A Guide to Building Three Impedance Jigs and Measuring Impedance and Thiel-Small Parameters in REW

May 27, 2024

 There are a few threads on various forums related to building an impedance jig and how to measure impedance using various software (typically the LIMP module of ARTA or REW). However, I do not know of one that sets out in a purely elementary way how to do these things in a single post. This article is intended to show (1) how to build three different impedance jigs that anyone who has soldered a simple passive crossover can follow and replicate, and then proceed to (2) show how to measure impedance and calculate Thiel-Small (T/S) parameters (see HERE and HERE for T/S parameters) using REW, showing every step and the exact wiring connections. (I will refer to terminology used in REW and LIMP but decided not to describe the measurement steps in LIMP in this article, as it is readily available elsewhere.)

The bottom line is that you do not need to purchase a commercial solution for measuring impedance and calculating T/S parameters. (Full disclosure, I have a DATS v3 that currently runs for approx. $130. It is convenient, but I will likely sell it soon now that I have built Jig 3 of this article.) Jigs 1 and 2 are literally just a single resistor and in my example, I do not even solder anything for Jig 1. You can build a cheap version of Jig 1 for under $10. Jig 3 has five resistors, a capacitor, and some jacks and switches. This write-up is for those people who are not quite sure about such jigs but might take the leap with a detailed step-by-step explanation.

The other thing I have noted is that people following the threads often make mistakes, not in the actual build of the jig, but connecting everything correctly. This is not surprising as LIMP adopts Left Output as typical, REW adopts Right Output as typical, and I often see write-ups that say one thing, but the screen shot says another. Other terminology also differs across write-ups. So, I will include numerous photos of the exact connections for each example.

I have seen different terminology referring to the audio interface – typically a soundcard, PC, or pre-amp. I will simply use the generic terminology “Audio Interface.”

TABLE OF CONTENTS

1. Impedance Jig 1

2. Impedance Jig 2

3. Impedance Jig 3

4. Measuring Impedance and T/S Parameters Using REW

A. Setup and Calibration in REW

B. Measuring impedance in REW

C. Calculating Thiel-Small Parameters in REW

5. Measuring Impedance and T/S Parameters Using LIMP (ARTA)

6. Calculating T/S parameters in VituixCAD

7. Comparison of Impedance, T/S Parameters, and box modelling

A. Impedance and T/S parameters

B. Box-modeling implications

Sections 1 through 3 will demonstrate how to build three different impedance jigs. Jigs 1 and 2 are basically the same thing, just with a different project box and different connection types. These will allow you to measure impedance (and correspondingly determine Thiel-Small parameters) using the jig and a two-input, two-output Audio Interface. You will be limited to the output power limitations of the Audio Interface.

Jig 3 provides extended functionality that allows switching a switch and taking SPL measurements for crossover design without changing a lot of cables. Additionally, it utilizes the output signal through a power amplifier, allowing impedance measurements at higher output levels, which can be more accurate (as well as the ability to evaluate the robustness of the impedance of a driver at increasing levels of power).

Of course, there are a variety of ways to build such jigs and the examples below are just a few. You do not need a nice project box or chassis; you can just solder some wires and a resistor together. Also, the placement of the jacks, terminal posts, and switches can be places wherever they are the most convenient for your own setup.

Below are schematics for impedance jigs from ARTA (LIMP) and REW. The diagrams/schematics do not mean a whole lot to me, I need something more intuitive so I present point-to-point wiring diagrams similar to a speaker crossover that should make it easy for anyone to see clearly what the connections are.

1. Impedance Jig 1

This is the simplest of the three jigs presented and follows almost exactly this “sticky” post at Parts Express Tech Talk.

Required Parts:

• 1 x Project Box

• 3 x RCA Jacks (or 1/4-inch TS jacks)

• 1 x Dual Speaker Binding Posts (or two single posts)

• 1 x 100-ohm resistor

• Hookup wire

I found some female RCA-to-screw-terminal connectors that I had bought years ago, so I built this jig with those and some small wire nuts and did not have to solder anything. The project box I had on hand was a little on the small side, so it was a pretty tight fit in the box.

To keep things generalized, I am simply using the labels RCA 1, RCA 2, and RCA 3. If you build your own, you should label it in whatever intuitive terms are best for your own needs.

• RCA 1 (Output from Audio Interface): This is connected to the audio output from one of the channels of the Audio Interface. In the REW diagram this is Right, in the LIMP diagram it is Left, so here it is generically labelled X. It does not matter which output is used, but it will affect how the inputs are connected and the setup within REW or LIMP.

• RCA 2 (Reference): This is referred to as the Reference Channel or Reference Input in REW and LIMP

• RCA 3 (Impedance): This is referred to as the Load Input in REW and Impedance Input in LIMP

• The 100-ohm resistor is referred to as the Sense Resistor in REW and the Reference in LIMP

Things to note:

• Each speaker terminal post has just one wire, one terminal post is positive (red) and the other is ground (black).

• Each RCA jack has two wires, one positive and one ground.

• The four ground wires (three from the RCA jacks and one from the speaker terminal post) connect together and do not connect to either side of the resistor.

• You only connect one OUTPUT from the Audio Interface to the jig (RCA 1). It can be LEFT or RIGHT (or possibly labeled 1 and 2 on your device), it does not matter. So, I have labeled these generically X and Y.

• You connect both Audio Interface INPUTS to the jig (RCA 2 and RCA 3). It does matter which one connects to which, depending on which output you used. Whichever OUTPUT you selected from the Audio Interface X (LEFT or RIGHT) should be connected to RCA 3 (“RCA out to Audio Interface INPUT X”) and the other channel that was not used as the OUTPUT connects to RCA 2 (“RCA out to Audio Interface INPUT Y”).

  • • To state this another way, RCA 1, RCA 2, and RCA 3 can be (Right Output, Left Input, Right Input) or (Left Output, Right Input, Left Input).

My Audio Interface (Motu M2) has both ¼-inch jacks and RCA jacks for outputs, but I am using the ¼-inch jacks, so I am using ¼-inch TS-to-RCA cables as shown in Figure 1-D. My connections are:

• Audio Interface connected to computer via USB cable

• ¼-inch-TS-to-RCA cable connected from Audio Interface LEFT OUTPUT to impedance jig RCA 1 (In Figure 1-F the sliver cable is a TRS-to-TRS cable with a TRS-to-RCA adapter on one end)

• ¼-inch-TS-to-RCA cable connected from impedance jig RCA 2 to Audio Interface RIGHT INPUT

• ¼-inch-TS-to-RCA cable connected from impedance jig RCA 3 to Audio Interface LEFT INPUT

• Two banana-to-alligator cables from the speaker binding posts to the positive and negative terminals on the speaker driver

2. Impedance Jig 2

Impedance Jig 2 is the simplest to use, as it incorporates the necessary cables in the jig so you do not have to keep track of any additional cables, but it is a little more difficult to build, only because I had to solder very thin wires in a very small project box.

Required Parts:

• 1 x Project Box

• 1 x ¼-inch TRS male-to-¼-inch TS male stereo Y-splitter

• 1 x 100-ohm resistor

• 2 x Alligator clips and speaker wire

The TRS to TS Y-splitter is simply to save money. Two TS-to-TS cables would work, cut them both in half and discard one of the halves so that you have 3-cables with TS jacks on the ends. The TRS to TS Y-splitter allowed me to just cut it in half and use only one of the channels of the TRS portion of the cable. This allowed me to have only two wires going into the project box from the Audio Interface rather than three.

It was more difficult because the copper wires inside of the PVC jacket are very thin. I could trim them back with a utility knife if I were very careful, but they were much too thin for any wire strippers that I had.

Things to note:

• Red and black wires to the speaker are just 16 AWG speaker wire that I had on hand and some heavy-duty alligator clips.

• The TRS cable coming into the jig has left and right channel signals as well as the copper shield as ground. I only needed one channel, so I clipped the other channel (small red dot in the diagram) right at the PVC jacket.

• The dual TS cables coming in from the Y-splitter have a left and right channel as well as the copper shield as ground.

• The three ground wires connect and do not connect to either side of the resistor

• You only connect one OUTPUT from the Audio Interface to the jig (labeled “from Audio Interface OUTPUT X” in Figure 2-A). It can be LEFT or RIGHT (or possibly labeled 1 and 2 on your device), it does not matter. So, I have labeled things generically X and Y.

• You connect both Audio Interface INPUTS to the jig (labeled “out to Audio Interface INPUTS X and Y”). It does matter which one connects to which, depending on which output you used. Whichever OUTPUT you selected from the Audio Interface X (LEFT or RIGHT) should be connected to INPUT X and the other channel that was not used as the OUTPUT connects INPUT Y.

In Figure 2-B you can see that the cable I bought had a black TRS connector split to Red and White marked TS connectors. I wired the jig such that Red served as the Reference connection. So, I can connect (Black, Red, White) as either (Right Output, Left Input, Right Input) or (Left Output, Right Input, Left Input). My Audio Interface (Motu M2) has both ¼-inch outputs and RCA outputs, but I am using the ¼-inch jacks. My connections are:

• Audio Interface connected to computer via USB cable

• ¼-inch TRS connector (black on my cable) of impedance jig connects to Audio Interface LEFT OUTPUT

• ¼-inch TS connector (red on my cable) of impedance jig connects to Audio Interface RIGHT INPUT

• ¼-inch TS connector (white on my cable) of impedance jig connects to Audio Interface LEFT INPUT

• Two banana-to-alligator clips from impedance jig attach the speaker binding posts to the positive and negative terminals on the speaker driver

3. Impedance Jig 3

[I would not have understood nor attempted to build this jig if user 4thtry would not have posted a detailed build thread on the MAC forum. The original schematic is by user dcibel on several forums and Reet on HTGuide.com.]

Impedance Jig 3 is slightly more complex, as it adds a few components to the jig. However, it is not only an impedance jig. Rather than relying only on the Audio Interface’s output signal to measure impedance, it runs the output through a power amplifier and can both (i) generate the signal for measuring impedance using the Audio Interface’s inputs (as with Jig 1 and Jig 2), and (ii) generate a more powerful signal to the loudspeaker to use an XLR microphone to measure SPL for crossover design. This simplifies the measurement process by eliminating the need to disconnect the impedance jig and change connections while transitioning between impedance measurements and SPL measurements.

When the Impedance/SPL switch is set to Impedance, INPUT X of the jig is connected to the Audio Interface and functions as the impedance input. When the Impedance/SPL switch is set to SPL, the 10-ohm resistor is by-passed, INPUT X is disconnected from the Audio Interface, and an XLR microphone is connected to the Audio Interface.

The other feature the jig adds is a switch to include a high-pass filter (a simple capacitor) to protect tweeters from potentially damaging low frequencies when taking SPL measurements. The Reference Input (INPUT Y) functions as the Reference signal for impedance measurements and functions as the Loopback signal for SPL measurements. When the High-Pass/Full Range switch is set to Full Range the signal comes into the jig and (i) goes out to the speaker to be measured by the XLR mic, and (ii) goes out to the Reference connection for Loopback measurements. When the High-Pass/Full Range switch is set to High-Pass, the signal goes through the High-Pass filter (i.e., the capacitor) and the signal (minus the low frequency part that has been filtered out) goes to the speaker (tweeter). The high-pass filtered signal also goes to the Loopback Input and the Audio Interface compensates for the effects of the inserted capacitor

The additional 47K and 22K resistors step down the voltage from the power amplifier from a high-level signal to line level to reduce the chance of accidentally damaging the Audio Interface from an excessive output signal. This does not completely safeguard the Audio Interface, so you should still be careful. The ARTA jig presented in the Application Note AN1: ARTA Measuring Box includes a pair of zener diodes for protection if you are interested. Additionally, for the resistors I just followed the schematic posted in the links already provided; they do determine the maximum power that can be measured without clipping the audio interface input, but that is something that I do not understand well enough to comment on.

Required Parts:

• 1 x Project Box

• 2 x ¼-inch TS jacks

• 1 x 10-ohm, 10-watt resistor

• 1 x 1K ohm resistor

• 2 x 22K ohm resistors

• 2 x 47K ohm resistors

• 1 x 22 uF, or similar, capacitor

• 2 x Single pole, single throw (SPST) rocker switches

• 2 x Dual speaker binding posts (or 4 single binding posts)

• Hookup wire

Things to note:

• Not included in the Required Parts is a Power Amplifier and necessary cables:

  • • 1 x TS/TRS-to-RCA cable to send the output signal from Audio Interface to Power Amplifier

    • 2 x TS/TRS-to-TS/TRS to connect the jig to the Audio Interface’s inputs

    • Speaker wire (and optional connectors, e.g., spade, banana) from Power Amplifier to jig

    • Speaker wire (and optional connectors, e.g., spade, banana, alligator) from jig to speaker

• You connect one OUTPUT from the Audio Interface to Power Amplifier. It can be LEFT or RIGHT (or possibly labeled 1 and 2 on your device), it does not matter.

• For impedance measurements, you connect both Audio Interface INPUTS to the jig (TS to Input X and TS to Input Y). It does matter which one connects to which, depending on which output you used. Whichever OUTPUT you selected from the Audio Interface OUTPUT X (LEFT or RIGHT) should be connected “TS out to Audio Interface INPUT X” (Impedance) and the other channel that was not used as the OUTPUT connects to “TS out to Audio Interface INPUT Y” (Reference).

My Audio Interface (Motu M2) has both ¼-inch outputs and RCA outputs, but I am using the ¼-inch jacks. My connections are:

1.       General (Both Impedance and SPL measurements):

• Audio Interface connected to computer via USB cable

• ¼-inch TS-to-RCA cable connects Audio Interface LEFT OUTPUT to Power Amplifier Left Input

2.       Impedance mode:

• ¼-inch TS-to-TS cable from impedance jig “TS out to Audio Interface INPUT Y” (Reference)connects to Audio Interface RIGHT INPUT

• ¼-inch TS-to-TS cable from impedance jig “TS out to Audio Interface INPUT X” (Impedance)connects to Audio Interface LEFT INPUT

• Two banana-to-alligator clips from impedance jig attach the speaker binding posts to the positive and negative terminals on the speaker driver

3.       SPL mode:

• ¼-inch TS-to-TS cable from impedance jig “TS out to Audio Interface INPUT Y” (Reference) connects to Audio Interface RIGHT INPUT

  • • • §  “TS out to Audio Interface INPUT X” (Impedance) on the jig is not connected for SPL measurements

• XLR cable from measurement microphone connects to Audio Interface LEFT INPUT

• A long run of speaker wire from the impedance jig binding posts labeled “To Speaker” connect to the positive and negative terminals of the speaker driver.

Figures 3-I and 3-J show the connections for taking SPL measurements rather than Impedance measurements. Since this article is focused on impedance measurements, how to take SPL measurements of a speaker for crossover design is not covered. While the measurement process varies a little depending on the software used, guides for taking measurements for VituixCAD in REW, ARTA, CLIO and SoundEasy are available HERE in pdf form.

4. Measuring Impedance and T/S Parameters Using REW

General instructions to measuring impedance and estimating T/S parameters in REW are HERE and HERE. This section will follow the procedures from REW and include screenshots that apply to the setups described in the sections above (e.g., Left Output, Left Input, Right Loopback Input).

A. Setup and Calibration in REW

i. Preference setup

Select preferences from the top menu bar or the wrench icon in the upper right corner of REW.

Select your audio device, sampling rate, inputs and outputs, and the calibration file for your audio device (if you have not created a calibration file for your audio device go HERE). As mentioned in the section on building a DIY impedance jig, I will use Output X = Left (Output 1 in the screenshot). There is only one output, so Timing reference output is also Left (1). Input is Left (1) and Loopback input is Right (2). [Note: the instructions at REW assume the RIGHT Output is used, so the screenshots and descriptions from the REW website are different from the ones here.]

The input signals need to have the same gain. In my case, my Motu M2 has LED lights on the front that make it easy to adjust. However, most Audio Interfaces do not have such LED lights. In this case, connect the impedance jig with the test leads open (i.e., not attached to a speaker or to each other) and use the signal generator in REW to play a sine tone at 1 kHz at the intended measurement level while observing the input levels on the REW Level meters (See image below for the Generator and Levels options). Adjust the input gains so the input levels match to within 1 dB and are not clipping.

The levels in Figure 4-B below are NOT good (the Right Input is 6 db less than the Left Input) and will generate an error message when trying to calibrate. The small knobs for the input gains on my Audio Interface were NOT far off, the Left Input was a little less than 12 o’clock and the Right Input was close to 11 o’clock. So just having the input gains “pretty close” is not good enough, so do not try to skip this step thinking you do not need to.

Adjusting the Input Gain knobs on the Audio Interface, I got very close, as seen below.

ii. Open circuit calibration

Open the Measurement panel and select Impedance in the top left. Double check your input and output settings. Mine are set correctly from entering the information in Preferences (in Step 1) but if you go directly to the Impedance panel they may need to be adjusted.

The Rsense resistor is 98.3 ohms (I measured my 100-ohm 5% tolerance resistor iwth a multimeter and this one was 98.3 ohms) because I am using Impedance Jig 1 for the screenshots; however, if I were using Impedance Jig 3 it would be 10 ohms. Select the 1. Open circuit cal button, which will be red if you have not calibrated yet.

Disconnect the test leads (i.e., not attached to a speaker or to each other) and click "OK". A sweep will run for about 15 seconds. If your input channels are not the same you will get an error message, but this should not happen if you already adjusted them as described above. In my case, a measurement was created called Impedance open cct that says, “Measured was 100.726% of reference.”

iii. Short circuit calibration

The Measurement panel closes, so open it back up and the 1. Open circuit cal button is now black and select the 2. Short circuit cal button which is now red. Follow the directions to short the leads together and select OK. After calibrating, my REW created a measurement called Impedance short cct that looked like a flat line at 0 ohms as seen below in Figure 4-H.

iv. Reference calibration

Lastly, I opened the Measurement panel again to calibrate the test leads. The 2. Short circuit cal button has turned black and the 3. Reference cal button has turned red. I took a 3-ohm resistor from my parts drawer and measured it with my multimeter - it measured 3.1 ohms. This third calibration step is less critical, but it also takes very little work other than measuring a resistor with a multimeter and connecting it to the test leads so I do it.

The first time I selected the “OK” button I got an error message “The highest level in the measurement is -48.43 dBFS, for greater accuracy increase the sweep level to raise the highest input level above -40.0.” So, I increased my Output level, and ran it again. REW created a measurement called Impedance ref cal that looks like a horizontal line at 3.1 ohms. (Note, the first measurement, although it generated an error message, looked the same.)

At this point there are three measurement windows in REW for each of the three calibration steps. If I close REW and come back later, although the three windows do not come back up, the calibration results are saved. So, once I have built an impedance jig and calibrated it, I can just select the Impedance option from now on and am ready to measure the impedance of a driver. Note, however, that the output/input setting and input gains may not be correct so you should check those. For example, when I opened up REW recently, I had moved the input gain knobs by accident when I moved my setup and also my calibration file for my Motu M2 Audio Interface had not loaded. Once I got those back to where they needed to be, I was ready to measure a speaker.

B. Measuring impedance in REW

At this point, all my calibration buttons have changed from red to black and I am ready to take a measurement. The driver is set on some boxes or books (which are set on my desk) that do not impede the pole piece vent and I put a little piece of felt on the desktop under the boxes to prevent any reflection from the surface, although in this case I’m using 5-inch midwoofer so it probably didn’t make any difference. I connected the alligator leads to my speaker and selected the Start button.

Impedance is just a single free-air measurement. Note, for crossover work you need the impedance of the tweeter and the in-cabinet impedance of the midrange and/or woofer. After selecting start and obtaining this measurement, I can go to File\Export\Export measurement as text in the menu bar and save the impedance measurement as a *.txt or *.zma file. (There are some other file extensions available that I do not use.) However, to calculate T/S parameters, you will need two or three impedance measurements.

In the REW instructions it says that, at a minimum, you should calibrate the Open Circuit. Given that I have the impedance jig right in front of me, I am not sure why someone would not do the other calibrations, but I measured the following:

• A partially calibrated measurement: Impedance with only the Open Circuit calibration performed

• An imprecisely calibrated measurement: Impedance with all three calibrations performed, but using 100 ohms for Rsense

• A fully calibrated measurement: Impedance with all three calibrations performed, and Rsense correctly set to 98.83 ohms. (I purchased 5% tolerance 100-ohm resistors, but I know the true resistance of this one is 98.3 ohms because I measured it with my multimeter.)

You cannot see the difference between the last two in figure 4-L, as they are so similar; i.e., the difference between using the nominal 100 ohms and the actual 98.3 ohms didn’t matter and the difference between the results with only the Open Circuit calibration vs the full calibration is quite minor (at least in this one case).

C. Calculating Thiel-Small Parameters in REW

There are a couple of different options to calculate T/S parameters. The sealed box method – which you can read about at the REW link if you are interested. The added mass method, which is the most common method, and the dual added mass method, which you can also read about at the REW site.

For the added mass method, you run an impedance sweep with the driver in free air and another with some additional mass added. Supposedly, the best practice is to have the driver positioned with the cone facing forward as it would be in an actual speaker, which also means you must attach the added mass with some sort of adhesive like Blu-tack. I do not do this, I just set the driver on some boxes or books on a worktable. It is still the case that I put some Blu-tack or masking tape on the weights so that they do not “jump” during the sweep.

The T/S parameters can be sensitive to the added mass, and I do not know what the “true” Mms is. The manufacturer’s data sheet says 10.2g and both the REW and VituixCAD’s guidelines suggest starting with an added mass of one-half of the expected Mms. I measured the device under test (DUT) (i) in free air, (ii) with 5 grams of weight added, and (iii) with 10 grams of weight added. Some further measurements investigated and compared in Section 7.

To calculate the T/S parameters, select Tools in the menu ribbon and select Thiel-Small parameters. When the T/S Parameters window opens, select the added mass method, the free air measurement, the added mass measurement, and the amount of mass added, and finally enter the voice coil DC resistance and the effective cone area. (All of the items circled in blue in the Figure below.) I measured 3.8 ohms DC resistance with my multimeter, but the manufacturer’s Re was 3.7 ohms which would have been okay in this case. I used 96 cm2 for the Sd from the manufacturer’s data sheet, but I could have used various calculators to get this if needed. Click the Calculate Parameters button, circled in red. I have already done that in the screenshot below, and the calculated parameters are circled in green.

5. Measuring Impedance and T/S Parameters Using LIMP (ARTA)

The measurement process along with most screenshots for measuring impedance and calculating T/S parameters in LIMP are in this thread. That thread doesn’t cover all of the calibration steps, but those can be found HERE. Since those resources are pretty accurate, I will not repeat them here.

6. Calculating T/S parameters in VituixCAD

VituixCAD cannot be used to measure the impedance of a speaker. However, if provided the appropriate impedance curves, it can calculate T/S parameters. It can calculate the parameters using the same sealed, added mass, and dual added mass methods as used by REW. The mathematical formulas may differ within the software, and I did not attempt a comparison of the calculations from VituixCAD and REW. However, VituixCAD does add an “Extended Impedance Model.” I do not know the practical advantages, and again did not attempt a comparison, but those interested in pursuing alternative methods (and likely best practices) can find more HERE and HERE.

7. Comparison of Impedance, T/S Parameters, and box modelling

A. Impedance and T/S parameters

In most cases, the user will only have one impedance jig and not be able to compare various alternatives. On the other hand, the user can vary the different amounts of added mass and use both the added mass and dual added mass methods to see what the variation implies and decide on which seems to be the most reliable. One rule of thumb I found was to start with an added mass of about half of manufacturer’s Mms and increase it until Fs drops by 25%. In fact, when using DATS, it will not proceed to calculate VAS using the added mass method unless Fs drops by at least 25%. Note, if you keep increasing the added mass significantly it will make a significant difference to Vas (inaccurately), which will in turn affect the box modelling, so you do not want to keep adding mass excessively. This section concludes with a few comparisons that addresses a few prominent issues.

Does the jig matter? Even though each jig has different connections and different wiring, the calibration steps should remove these effects. Below is the free-air impedance using Jig 1, Jig 2, Jig 3, and DATS v3.

Before I used the DATS, I settled on the added masses of 5g and 10g. This reduced Fs by 24% using Jig 2. Below are the measurements of free-air impedance, impedance with added mass of 5g, and impedance with added mass of 10g, for each of the three DIY jigs. (Note: I have read that a U.S. nickel weighs about 5 grams. I bought a precision scale and measured about 20 nickels and each one weighed 4.99 grams.)

After measuring with the DIY jigs, I measured using the DATS and I had to use an added mass of 8g to avoid the error message that Fs had not decreased by 25%. In the table below, I show the T/S parameters provided in the Manufacturers data sheet; those determined by DATS using +8g and +13g added mass, and those determined using each of the three DIY jigs in REW using the added mass method with +5g, +10g, and finally using the dual added mass method relying on the +5g and +10g measurements.

A couple of comments on the variation of T/S parameters. Just quickly reviewing the table of results, the measurements appear relatively consistent, but upon closer observation some of the measurements differ by close to 40%. For a rigorous comparison, I should have used the same added masses and ensured the same voltage levels for all measurements. I measured with DATS after I had used the DIY jigs and could not proceed with +5g and had to increase the mass to +8g to measure VAS. I choose not to go back and redo all of the previous measurements. Similarly, the voltage level for Jigs 1 and 2 was left at the same level, but Jig 3 uses the power amplifier, and I did not attempt to keep the voltage level the same; and finally the DATS does not have an adjustable level so those measurements are based on whatever the default output level is. The table demonstrates that the results are sensitive to the measuring conditions and the user should attempt to follow best practices as closely as possible to get reliable parameters.

I often see forum posts where someone finds out that their driver measures significantly different from manufacturer’s specs and panics, only to find out that it did not make any difference when modelling the box. So, we next turn to the practical implications of the differences in T/S parameters in Table 7-1 above.

B. Box-modeling implications

To begin, I entered the manufacturer’s T/S parameters into WinISD and modeled a ported speaker box using the recommended Chebyshev alignment, 19 Liters tuned to 49 Hz. I then used this box volume and tuning to see how different the response would be using the T/S parameters from the other calculations in Table 7-1.

Figure 7-C shows the SPL simulation in WinISD from the manufacturer’s parameters and each of the DIY jigs with an added mass of 5g. The difference between the manufacturer’s specs and Jig 1 is 1.5db at 500Hz, but the three jigs are within 0.4db of each other.

Figure 7-D shows the SPL simulation with the T/S parameters derived when +10g was used for the added mass.

Finally, Figure 7-E shows the SPL simulation for the box models using manufacturer’s T/S parameters and those calculated using the Dual Added Mass Method in REW. The results here are considerably different than those in the two Figures above.

As I mentioned, I have not used the Dual Added Mass Method before, so there might be some better practices that I could have used for these measurements. On the other hand, these could be most accurate, and the prior ones inaccurate; after all the Dual Added Mass Method is supposed to be a more accurate method. I did not investigate further as my goal was not to determine the correct T/S parameters of this driver but to investigate the effect of the different jigs and different methods, and I think these last 3 Figures provide a few good concluding points.

• The different jigs are not driving significantly different results – if you build a DIY jig, whether something simple or something a little more complex, and calibrate it, you should get similar results.

• Jig 1 and 2’s results were nearly identical, whereas Jig 3 differed a bit. I conjecture this is because Jig 3 passes the signal through a power amplifier rather than just using the output of the Audio Interface and the stronger signal is producing more accurate results. (I did not investigate this further).

• Using different amounts of mass, as well as different methods (added mass vs dual added mass), can make a significant difference. You should research best practices as well as take measurements using a variety of alternatives so you can see the implications. You will not know the “true” T/S parameters, but you will have data to make better informed decisions on the appropriate box volume and tuning for your project.

• With any impedance jig (DATS or DIY) there is no reason to rely on manufacturer’s T/S parameters. Adopt best practices to get reliable T/S parameters and model your speaker box with REAL parameters. The Figures above likely exaggerate the effect of the differences in parameters from manufacturers specs because the box volume and tuning were constrained to be the same as the original model form manufacturers specs.  A little smaller or larger, a little lower or higher tuning, and it is often the case that you can get the desired results. [However, the worst-case scenario is that you purchase a driver based on manufacturers specs and are constrained to a certain box size. If the actual T/S parameters differ enough, it may simply be impossible to achieve the originally intended goal with that driver.]