Hello all!
My name is Phill Graham, and I used to be active around these parts when it was still the basslist email listserv, and hosted by Dylan at UTAustin.
Seems it has grown somewhat!
Since that time I went to college, worked as speaker design consultant during college, and have been in graduate school for a number of years. I am pretty much only involved with school (and pro audio) these days, but I am looking to build a personal project right now.
I would like to get started on the boxes, but I need some information to do so. The box is an mtm, and will use two of the Dayton RS-180s mids, and the tweeter the Peerless HDS 810921. The baffle will be \_/ shaped, with the woofers angled out to align the acoustic centers of the tweeter and mids.
Has someone determined the acoustic center of these drivers by the Hilbert transform, or some other method? I need to give the geometry to my friend (and woodworker for this project) so we can get the boxes built, and then work on the XO designs.
Thanks for any help!
Anyways,
My name is Phill Graham, and I used to be active around these parts when it was still the basslist email listserv, and hosted by Dylan at UTAustin.
Seems it has grown somewhat!
Since that time I went to college, worked as speaker design consultant during college, and have been in graduate school for a number of years. I am pretty much only involved with school (and pro audio) these days, but I am looking to build a personal project right now.
I would like to get started on the boxes, but I need some information to do so. The box is an mtm, and will use two of the Dayton RS-180s mids, and the tweeter the Peerless HDS 810921. The baffle will be \_/ shaped, with the woofers angled out to align the acoustic centers of the tweeter and mids.
Has someone determined the acoustic center of these drivers by the Hilbert transform, or some other method? I need to give the geometry to my friend (and woodworker for this project) so we can get the boxes built, and then work on the XO designs.
Thanks for any help!
Anyways,
More details...
Hello,
If I am going to have to do the process myself with a FFT analyzer, then simply looking at the time differential between the predominant energy spikes of the drivers' individual impulse responses does NOT give the full picture of the group delay behavior of the driver, but it will usually be close. It will give you good idea of the driver's acoustic phase in bandpass where the bulk of the arrival energy arrives, which may or may not be the frequency range you desire to measure this value in.
If I was doing the measurement using SMAART (a dual channel FFT analyzer used in pro sound), I would begin by setting the delay between the reference and measurement channels by looking at the "spike" of the arrival time of the impulse response, or ETC.
Then I would switch to bode plot mode, looking at the behavior of the phase trace in the region near the desired crossover frequency. Then I manually tweak the reference delay time so the slope of the phase trace is zero in the region of the desired crossover frequency.
Since the slope of the phase has been forced to zero in that range, then by definition the group delay is zero in that range, and the reference to measurement delay can be taken to be the true group delay in the desired frequency range. At that point the physical distance needs to be subtracted, and the remainder is the group delay to the driver's acoustic center in the range you desire.
Another way to do this is to calculate the phase of the speaker you are measuring with the Hilbert transform, and compare that to the phase you calculate from a FFT. You tweak the delay time till the FFT lines up on the calculated phase. Its the same basic idea as above. The MLS analyzers I have seen work this way.
So, in summary, I know how to do the measurement I am still hoping that someone has already done them, and can give me the results, so that cabinet can be worked on before the drivers arrive. My partner in crime lives in a different state, so it's a little more complicated than meeting on the weekend to do the measurements.
Thanks again!
youyoung21147 said:
Hello,
If I am going to have to do the process myself with a FFT analyzer, then simply looking at the time differential between the predominant energy spikes of the drivers' individual impulse responses does NOT give the full picture of the group delay behavior of the driver, but it will usually be close. It will give you good idea of the driver's acoustic phase in bandpass where the bulk of the arrival energy arrives, which may or may not be the frequency range you desire to measure this value in.
If I was doing the measurement using SMAART (a dual channel FFT analyzer used in pro sound), I would begin by setting the delay between the reference and measurement channels by looking at the "spike" of the arrival time of the impulse response, or ETC.
Then I would switch to bode plot mode, looking at the behavior of the phase trace in the region near the desired crossover frequency. Then I manually tweak the reference delay time so the slope of the phase trace is zero in the region of the desired crossover frequency.
Since the slope of the phase has been forced to zero in that range, then by definition the group delay is zero in that range, and the reference to measurement delay can be taken to be the true group delay in the desired frequency range. At that point the physical distance needs to be subtracted, and the remainder is the group delay to the driver's acoustic center in the range you desire.
Another way to do this is to calculate the phase of the speaker you are measuring with the Hilbert transform, and compare that to the phase you calculate from a FFT. You tweak the delay time till the FFT lines up on the calculated phase. Its the same basic idea as above. The MLS analyzers I have seen work this way.
So, in summary, I know how to do the measurement
I am still hoping that someone has already done them, and can give me the results, so that cabinet can be worked on before the drivers arrive. My partner in crime lives in a different state, so it's a little more complicated than meeting on the weekend to do the measurements.Thanks again!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.