UPDATE 12/04/2016: Paul, N1BUG reported that he used FT37-W cores in his mod. These are the same mix used for port isolation in the W1VD combiner and are pretty stout as far as mixes are concerned. Paul reports that for 630-meters he needed 17 turns of #24 AWG wire to move the range between 280 and 590 kHz. For 2200-meters, he used 30 turns of #30 AWG wire to move the range between 105 and 220 kHz. He used a DP3T to select 160-meters, 630-meters and 2200-meters.
UPDATE 01/25/2017: I have recently been corresponding with John, GW3VVC, who reports that his early 259 (board dated 1992) is apparently using a different approach to generating a signal and the mod does not work.
TESTING NOTE 10/09/2017: Roger, VE7VV, recently installed this mod for 630m and 2200m and reported the number of turns and core material that he used as well as test load data from the finished product. As expected, 2200-meter numbers are not very good, but they get you in the ballpark. Roger’s data follows:
“Switch 1: 87T on a FT82-77 toroid core, 8.8mH, giving a tuning range of about 103-178 kHz.
On 137 kHz, 50 Ohm resistor measures R=49, X=12, SWR=1.2; 24 Ohm measures R=23, X=1, SWR=2.1; 100 Ohm measures R=92, X=29, SWR=2.0.
Switch 2: 27T on a FT87-77 toroid core, 0.85mH, giving a tuning range of about 330-630kHz.
On 475 kHz: 51 Ohm resistor measures R=51, X=2; 24 Ohm measures R=24, X=0; 100 Ohm measures, R=101, X=0.”
John Langridge KB5NJD / WG2XIQ (Originally written in April 2012)
With WTC 2012 giving the ultimate Valentine’s day present of a new amateur allocation from 472 kc – 479 kc in the 630 meter band, a lot of us that are low band stalwarts began seeing an opportunity to return to the old school paradigm of radio that was designed, built and implemented by the operator, rather than placing an order for a turn-key station from a supplier on the Internet.
I intend to do my best to chronicle my quest for the bottom of the medium wave band, slated to officially open to US amateurs in the future.
In the past, I have written about designing loading coils for electrically short antennas. Those principles can be applied to any amateur band and work very well at 630 meters where amateur antennas will most likely always be shorter than a quarter wavelength. While I can build a loading coil to interface with my antenna system based strictly on the calculations, it’s always nice to be able to take performance data of the system to optimize it. At these frequencies, small changes in frequency can result in large changes in impedance, impacting SWR. Since I have decided to work on a transmit antenna before working on a transmitter, one valuable tool in the design and implementation is a signal source and impedance bridge. Many consumer grade antenna analyzers generally have a frequency limit on the low end around 1700 kc, quite a long way from the area of the band I need to test. Higher end units like the RigExpert and AIM cover the range with a high degree of accuracy but the trade off if a very high monetary cost. Some work done by Ed, KL7UW, who is a member of the 600-meter research group (http://500kc.com), related to the modification of the MFJ-259B to operate on 500 kc. After a number of conversations with Ed, I decided that performing this modification to my MFJ antenna analyzer was a good first step.
Opening up the MFJ-259B analyzer, I noticed a number of potential complications, most notably the number of connections directly from the main PC board to connectors on the chassis. This meant that I might have to completely disassemble the unit, unsoldering connections and then getting them all back together when I was done. This would make the prototyping process difficult, as this modification requires a lot of “cut and try”. The goal of this mod requires gaining access to the band switch solder pads that link to the 66 microHenry inductor (L1) for the frequency range roughly between 4 – 1.7 MHz. Fortunately for me, removing the battery case provided good access. It helped that the battery leads from the AA case to the lower board were very long.
Note the circular pads in the lower right quadrant.
This is the band switch. L1 is the low band range inductor. Identifying the proper solder pads on the band switch without disassembly would have been impossible without the schematic but fortunately that is easy to find on the Internet. Moving the oscillator frequency is accomplished by switching in series inductance in stages for each band range. In order to accomplish 630m, I would simply have to switch in an inductor to move the oscillator to the frequency range that I wanted it. Fortunately, the frequency counter is simply sampling signal and determining frequency so only the oscillator circuit needed to be modified. In order for me to preserve coverage of 80 and 160 meters, I would have to make the mod switchable, shorting across my mod with a switch to use 80 and 160m. I had already cracked the case open so what was wrong with drilling a hole in the chassis for a switch at this point HI!
Tracing the inductors around the band switch, the last one was L1. Using an Exacto knife, I cut the trace between L1 and the band switch.
Next, I insert wires between the inductor and the band switch. Any wire is fine, just don’t use anything too large. Small speaker wire or even zip cord is fine. I had some twisted pair scraps from winding bifilar transformers, so I used that.
Adding wires to switch in inductor for 600m while allowing use on 80 and 160m
Next I would take both of these wires to a mini toggle switch. Connected in parallel with these wires is my new inductor. I didn’t know how much inductance I needed so this was going to be a cut and try. Ed, KL7UW, indicated that his prototyping required 11 turns on a double stack of FT-37-77 ferrite cores. I was unable to identify the cores I had on hand so I decided to develop a baseline and wind a double stack of 11 turns on the cores I had and parallel it with the switch and see what happened.
This didn’t work so well and only dropped the frequency to about 1500kc. This was progress and I was headed in the right direction.
I had recently wound a rather large double stack so I figured it would not hurt to plug it in and see what happened.
I was getting warmer, bottoming out near 800 kc. Just for the heck of it, I decided to try a double layer slug inductor that I had gotten from Paul, KD5IVP, some time back for another project I was working on. Sure enough, it was near 700 kc. The truth is that I really did not want to wind any more turns on the ferrites. Using small wire for a large number of turns is really tough to do without overlapping turns and other problems. I decided that since I was already using a double layer inductor I could afford to make it a triple or quadruple layer inductor and use the slug to fine tune. The key to winding multi- layer transformers is ensuring that the turns for each new layer always start on the same end. That’s a difficult concept to describe here so if you are interested in this project, I will be happy to show you. Needless to say, I came up with a monster inductor and a frequency range right where I needed.
After drilling a hole in the front panel to mount the toggle switch, I used hot glue to insulate the leads on the inductor and wrapped it in electrical tape to keep the windings in place. A little hot glue attached it to the side of the battery compartment.
One observation after replacing the cover was a slight increase in frequency on the bottom end, most likely from capacitance between the coil and case. This does not create a problem but just be aware of this if you intend on replicating this modification. Give yourself a little overhead.
The diode detector of the MFJ-259B is not perfect and the accuracy is not high. Broadcast signals can interfere and give erroneous results. What the MFJ-259B offers is a low cost system that allows the ballparking of a 630 meter antenna system and loading coil. Don’t expect to receive exact values of R and X and understand that any calculations made using measurements from the MFJ-259B will need to be considered as part of a range of values. I can report that the measurements taken with the MFJ unit are generally in agreement with other, more accurate and more expensive means of measurement. Some of those will be discussed in a later article. Additionally, MFJ has recently released a new analyzer that has this functionally as standard – for an additional $100. I hope to one day compare the results of two units side by side.
Special thanks to Ed Cole, KL7UW, for his work in prototyping this modification and answering questions.