Bass response and inductance humps

[augerpro]

I'm going to use the Dayton Esoteric ES180TiA as the woofer in a 3 way. Modeling the bass I see this is a woofer that likes small boxes and/or low tunes for flattest response compared to other typical drivers, otherwise there is a considerable hump at tuning. A 17 Lt. box @ 35hz produces a fairly flat response (blue in the box model pics). But as this will be crossed at 500hz I knew there would be a considerable inductance hump around 100hz - pretty typical. It occurred to me that I could use a bigger box (22 Lt.) @ 40hz (red in the box model pics) to offset somewhat the effects of the inductance hump. To test out the probable inductance hump I made modeled a 4th order crossover for a driver I have measurements for and is similar in impedance: SBA SB15 poly cone. Here is a pic of that with raw driver response in box, crossover applied, and crossover transfer function.



So comparing that against the box model looks like it does complement the inductance hump nicely. However, group delay is worse, and I need advice on whether the result is concerning?

[JonMarsh]

The thing you've got to ask yourself, is the cure worse than the disease, as regards the reproduction of music? You've got an underdamped filter with the red curve, with that vented alignment, with peaking at the corner frequency, and of course with attendant GD affects.

Crossing woofers low if there own box tuning isn't super low always results in these issues, with the impedance curve of the driver and the filters interacting with the box tuning. some damping in the enclosure might help, but I'd also look at re-aligning the box tuning- seems to me from what you've shown that the tuning is higher than optimal. But without all the files to play with, I'm kind of whistling in the dark here...

[augerpro]

Do you want files to play with?

So the GD is worse, but if you saw only the one result would you say it is objectively bad? Brings up a side issue where some EQ the response of a vented box to look like a sealed box and claim that the time domain performance is the same. I can't see how if you still have two sound sources that are not in teh same place and have different responses and phase. Still I've seen GD much improved from that method. So similarly if the crossover hump offsets the dip in the box response, and overall response is flattened, is the GD not also improved? I think SE might show this in the crossover CAD now that I think about it.

[augerpro]

In SE I added an active boost of 3db and Q=3 to 60hz and 40hz of the woofer circuit above. The end effect would be a flattening of the response relative to area 100-200hz. GD is worse for both. I guess my thinking was that, like phase, if response is flattened then time domain should be improved because of it. That doesn't appear to be the case. It seems to be any manipulation of the frequency response whether it is flat or rolling off is impacted from the starting response, lowering response (say with a PEQ) lowered the group delay, while boosting response worsened it. Or maybe it is not actually a change from the starting response, but how close the result is to a Bessel roll off (maximally flat GD)?

[olu78]

Hey Auger. Out of curiosity, what driver are you planning on using as your midrange? I'm working with es140ti's and ne123w 's for one design and es180ti-a's and mr13p's, as soon as they arrive. I am interested in your progress.

Sent from my SM-G935T using Tapatalk

[augerpro]

I'll be using the Tectonic Elements 46mm BMR. It's a WMTMW layout with the design goal of wide horizontal dispersion and narrowed vertical dispersion through nearly the entire passband of the speaker. Tweeter is probably going to be the Dayton AMT2-4, but I'm considering the BG Neo3 as well.