size) for test setup; 100 MHz oscillator PC board is 1.25 H x 0.9"Wt to fit inside the waveguide in Figure 2.

ENS PL^Tf

FHOf,T f.L ATi

Figure 2 The thermistor is placed in tube on the inside of the waveguide with thermal grease,

C Tk

0 0

Post a comment