Yesterday (April 1, 2006) was a typical Spring day in Eastern Idaho -- a little rain, a little snow, a lot of wind and the temperature a balmy 40 degrees F. It seemed like a good day to spend inside  -- preferably doing something associated with radio that i had not had time for before. I decided to measure the S-meter readings of my FT1000MP as a function of the RF signal supplied to the radio. I had installed the INRAD roofing filter modification and re-adjusted the IF gain (to 11 in my case). I was curious to see just what the S-meter curve would be for IF Gain settings of 10, 11 and 12. These settings were the choices I thought made sense when I was adjusting the gain after installing the roofing filter. The objective was to have the S9 reading on the radio be a -73 dBm signal. [ 50 microvolts across a 50 ohm load is -73 dBm]. I selected 14.100 MHz as the input frequency although the Service Manual suggests 14.200 MHz for setting the S-meter.

I put together a test setup using a WaveTek 3000 RF Generator driving a 10 dB Mini-Circuits inline attenuator (HAT-10) which was connected to the Antenna A input of the FT1000MP.By using the pad as a load at the receiver I was assuring that the output of the generator would not be affected by the input of the receiver not being 50 ohms.

The S-meter of the FT1000MP is a bar graph with the odd values from 0-9 shown and with three bars between the bar assigned to each of the numbers. That meant that the best resolution I could achieve would be 0.5 S unit. In addition, when I started injecting signal I noticed that there was a small amount of hysteresis as I would increase and decrease the signal amplitude enough to make a bar light or go out. I decided to adopt the practice of increasing the signal amplitude to light a particular bar. In that way I felt that the required amplitude for a given bar would be the same every time I adjusted things to make it light.

I also started at the highest reading (S9) and worked my way down. I decided to make signal amplitude readings at each of the odd numbered S-meter values and the bar above and below that value. For example, I would make a reading with a generator output setting that provided an S3 reading and then additional readings at S2.5 and S3.5 . Each time making the output setting adjustment so I was increasing the output to get the bar to light.

The graphical result for the three IF Gain settings is shown below. The specific measurements are in the tables following the graph for people who may have interest in such details. Some additional discussion follows the graph.

I used the EXCEL trendline program to fit fourth order polynomial curves to the three data series. The bold line is the one that characterizes the IF Gain =11 measurements which is the setting I use for my radio. Ideally, we would like the curve to intercept the S9 vertical line at -73 dBm. In the three cases I measured, the gain choices gave an  intercept of -71, -74 or -77 which correspond respectively to 71, 45 and 32 microvolts. Since 45 microvolts is the closest to 50 microvolts that is the one I selected.

The curves are pretty linear until you get to about -93 dBm which is about 5 microvolts. I am guessing this is the point that we begin to see AGC action (on this radio) reducing gain.

As an experiment I added 5 dB and then 10 dB of signal to the S9 reading for each of the IF Gain settings. Each setting showed S9+5 and S9+10 as expected.

The actual data from the measurements is shown below: