Starting at our November 2006 meeting our members embarked on a construction project.The ultimate goal is to build a homebrew version of the MFJ 259 antenna analyser pictured here. Some of our members already have one of these and have
shown that it is of immense help when setting up antennas. But these are relatively expensive items of test gear, so the aim is to produce a cheaper version suitable for homebrew construction so that they may be more widely used by our members.
Our design will be based on that of the Tenna Dipper.
As this is a fairly involved project it will be conducted in stages. The first stage will be to build a frequency counter with lcd display which will be the readout of the unit. Later will be added a variable rf oscillator and bridge components for determining the impedance – an antenna analyser is in effect a calibrated SWR bridge and low power transmitter which can be tuned over the range of interest.
Stage 1 – Frequency Counter
The first component we shall be building for our project will be a frequency counter and 16 character LCD display module. This unit was described by EI9GQ in his Homebrew column in RadCom October 2006, and reference should be made to that for further details. In itself this is a very versatile counter based on a PIC microcomputer and should be a great asset to any shack. The club will be obtaining parts for our members and providing preprogrammed PICs, though there is nothing to stop you programming the PIC yourself.
The project was introduced at our November 2006 meeting where there was considerable discussion on what form the final analyser will take. It was agreed that the Tenna Dipper is pretty basic and limited in its potential. A design with an improved sine wave oscillator and a meter or lcd display for display of resistive impedance is preferred. At this stage the counter can be used as a general purpose item of shack equipment and suggestions for the rest of the analyser will be decided later.
Although some components had been obtained there were still some items outstanding, in particular the LCD displays. It was not possible to distribute kits at the meeting but these were supplied at the following meeting less the displays, which because of cost builders were asked to buy themselves from Maplin.
One or two members have now built the kit with some success but there has been no further progress on the impedance meter part of the project.
Construction comments from G3YMC February 2008
Like some other club members I took my kit away from the 2006 meeting meaning to build it up and never got round to it. It was a little time before I placed an order with Maplin for the LCD display, a combination of finding an excuse for an order large enough to circumvent their carriage charge and also technical hiccups placing an order via their website. Some six months later though I had the display in my hands, whereupon it sat awaiting a suitable opportunity among other things, work and amateur operating. That opportunity arose in February 2008 when I had a little time on my hands, so one day I picked it up, refreshed my memory, dug out the RadCom from the pile and set to work.
Construction was relatively straightforward, though a bit tedious on the supplied square of Veroboard – one designed for DIL integrated circuits though not that suited to a circuit with quite a few discrete components. Newcomers to construction would find a proper circuit board as shown on EI9GQ’s site much easier. It was a bit tedious finding the best place for the crystal and its trimmer cap for instance, an area where short wiring is ideal. After a couple of hours though everything was in place on the main board. Dry testing was done before inserting the PIC chip to ensure that the 5V regulator was working and the supply to the PIC and display was correct.
On to the display. I connected this to the main board via lengths of ribbon cable cut from an old PC cable. Stripping the ends was found easiest by melting the plastic insulation with my soldering iron, the leads then did not take solder that easily as
these leads are designed for IDC connectors and not tinned. But eventually all was well and the display connected, two ribbons with 4 and 5 leads as described in the article. At this point came the acid test. The PIC was inserted, a PP3 battery connected,
followed by a blank display. Adjusting the display contrast potentiometer brought up black squares but no digits. Then it occured to me I had omitted to add the link on the display between pin 5 (R/W) and pin 1 (GND) as instructed. Quickly done followed with the satisfaction of seeing “COUNTER Ver 1.0” and then the frequency readout appear. It works!
Initial testing was done simply by placing it close to my K2 and setting that to its tune position. I noticed immediately that it seemed rather insensitive and although it readily showed the K2 transmit rf I got no indication just by touching my fingers on the input which you normally get with other counters. More significantly the frequency was off, as much as 3kHz at 10 metres. The easiest way of calibrating the crystal oscillator was to listen to it on the K2 at 10.240MHz on the 30m band, and this showed it was indeed off frequency and the adjustment range of the trimmer was insufficient – it was still well off with the trimmer at maximum capacitance.
The solution to both these issues came when I found some notes by G8MNY here. John had also used this kit as a club project and found the same two problems. I implemented John’s mods.
Adjustment of the crystal frequency is suggested in the article as being done by changing the value of the two 22pF capacitors C5 and C6. However a far better way used by John was to add a small capacitor across the crystal itself. John suggested 22pF,
but I found that was too large, possibly because of the extra wiring capacitance in my Verboard version. With a 10pF ceramic capacitor between pins 15 and 16 of the PIC it was found the crystal could be set correctly to frequency with the trimmer in about its mid position. Final adjustment will be done when the counter is inside its box, but by adjusting for zero beat at 10.240MHz I found the counter within 20Hz at 28MHz, which is within the accuracy of my K2!
John reported that his counter required 1V of input to drive the PIC with rolloff at higher frequencies and suspected a problem with the BF199 preamplifier. He decided to change this to a 2N2369A, also changing bias resistor R1 to 18k, and found he had a much more sensitive unit, down to 10mV over most of the frequency range. I decided to do this mod as well and fitted a 2N2369 from the junk box. I also increased the input coupling capacitor C1 to 10nF to allow counting at lower frequencies. I have not done quantitive measurements but it does seem far more sensitive and produces readings if I
just touch the input.
The next stage is to put it into a suitable box and tidy it up so it can become a useful shack tool. My metal work skills are not the best so I imagine it will be a little while yet before this happens…