The measurements shown here were made using the Just MLS plug-in from LSPCad standard 5.25 along with the Sound Blaster Live! sound card in my lab PC.
The sound card feeds a Sansui AU-7700 integrated amp to drive the test speaker. I use a switch box to provide the different connections necessary for doing the frequency response and impedance measurements. The switch box I'm using is based on the "Wallen jig" from when I was using Loudspeaker Workshop. Some of the values have been changed to fit my needs. I'll get a schematic of the modified jig drawn up sometime.
The microphone is a homebrew using one of the Panasonic electret capsules available from Digikey. The microphone preamp is based on the TLO84 dual J-fet input op-amp. It is setup with the first stage having 20dB gain feeding the second op-amp. The second op-amp has switchable gain of 20dB or 40dB. This gives the microphone preamp 40dB or 60 dB gain. The preamp is powered by a pair of 9V batteries. The microphone is not calibrated yet so don't take the responses and sensitivity as gospel.
Soap box time!! Be very careful if you are trying to do measurements using your soundcard! Most sound cards have minimal if any input protection. What this means is if you overdrive the input of your sound card it rapidly becomes dead... For protection on the inputs what I like using is a standard red LED in series with a 1N4448 (1N4148). Build 2 of these combos and arrange them in parallel pointing opposite directions. This makes a clipper that has a limit voltage around 2.2 volts peak. With the LED's in the circuit you get an indicator that shows when you are overdriving the inputs. Here is a schematic for a protection clipper.
The basic test setup is a large infinite baffle (wall in my basement) with a 2' x 2' cutout in it. I have made several test baffles that fit in the hole so that I can test different size drivers. For the IB measurements I have the gating on Just MLS set for 5ms. This seems to be a good compromise of Lf response Vs reflections off the floor and ceiling. Below is a crude picture of the test IB.
The first set of drivers are 4 Radio Shack 40-1354a drivers. I came across an interesting bit of data making these measurements. There were 2 distinct different frequency responses. One pair rolls off at 10K. The other pair's frequency response is extended out to almost 20K with the exception of a really narow notch at 18-19Khz. After looking at the results and re-testing several times to make sure that I was really seing this I inspected the drivers. The extended frequency response drivers were made in Taiwan, the drivers that roll off at 10K are made in Thailand. Drivers #1 and #4 are made in Taiwan, Drivers #2 and #3 are made in Thailand.
Next set of drivers are RS 40-1285d 6" widerange speakers. The frequency response of these seems to be influenced by the country of manufacture also.
Electrovoice SP12B drivers.
Electrovoice SP12C drivers.
Radio Shack's version of the JVC ribbon tweeter from many years ago. This also was a test for the microphone.