VARIABLE FREQUENCY OSCILLATORS

VARIABLE FREQUENCY OSCILLATORS

By: Guillermo H. NECCO, LW 3 DYL

While I was watching at some friends while assembling the CW equipment presented in a previous edition, I noticed that there was no problem in assembling the PCBs for the receiver and the transmitter but the troubles began at the time of adjusting the VFO (variable frequency oscillator). Because of this I will try in this article to clear all doubts and questions that you may have regarding to this, the most delicate part of the equipment, because on its stability depends its quality.

Failure in the start of the Vackar oscillator:

When Roberto, LU8DIW came to show me how was his GACW40 going I took the opportunity and I tried to calibrate it. At that moment we noticed that it did not oscillate. As it is usual in this cases we check the card looking for a misplaced component or a failed one but no, all voltages were right but it did not start. Why do these things happen? In the beginning because the disparity in the components that are used. Being quite a few of them pull-ups, if you use them in an audio circuit they may work fine but when you use them in an radio frequency circuit it isn’t always so. I use to say than together with Ohm’s and Kirchoff’s laws in RF we are also subject to Murphy’s law ( that that says that if a toast fells over the rug it will do it over the side with jam). Why this oscillator does not start if I assembled a bunch and all them worked? In this particular case it accomplish with Flage’s Law on the unborn perversity of the inanimate things: from any inanimate object no matter its composition or configuration, could be expected that at some time it will act in a totally unexpected way because of reasons absolutely obscures and mysterious.

When an oscillator doesn’t start means that it doesn’t have enough positive feedback as for to begin the oscillation and maintain it in the time. In the particular case of this oscillator we take out the .001uF that is connected between the base of the transistor and ground (See figures 1 and 2). As soon as we did this it started oscillating. We should then keep trying different capacitors (470pF, 220pF, etc.) till it stops oscillating again. Then we should solder the condenser with the previous value to when it stop oscillating. We do not have to leave it without the capacitor because it becomes unstable. This is a case of Skiness Constant, that is the amount that has to be added or subtracted or by which has to be multiplied or divided the obtained result to get the one that must have be reached according to the accepted theory.

THE UNEASY 1N4007:

My friend Saul, LU7EJT commented that in his equipment 3DY the oscillator had frequency drift and that he solved the problem by eliminating the diodes 1N4007 that it use like varicaps and installing a variable capacitor with plates and air dielectric, with which the oscillator became stabilized immediately. This happens because the junctions are sensible to the room temperature. In this case, I use common power supply diodes because they are easy to get, but although I did not have problems, many have commented on their instability. This is because there are many production differences among components identified with the same number and also because: it is a power supply diode that we are making work in a zone it was not designed for. Why do we do this? Basically because is a component easy to get. It would be easy for me to use varicaps but will everybody be able to get them?

A stable VFO with usual components:

Because the problem was with the junction of the diodes used as varicaps I designed another oscillator for those that have had problems with the Vackar and wish more stability. It is a Colpitts oscillator with parallel tuning (see figure 3). You will notice that I have changed the 1N4007 by two transistors BD 139, of better quality, of which I use their base-collector junctions. There is to be taken into consideration in this oscillator that the .001uF feedback capacitors should be polyester or silver-mica; do not use disc ceramic capacitors because they provoke instability due to variations in the temperature. Also the coil is more critical. I advice to begin with something like 25 turns of wire # 24 (.50 mm2) and measure the frequency till reaching the 3MHz. Remember that if the frequency is too low you have to take out turns one or two at a time till you reach the specified frequency. Once you get the required number of turns it is advisable to dip it in varnish. The card is shown at figures 4 and 5.

How can I measure the frequency?

Obviously with a frequency counter. Those that doesn’t have this instrument and doesn’t have the economic means for making it (in the magazine there are various published) will have to check if a friend or the Radio Club may have a general coverage receiver with which we may be able to tune the signal from the oscillator and read the frequency in its display.

For those that do not have any of this alternatives but have for instance a multimeter that measure frequencies generally up to 200 KHz, I’m presenting a little gadget that divides the input frequency (the VFO output) by10 or 100, so a 3MHz frequency is transformed in one of 30 KHz easily measured with the economical multimeter. See figure 6.

Another way, included in the card, and that is selected by a switch, is a by 2 and by 4 divider. What is this for? Let’s suppose the case of the oscillator at 3MHz. If we divide it by 2 and we near it to an AM receiver we will hear a signal at 1500 KHz. If we have an oscillator, by example, at 4.920 MHz and we connect a little piece of cable from the divider by 4 to the AM receiver we will hear its signal at 1230 KHz.

With this trick it is possible to "hear" frequencies higher than 20 MHz. With this I would like to show to the newcomers that in electronics we are able to replace money with ingenuity.

Translation by Juan Ferrari KG4FSN