PROJECTS

TINKER TOOLS

The ANALOG / DIGITAL / ANALOG (ADA) DIAL is a different kind of "Dial" for your home-brew VFO

Building a reasonably stable VFO for your home brew rig is fairly easy these days even though finding a variable capacitor for the tunable LC circuit can be a daunting task.  One  way around variable capacitor dilemma is to use voltage controlled oscillator (VCO) which is tuned with a potentiometer.  So, you might ask, how does one go about translating the twist of a potentiometer into a tuning dial?  And, for that matter, how does one go about tuning an oscillator with a potentiometer?

Let's look at the "dial" first and get into the circuit, later.  Figure 1 shows a photo of the control panel for a QRP rig I recently built, and Figure 2 shows the template I used for drilling holes in the control panel.  A copy of this template, printed on card stock, can also serve as a "pretty face".

Back to TOP

As you can see, the template in Figure 1 is calibrated to cover a much wider portion of the 40 meter CW band than does the one shown in Figure 2.  Note, also, that this "dial" consists of nine light emitting diodes (LED's) spread across the top of the front panel.  Each diode corresponds to a specific frequency, much like the marks on a traditional analog dial. When that particular LED lights, you simply read the frequency below the LED.  Since this is a 40 meter rig, it is understood that a seven precedes each of the decimal values.  What could be more simple and easy than that for a calibrated analog dial? 

How did I do it?   It is an adaptation of the LM3914 "dot/bar display driver for linear scale" (quite a mouth full) integrated circuit.  Tuning is accomplished using a varactor, sometimes called a voltage variable capacitor (VVC).  In order to vary the capacitance of a varactor, one must vary the reverse bias, small reverse bias yields large capacitance and large reverse bias yields small capacitance.  As the capacitance varies, the frequency varies accordingly (more C, lower frequency; less C, higher frequency).  The varying reverse bias voltage is sensed by the LM3914 and used to drive the LEDs.  Thus, there are a large number of frequencies that can correspond to a large number of LEDs.  My front panel is a bit small for a large number of LEDs, but nine of them fit nicely across the five and one half inches I have available, and there you have it.  My "dial" can indicate nine individual frequencies within the tuning range, and the operator (you or I) can interpolate between LEDs, just like on any analog dial.  Twisting a small vernier dial connected to a ten-turn pot tunes the VFO, so there is lots of band spread.

Back to TOP

 For reference, Figure 3 shows the VFO circuit which is a textbook Colpitts oscillator.  I've used this circuit as the basis for several VFO?s over the years, and it always performs well.  As you can see in Figure 3, my circuit uses 1N5400 diodes as a varactor instead of the traditional variable capacitor.  D1 and D2 do the tuning and D3 acts as a temperature compensating diode.  These are rather hefty diodes and I have found they perform better than some of their smaller cousins in this particular application.  A "real" varactor can, of course, be used, but most of the varactors readily available these days are the surface-mount type, and I don't like working with itty-bitty components. 

Notice, also, that tuning is accomplished using a 10k potentiometer.  A ten-turn pot serves nicely here to give lots of "band spread".  The problem is that dials for ten-turn potentiometers are seldom calibrated in frequency (like, never!).  The "pretty face" shown in Figures 1 is for a 40 meter QRP CW so the "dial" covers only that portion of the band that is of interest.  The dial can, of course, be calibrated to cover any portion of the band or the entire band according to the desire of the builder. 

So, how does one know the frequency at any given setting of the potentiometer?  Like most things electronic, there are a number of ways to accomplish this, some of which can be quite complicated and/or expensive.  The simple, easy, and cheap way is to use light emitting diodes (LEDs) as indicators.  Here's the way it works:  Once the VFO circuit is built, debugged, and tweaked to whatever range of frequency your particular VFO covers, you then crank the dial and carefully record the exact frequency at which each LED comes on. 

That's all there is to it.

Back to TOP

 Well, almost all.  There is a bit of LED driver circuitry involved, shown in Figure 4.   As you can see, there is not a much circuitry involved, just a single LM3914N, nine LED's, and a few other odds and ends with wires to connect everything together.  I used three 30k resistors in series to create the 90k resistor that is connected to Pin 5. 

The LM3914N senses the tuning voltage and drives the LED's.  The most difficult part (as usual) is the mechanics.  You know, drilling all the holes in the front panel so the LED's line up nicely, etc.  The two trim-pots (10k & 1k) are for tweaking band-spread and calibration. 

I built my "dial" on two pieces of perfboard: one holds the LEDs and the other holds the LM3914N and associated circuitry.  The two boards are held together with 1/2 inch spacers and both are piggy-backed" on the front panel with 1/4 inch spacers so the LED's match the holes in the panel.  I connected the two boards electrically via flat cable.

The LM3914 board, unlike the LED board, can, of course, be any convenient size and/or shape, and can be placed anywhere within the box that contains all the other circuit boards, then wired to the LED board on the front panel using ordinary hook-up wire or flat cable.  I mounted my LM3914N board directly behind the control panel, as shown in Figure 5. 

The spacers between the front panel and the LED board are 1/4 inch and the spacers between the LED board and the LM3914 board are 1/2 inch.

 Back to TOP

The component layout I used is shown in Figure 6.

 If you have room for them on your front panel, the LM3914 will drive up to ten LED?s.  Most QRP rigs, however, will probably not have room for that many.  If you really want to go wild, two LM3914's can be wired in series to drive 19 LED's (you can't do 20 because one of the positions is used to daisy-chain the LM3914's).  Or, if your VFO has a very limited range and/or your panel is very small, you may need only 4 or 5 LED's to do the job.  That's the beauty of building your own stuff - -  you can, within certain limits, build it any way you want!  I only have about 5  inches to work with across the face of my front panel, so my LED board is about 5 ⅜ x 9/16 inch and my LM3914N board is about 5 ⅜ x ⅞ inch.  From the center of one LED to the next is one-half inch. 

Back to TOP

Figure 7 shows a photo of the circuit boards for one version of this LED "dial" mounted on a front panel. I cut both boards from larger perfboards.  Note the flat ribbon cable I used to connect the LM3914 board to the LED boards.  You can see the rear of the ten-turn pot used for VFO tuning near the lower right corner in Figure 7.  The other controls and switches, as well as the components on the LM3914 board are hidden in this photo. 

By the way, varactors can also be used to tune a variable crystal oscillator (VXO).  I?ve had success using two crystals (salvaged from ancient CB radios) in a VXO circuit; a 17.xxx MHz and a 10.xxx MHZ crystal to obtain a 7.xxx MHz signal.  By tuning the higher frequency crystal as a VXO, a fairly broad range of frequencies can be had with rock-solid stability, but that's another story. 

Caveat: Don?t expect to obtain precise calibration with this device.  It will let you know approximately where you are in the spectrum, and you can certainly know whether or not you are ?in the ballpark?.  The first problem that prevents a high degree of precision is that this is an analog device.  Like any analog read-out, there is a certain amount of interpolation and interpretation required.  Secondly, the LEDs light up at different frequencies, depending upon which direction you are "dialing".  Going from left to right, a given LED will light at a different frequency than when going from right to left.  The difference is slight, but you need to be mindful of it.  I have not yet discovered how to completely overcome this problem, but there is probably some way to get around it.  What I do (until I find a real solution) is to calibrate left to right, then always dial left to right when I want the most accurate reading possible, given the limitations of the device. 

In any event, these are not major problems for me, and I expect you will not find them to be intolerable. 

I hope that all you builders out there (both of you) give this "dial" a try.  I think you'll like it. 

73,  Dick,  W6BKY

  HOME    PROJECTS   TINKER TOOLS   TUTORIALS   PRODUCT REVIEWS   MARKET PLACE   OTHER

Comments?  Observations?  Suggestions?  Please email:  w6bky1@gmail.com