Speaker Impedance Measurement

Discussion created by KJBob on Nov 14, 2012

     Many SigmaDSP users are designing speaker systems, for which measuring a speaker's impedance vs. frequency is often useful.  Likely those working with SigmaDSPs already have an evaluation board.  When not needed for project development, the board can be used with a simple adapter to rough out a speaker's impedance curve, as demonstrated here.


     To make the impedance measuring adapter, cut an ordinary 1/8" to 1/8" stereo cable in half.   At each cut end, identify the wires for the barrel (shield), tip, and ring of the 1/8" plugs.  Solder these wires, a 10 ohm resistor, and the free ends of an alligator clip lead (also cut in half) as shown in the diagram below.  Insulate the joints with shrink tubing.  This adapter creates a voltage divider with the unknown impedance, while bringing out measuring points to the ADCs.



     Load the attached project file into SigmaStudio.  It runs as-is on a ADAU1701MINZ eval board, and could be modified for others.  It includes these functions:


  • A sweep generator with three frequency ranges:    20 -- 120 Hz,  100 -- 600 Hz, and 500 -- 3000 Hz.
  • A synchronous demodulator that phase-locks to Zx's current waveform, to measure Rx (in-phase) and Xx (90 deg)
  • Math to present Zx as a magnitude (1 - 50 ohms) and phase (-45 to +45 degrees)


     When trying this for the first time, connect another 10 ohm resistor at Zx.  Download the project to the eval board.  Open its two output real-time display graphs.  The magnitude graph should read 1.0 (x10 ohms), the phase graph should flat-line at 0 (x10 degrees).   Try the other frequency ranges with the mux buttons.  Sweeps repeat every 10 seconds, you can usually see a glitch in the curves where the frequency flies back from max to min.  Freeze each real-time display as the glitch approaches the origin to produce a static graph.


     Short the Zx clips together, of course you should read zero ohms with a near-zero phase.  On the other hand, with the clips open-circuit, Z goes off the graph.  Current through Zx is now zero, causing the sync detector to lose lock and produce erratic phase readings.


     Once the setup is working to your satisfaction, you can make unknown measurements.  Below are impedance phase (top) and magnitude (bottom) graphs I measured on a small DJ speaker.  Observe the resonances near 60 and 100 Hz: