What is baudline?
Mystery Signal
Mystery Signal of the Month   Feb 13 2001
We've got signal, but what the heck is it?
That's your mission.  Analyze this bébé.

mystery spectrogram                                                                                                                                                          

download the audio file 4.wav.gz

hints :

  • Setup baudline to be a Netscape helper application as described in the FAQ.
  • Turn on the 60 Hz notch filter in the play deck.
  • Try playing the mystery signal at different speeds, slow it down or drop it an octave.
  • Loop on the transitional points, play it again and again until you start to hear things.
  • Use baudline's harmonic helper bars and delta measurement bars to determine the harmonic relationships.
  • What is the folding frequency?  What direction are the ramps going?  What does it mean?

What is it?

This was a very elusive signal that was fairly quiet and it occurred every couple days at the strangest times.  That 11 Hz fundamental tricked me into thinking it was something really big and for the longest time I thought it was some heavy machinery on the power grid.  It wasn't until I happened to put a computer with a microphone in the room next to it that I made the connection and discovered what it was.  I know, you're saying right now "shut up about your stupid story and tell me what it is!"  Well, here we go.

This mystery signal is in fact a Kenmore Ultra Fabric Care Heavy Duty series 90 washing machine circa 1970.  Other models of washing machines have different signatures but they all share the universal Wash, Rinse, and Spin cycles.  When you look at the spectrogram you will see three distinct discontinuities in the 120 Hz tone (at the 35, 155, and 275 second marks).  This is a timed relay in the washing machine switching modes of operation (cycles).  Let's look at these modes in detail. 

  1. The audio file starts most of the way through the Wash (churn) cycle at time zero.
  2. At 35 seconds a relay flips, the water is turned on, and the heavy Spin cycle begins with the basin struggling under load to rotate.
  3. At 105 seconds the water input is turned off and the water slowly begins to drain.  As this happens it takes about 25 seconds for the slow power up ramp to reach the 11 Hz fundamental which I am guessing relates to 700 RPM.  This is the initial centrifugal spin speed
  4. Then there is 36 seconds of steady 11 Hz fundamental which signifies the top rotating speed have been achieved.
  5. At 155 seconds a relay flips and the spinning stops.  Next is the secondary spin cycle.  Due to the water spun off in the last cycle, the load is much lighter, and it only takes 7 seconds to reach a much faster centrifugal spinning speed (a couple thousand RPM?).  This lasts for two minutes.  Notice the faint 3 Hz fundamental in the 70 - 80 Hz region, this is minor wobble caused by a slight imbalance.
  6. At 275 seconds a relay flips, the spinning stops, and the Rinse cycle begins.  The sound of the water flowing into the washing machine looks spectrally flat and noise like.  The bands and dips that you see are caused by the resonance of the pipes and standing waves in the room and the machine.
  7. The file ends.  Since this is off screen, you don't see it, but the Rinse cycle continues, followed by another churning Wash cycle, and then the final Spin cycle.
Now I haven't answered the most puzzling thing about this file which is "what is causing the mirror images around 120 Hz?"  A frequency ramp (up or down) usually represents a change in the speed (up or down) of a spinning machine.  Since my washing machine only has one rotating bin, it can not obviously be both increasing and decreasing in speed at the same time, much less be spinning at two different speeds at the same time.  So something is going on.  I believe the mirror image is a frequency reflection (think complex?) or aliasing which is caused by poor (or zero) filtering in the main motor.  If you have a better idea, then please let me know.

Jun 5 2006 update:
    A mystery signal enthusiast points out that induction motors create strong 120 Hz frequencies and that any sort of amplitude modulation nonlinearity could be the source of the mirror like IMD tones.  Rotor slippage could also be at work but the slope of the frequency change is in the opposite direction of the washing machine's drum acceleration.  This opposite direction is a contradiction so slippage may not be a factor.  Part of the puzzle is now solved but unfortunately many new questions are asked.  If you have any thoughts about the mystery of the washing machine spectrogram then please contact us.

    To learn more about induction motors, check out this link:

Copyright © 2001 SigBlips.com - group - blog - site map