Scribd
Upload
7 views10 pages

DATS V2 Measurement and Accuracy

The document provides an overview of small-signal loudspeaker parameters and their measurement using the Dayton Audio Test System (DATS), which calculates parameters like free air resonance and electrical Q from impedance measurements. It distinguishes between small-signal and large-signal parameters, emphasizing the importance of measuring within the linear region to avoid distortion. Additionally, it evaluates the accuracy and precision of DATS, noting that it typically performs better than its unofficial accuracy specification of +/- 2%.

Uploaded by

Novais Silvio
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
7 views10 pages

DATS V2 Measurement and Accuracy

The document provides an overview of small-signal loudspeaker parameters and their measurement using the Dayton Audio Test System (DATS), which calculates parameters like free air resonance and electrical Q from impedance measurements. It distinguishes between small-signal and large-signal parameters, emphasizing the importance of measuring within the linear region to avoid distortion. Additionally, it evaluates the accuracy and precision of DATS, noting that it typically performs better than its unofficial accuracy specification of +/- 2%.

Uploaded by

Novais Silvio
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 10

DATS Loudspeaker Parameter Measurement Overview

and Impedance Accuracy Evaluation


by John L. Murphy, True Audio

Small-Signal Loudspeaker Parameters:


The so called “small-signal” parameters characterize the resonance of a speaker in normal operation.
Like many other electronic devices, loudspeakers have an input signal range where they work
normally. Above that range is some threshold above which the speaker exhibits misbehavior such as
increasing distortion, reduced output, or even catastrophic failure. The signal range where the speaker
operates normally is called the “linear region,” versus the “non-linear region” where distortion rises
rapidly. The small-signal region is that region between the noise floor and the point where the system
is no longer linear. This is the signal range where a loudspeaker’s small-signal parameters are
measured. Small-signal testing is usually performed at the lowest signal level where there is adequate
signal-to-noise ratio. See Figure 1 for a comparison of small-signal and large-signal ranges of
operation.

Figure 1: Small-signal versus large signal operating ranges

DATS calculates a loudspeaker transducer’s small-signal parameters from its impedance


measurement, using techniques established by Thiele1 and Small2 themselves. In addition to the
small-signal parameters DATS measures directly, it also calculates various other parameters.

The small-signal parameters of a loudspeaker are1:

fS The free air resonance


QES The electrical Q
QMS The mechanical Q
QTS The total Q
SD The piston area
VAS The equivalent volume of the suspension
RE The DC resistance of the voice coil
In contrast to the small-signal parameters, the large-signal parameters are intended to characterize the
driver’s performance limitations as it becomes non-linear. Large-signal parameters are usually best
left to the manufacturer to measure, as they may require disassembly or destructive testing. The large-
signal parameters include:

PE(MAX) The thermally limited input power


XMAX The maximum linear excursion
XMECH The mechanical excursion limit
VD The maximum displacement volume

There seems to be a bit of confusion in some DIY speaker circles about what signal level is
appropriate for measuring small-signal parameters, so let’s look at this question more closely. In his
groundbreaking paper titled “Loudspeakers in Vented Boxes,” Neville Thiele1 discusses the test signal
level and states:

“The value is not of great importance, but a standard test figure is 1 volt.”

In his excellent book titled “Testing Loudspeakers,” Joe D’Appolito3 writes:

“The T/S parameters are “small signal” parameters. It is important …to


keep drive levels as low as your instrumentation will allow while still
providing reliable results.”

So there is no reason to measure Thiele/Small parameters at any particular level, as long as the
measurement is not contaminated with noise and is well below the large-signal threshold. DATS
allows the user to set the level of the test signal sweep anywhere in the range from +5 dBu (2 volts
peak) to -10 dBu (0.35 V peak) within the software. This constitutes the small-signal range for the
vast majority of speakers, but note that some micro-speakers could be pushed into the non-linear
region at DATS maximum output level.

The objection is heard occasionally that fS can change with drive level, so it would seem that the
parameters should be measured at higher power levels. It is easy to demonstrate—in the case of well-
designed transducers that are operating normally (i.e., not damaged)—that the fS of a driver does not
change significantly over a very wide range of operation. In fact, a shifting in fS at high drive levels
would indicate the onset of nonlinearity and would not constitute a valid small-signal measurement.
At the other extreme, a shift in fS at low drive levels is an indication that a driver has mechanical
obstructions such as debris in the magnetic gap. This behavior is the basis for the DATS rub and buzz
test. For example, Figure 2 shows the impedance of a transducer (Dayton Audio RS-100) measured at
eight different drive levels over a 70 dB range in 10 dB steps. At the lowest drive levels, the
resonance vanishes into the noise but the shape (Q) and center frequency (fS) remain unchanged over
the 70 dB range of measurement signal level. This is typical behavior for a good driver.
Figure 2: Impedance Response at Progressively Lower Signal Levels for a Properly Functioning Loudspeaker.

In comparison, Figure 3 shows the same test repeated on a different driver with a rubbing voice coil.

Figure 3: Impedance Response at Progressively Lower Signal Levels for a Defective Loudspeaker.

DATS Accuracy and Precision:


The Dayton Audio Test System, DATS, sets a high standard for speed, accuracy, and precision in
audio impedance testing and loudspeaker parameter measurement. While the unofficial accuracy
specification for DATS impedance measurements is +/- 2%, the product actually performs much
better than this and is limited primarily by the 1% tolerance of the calibration resistor. Basic accuracy
and repeatability (precision) test results for six randomly selected DATS units are given in
Appendices A through F. A summary of the results is given in Figure 4 and shows a worst case error
of -2.179% at 1 Ohm compared to error under 0.5% at higher frequencies. The error at 1 Ohm may
seem like a high percentage, but never exceeds 0.025 Ohms. From the raw measurement data shown
in the appendices it is clear that the repeatability of DATS measurements is excellent, and in some
instances the same value is measured for all ten measurements. Frequently the first three digits are
constant for all ten measurements, indicating excellent measurement precision.

Unit ID 1000 Ohms 100 Ohms 10 Ohms 1 Ohms

A -.032% +.040% -.131% -2.179% (-.022 Ω)


B -.023% +.034% -.108% -1.899% (-.019 Ω)
C +.132% -.329% -.482% -2.142% (-.021Ω)
D +.162% +.146% +.065% -.815% (-.0081Ω)
E +.047% +.301% +.284% -.083% (-.0008Ω)
F -.032% -.015% -.299% -1.604% (-.016Ω)

Figure 4: Mean measurement error for ten trials at each impedance for each unit.

The six test units performed with accuracy better than +/-1%, except at 1 Ohm, where the highest
percentage error was -2.179% for unit A. This may sound high until you realize that the actual error
never exceeded 0.025 Ohms. The accuracy specification for these six units (calibrated to +/-
0.1% at 1k Ohms) would be +/- 0.5% or .025 Ohms, whichever is greater. When calibrated with a
1% resistor the accuracy specification would be degraded to +/- 1.5% or .035 Ohms, whichever is
greater. Similar testing on capacitors and inductors reveals this same high level of accuracy. Out of
the box and uncalibrated, it is normal for DATS units to exhibit a basic accuracy of around +/-5%.

References:
[1] A. N. Thiele, “Loudspeakers in Vented Boxes: Part I and II, Loudspeaker Anthology, vol. 1
(Audio Eng. Society, New York, 1978).

[2] R.H. Small, “Vented-Box Loudspeaker Systems, Part I: Small-Signal Analysis, “J. Audio Eng.
Soc., vol. 21, (June 1973).

[3] Joseph D’Appolito, “Testing Loudspeakers”, published by Audio Amateur Press, 1998.
Appendix A: Unit A Evaluation Results
Appendix B: Unit B Evaluation Results
Appendix C: Unit C Evaluation Results
Appendix D: Unit D Evaluation Results
Appendix E: Unit E Evaluation Results
Appendix F: Unit F Evaluation Results

Revised: 26Nov13

You might also like