2008-12-23
Fixed the MFJ-207 FM Problem
Upon hearing that other HAMs have MFJ-207s free from the FM modulation problems my device shipped with I decided to take another crack at fixing the loop. Inspired by the design of the MFJ-269's ALC loop I simply added an AC feedback network to the LM324 lowering its AC gain but leaving DC unaffected. A easy, two component fix: 10 nF cap and a 1 Meg resistor, in series from pin 1 to pin 2 of the LM324 (output and inverting input).
After this addition I reset the ALC loop control point to that indicated in the manual. The oscillator is now nice and quiet, with only moderate noise sidebands, no more FM!
This fix does nothing about the buffering issues. The oscillator is still pulled around by the load on the bridge (and counter output), but it is now much narrower and suitable for use as a stand-in signal generator. It still doesn't particularly like narrow loads, but it is far more usable.
I took the opportunity to more carefully test the SWR scale. To do this I constructed a series of precision dummy loads into BNC connectors. I've been meaning to do this for a while, and have been getting by with a cermet trimmer pot wired to a BNC, these spot loads are much more expedient and allow doing a calibration (especially of the MFJ-269) in only a few minutes. I lacked some of the smaller resistors in sufficient precision, so I took the next smallest preferred E12 value I had and filed away part of the film until the required value was achieved. The file knick was then filled with liquid electrical tape to prevent ageing effects on the exposed element edges. I was careful to null my multimeter and use the most accurate one I have which seems to be around 1% on the 200 Ohms range when compared to voltages dropped across the similar power resistors.
The 207's SWR meter calibration is still utterly abysmal. You can only adjust it so one load resistance gives an accurate reading. Even reciprocal loads like 100 and 25 ohms, both 2:1, produce quite different readings. With loads on the lower side of the reference 50 Ohms the bridge seems to read low. Doesn't detract too much from its usefulness when tuning up a matching unit, it still dips sharply at 50 Ohms.
2 comments.
Parent article: MFJ-207 Review.