Introduction
This website documents my success in solving excessive VFO drift on my 20 meter MFJ Cub (MFJ-9320), as well as excessive drift on KB9RPG's MFJ-9320, MFJ-9330 and MFJ-9340. Similar solutions should be applicable for all MFJ Cub transceivers that exhibit the same problem since they all use the same part for the slug tuned 3.5 uH inductor identified as L3 which was the root cause of excessive drift. Plots showing the before and after modification results are shown further down this page. A 10 times improvement in VFO drift was obtained, and the radio is very stable within 4 minutes of power on. My modification requires removal of a slug tuned inductor and replacing it with a Toroid Core, and the addition of one fixed capacitor (optional), and slight adjustment of the spacing of the turns on the Toroid Core to adjust the VFO low end frequency.
My modification surely voids any MFJ warranty, so take that into consideration before attempting any such modification on your MFJ Cub. I opted to modify my MFJ Cub since it was not usable as is, and I also thrive on problem solving.
Note: I use an MRF237 for the output transistor on both my MFJ-9320 and MFJ-9340, and have their outputs adjusted for 2.0 watts out. I also cover the hole in the top of the case with a piece of tape to prevent internal contamination, and the excess drift I encounter occurred regardless if this hole was covered or not. All of my drift data shown below was obtained with the tape covering the hole.
Full Details
I built a MFJ-9340 back in 2008 and I'm very satisfied with the minimal drift I encounter with that radio (its VFO drifts slightly upward in frequency). This month (July 2020) I bought and built a MFJ-9320 and the drift was horrible. My MFJ-9320 VFO drifted 5 KHz downward in frequency over the first 20 minutes of power up.
I messed around with capacitors in the VFO and felt I was not really attacking the root cause of the problem and decided to do bench top testing using a spare SA602 in a DIP package to quantify the temperature coefficient of the individual VFO components and found that the root cause of excess drift was the VFO 3.5 uH slug tuned inductor identified as L3 which had an excessive positive temperature coefficient causing the VFO to drift downward in frequency with an increase of temperature. I then did comparison tests of the MFJ slug tuned inductor versus a T50-2 toroid core with 27 turns in place of the slug tuned inductor, and the results were startling. The T50-2 toroid core immediately changed the very negative drifting VFO to a slightly positive drifting VFO.
I have permanently replaced the original MFJ slug tuned inductor L3 with my T50-2 toroid core and added a 22 pf cap to dial in the VFO frequency (I adjusted the turns spacing slightly on the toroid core to fine tune the VFO frequency versus installing a small variable cap). My MFJ-9320 VFO now has slightly positive frequency drift similar to my MFJ-9340 when in receive mode and transmitting approximately 4 seconds every two minutes to obtain the output frequency, and I find this slight positive VFO drift ideal since the VFO drifts slightly down in frequency when I do a one minute key down test which generates a lot of internal heat from the output transistor. I do the one minute key down test to simulate a worst case contest situation where I'm running a frequency and generating a lot of internal heat from the output transistor.
I'm now in contact with MFJ asking them to investigate the L3 situation since numerous people are reporting that their MFJ Cubs are not usable due to excess drift, and suspect MFJ needs to do a better job specifying and/or sourcing this slug tuned inductor. The MFJ Cub really is a nice compact radio and hopefully MFJ will find a solution for L3 as it would benefit both MFJ and its customers.
Below is a plot of my MFJ-9340 (40 Meter Cub) which had very low drift as built and therefore did not need modification.
Removal of the Slug Tuned inductor (L3) and installation of new parts
Slug Tuned Inductor L3 (3.5 uH inductor) was replaced with a T50-2 Toroid Core with 27 turns of number 26 gauge enamel coated wire. Removing L3 is not an easy task, but my technique worked well. I used a solder sucker to remove solder from the 5 pins and 2 tabs associated with L3. Removal of the solder from the 2 tabs is the most critical part. After the 2 tabs are free from solder and free to move I used a pair of pliers to wiggle the metal case on L3 and kept pulling on it until it broke free from the base of L3 and totally out of the unit. I then used small side cutters to cut the remaining plastic base between each pin on L3 which then allowed me to heat up each pin individually and pull each pin out of the board since I did not have a hot air station that would have allowed me to heat up all the pins and tabs at one time for easy removal. The T50-2 toroid core installed is shown in the picture below, and it's the largest of the 3 red toroid cores. Note: you will see a polystyrene capacitor at C9 in my picture below in place of the small MLCC NP0 capacitor supplied with the MFJ Cub. I initially changed this capacitor to the polystyrene capacitor in hopes that it would help fix the original drift problem but it had very little impact, and therefore I see no need in changing C9 from the one that's supplied with the MFJ Cub.
22 pF C0G (NP0) capacitor connected from the junction of C6, C9, C10 and C35 to ground.
KB9RPG MFJ Cubs before and after modification (October 2022)
KB9RPG's 30 Meter Cub was not useable prior to my modification. Prior to modification the radio drifted 8.082 KHz in the first 20 minutes whereas after modification the radio drifted 0.381 KHz during the first 20 minutes which was a 21 times improvement. The after modification results shown below were after I had dripped some candle wax on L3 to lock the turns in place (I now use fingernail polish to lock turns in place). A supply voltage of 13.8 Vdc was used for the drift tests shown below. I then also modified his MFJ-9320 and MFJ-9340 as they also had excessive drift and their before and after modification drift test results are also shown below.
Website created and maintained by Don Kirk (wd8dsb) July 25, 2020