What is the mic distance? For such a small speaker 60-70cm would be fine.
Mic stand is also likely the source of the reflections. Photo of test setup is always helpful for us to judge the potential sources of the reflections.
Mic stand is also likely the source of the reflections. Photo of test setup is always helpful for us to judge the potential sources of the reflections.
Check if you don’t have some Windows sound effects setting wrong (you would not be the first…). And do a loopback measurement from amp ->voltage divider (>40dB) -> mike input. That should give a near perfect impulse response.
@sheeple I was trying to intentionally introduce some reflections to see if it would change the two irregularities in the impulse. I think the issue is coming from the computer source or amp. I will try a different computer, different connection to the computer (headphone output vs hdmi audio), and a different amp to see if I can isolate the issue.
I believe I got it sorted out. Here in Red is the woofer near field, and in purple is the blended far/near field measurements (port not included right now..since I was only trying to chase this silly 2.5ms/4ms issue you guys pointed out for me). I seemed to have resolved it. It turns out my yamaha HT receiver was the creator of it (I tried with HDMI cable and RCA cables and it was still there). I switched to my dads old 1980's hitachi S903 and it was miraculously gone.
I am hoping I can get your help at some point once I collect the updated data to check my amateur cross over design skills.
Impulses... RED is Near field, and Green is far field. These were taken in my garage and look to be dirty impulses, so I will need to repeat outside, or at least with the garage door open. The data looks much cleaner now. The big dip in the response is the enclosure tuning frequency, per Bagby's white paper guidance, right?
I am hoping I can get your help at some point once I collect the updated data to check my amateur cross over design skills.
Impulses... RED is Near field, and Green is far field. These were taken in my garage and look to be dirty impulses, so I will need to repeat outside, or at least with the garage door open. The data looks much cleaner now. The big dip in the response is the enclosure tuning frequency, per Bagby's white paper guidance, right?
Yes, you are right, just apply the diffraction response for yor baffle and then measure FRs off axis to have power response and directivity info.
These HT things have lots of sound processing settings, you probably didn't check the one without reverb 😉. Get some straight amp for acoustic measurements.
simplicity wise
sounds like Hitachi SR-903 is good
some anomalies were removed and
process moves forward.
sounds like Hitachi SR-903 is good
some anomalies were removed and
process moves forward.
Finally... some more action. I put together the ARTA jig and collected impedence measurements of the Scanspeak and Morel (both in the enclosure). Have a look. Is it me, or does the Scanspeak peak impedence values look really high? Is thta a bad thing? I compared them to revel clone I am basing my speaker off of, and their peak was in the 35ohm range.
Do I finally have enough information to design a crossover?
Do I finally have enough information to design a crossover?
The Scanspeak has an Fs of 39-40hz allowing for variations…..yours is showing quite a bit lower indicating that there still may be a calibration issue within your rig somewhere. Did you tune the box to 70hz or is that off too?
I believe the box is tuned to around 40hz (if the previous measurements, showing the large dip, are any indication).
Here is the jig... 100ohm resistor (measures 100.0 with multimeter). I converted this RCA switcher to this jig. The RW connections aren't connected at all to the circuit board, and the green ground is soldered to the ground bus on the underside of the board.
I have taken measurements using Limp and REW (I used REWs 3 step calibration too) and the peak both occur at the same frequency. When connecting to my speaker I am using a 10' length of some Monster Cable speaker wire I had lying around (maybe 10-12awg or so).
Here is the jig... 100ohm resistor (measures 100.0 with multimeter). I converted this RCA switcher to this jig. The RW connections aren't connected at all to the circuit board, and the green ground is soldered to the ground bus on the underside of the board.
I have taken measurements using Limp and REW (I used REWs 3 step calibration too) and the peak both occur at the same frequency. When connecting to my speaker I am using a 10' length of some Monster Cable speaker wire I had lying around (maybe 10-12awg or so).
Apart from a slight mistuning (only theory…) I don’t see strange things. Did you apply damping in the enclosure already?
Yes. The enclosure has foam lining the interior walls. (Melamine sponge type. Like Mr Clean sponges).
It's a ported box so the tuning frequency is at the impedance minimum between the two peaks, in this case ~40Hz. A nearfield measurement of the woofer will show a dip at the same frequency.