dBm s-meter vs side scale

  • 1
  • Question
  • Updated 4 years ago
  • Answered
Just curious why these numbers are so far apart?
Photo of Jon - KF2E

Jon - KF2E

  • 703 Posts
  • 234 Reply Likes

Posted 4 years ago

  • 1
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1057 Posts
  • 1101 Reply Likes
Official Response

Quick analogy: Think of the signal you want to hear as a gold watch and noise as muddy water. I give you a bucket with several inches of muddy water and the gold watch in the bottom.  You can see a glimmer of the watch and know it's in there somewhere (you are zoomed out on the panadapter, the noise floor is high and you can see a signal occasionally peak out of the noise).  So you decide to see if the watch is in the bucket and you split the water between two buckets (zoom in on the panadapter).  This causes the watch to be in one of the buckets only and now you can see it better through the mud.  As you continue pouring into more buckets, 4 then 8 then 16 buckets, the water level goes down by half each time (3dB lowering of noise floor in panadapter).  Eventually, the watch is sticking out of the water and you can see exactly what it is.

You'll notice that this behavior is consistent -- as you zoom out the noise floor rises by 3dB with each zoom-out because the bucket (called the bin) has twice the noise in it.  The net-net is the amount of noise you see is dependent on the bandwidth in which you measure it.  So in order to talk about "noise floor," hams have adopted a standard.  We talk about the noise floor in a 500Hz detection bandwidth.  To measure this "official ham noise floor level," just set the width of the filter in your slice receiver to 500Hz and then read the meter.  This is roughly equivalent to zooming all the way in on your FLEX-6000 and the zooming out 6 times (FLEX-6500 or FLEX-6300) or 8 times (FLEX-6700) and then the number read along the panadapter (In reality, the bin size at this point will be ~374Hz and so the panadapter reading will be off by about 1dB).

The slice meter S-meter measures the total power of everything in the bandwidth of the slice filter -- noise and signal -- and provides this as a single power reading.

Another analogy might be to look at a pail of dirt spread across your living room floor.  If you look through a paper towel tube, you might tell your wife that the dirt level is only "a tablespoon."  Your wife wife might look at the whole room and say "No!! There's an entire pail of dirt in here!"

One final thing -- you will also hear hams say "I can see 10dB into the noise floor" or you might look at WSJT and see that it has numbers like -20dB meaning it sees a signal 20dB into the noise.  How is this possible?  You clearly cannot see a signal below the noise floor in your panadapter, right?  Well again, the noise floor is defined as the 500Hz noise floor.  What's really going on is that they are restricting the detection bandwidth (think narrow CW filter) and they they can hear the signal with less noise.  Your ears (brain) do this naturally on CW -- you can hear a single tone stick out of the noise because your brain does the equivalent of an FFT, divides it into buckets and then tells your consciousness "there's a bunch of noise but there a tone of roughly 1000Hz in there."  WSJT uses very narros bandwidth detectors and time-averaging to see "below the noise floor" but what's really happening is that it's reducing the detection bandwidth, splitting the noise into buckets and then looking for signals in the presence of less noise.