Measuring Resistors

A number of customers have asked for real results of using Speaker Workshop. I tried to use Speaker Workshop with the Wallin jig to test resistors.

Test Bed

a Wallin jig with changes

I used banana jacks with alligator clip leads to measure resistance

an HP 3455 Volt/Ohm Meter

a PC with 350 MHz Pentium II using a Creative Labs PCI128 sound card

Methodology

I carefully tested the two reference resistors in the Wallin Jig. My jig uses 1/4W metal film resistors and I measured them in-circuit with the 3455 (to include solder resistivity and switch resistance). One resistor was 9.975 ohms and the parallel (a and b) combination was 4.996 ohms.

Then I calibrated Speaker Workshop using the Options / Preferences impedance calibration (Test... button). I left input impedance at 1000KOhms.

I measured 5 sample resistors (each 1/2W metal film) using the 3455 and then quickly put them on the jig and measured them using Speaker Workshop.

The Speaker Workshop sample rate was 44100 and the sample size was 16384.

This entire process was done within about 10 minutes to minimize temperature drift. The 3455 was warmed up for over an hour and the PC was left on all night before testing.

Results

Analysis

Speaker Workshop is really using a simple voltage divider to measure resistance. As a result, when you use something like the Wallin jig with a fixed voltage divider resistor, your results are extremely accurate within a limited range and accuracy drops off outside that range. The program uses the reference resistors - in this case 10 and 5 ohms - to figure out the accurate band.

What we see with the above results is excellent precision from 2-15 ohms with dropoff to about 1% error at 33 ohms. This is virtually perfect for speaker testing and building - hence the value of a jig with soldered in resistances and consistent results.

If you want to test larger resistors I recommend a bigger dropping resistor.

Finally, I was surprised and impressed with the accuracy shown above - it may have been a fluke but the test was done without any attempt to fudge the figures. I was extremely careful with the calibration and getting results quickly so resistors didn't drift, however.

Precision is a function of the linearity of the ADC used on the sound card and the linearity and noise of the input buffering circuitry and the noise of the output amplification.