Well, I have now replaced the older drivers back in the bottom four holes of the big manifold. Since I had spent some time adjusting filters on the BFD for the experimental "sealed" boxes I had to start from scratch. It would have been more sensible to have started a new filter series but one thing led to another. A small change just to see what happens and then another...
A quick check with a 9V battery on the speaker cables, at the EP2500 amp end, confirmed that the drivers were all moving in unison. I had wired them in series-parallel again to obtain a combined load of 4 Ohms.
I started with an REW sweep of the newer drivers with no BFD. (or rather with the BFD bypassed and no LED lit on the top, right button.) No obvious problems with 1/3rd octave smoothing. I'm not going to bother trying to improve this with the BFD.
The older drivers proved not to be quite so accommodating. I added two filters. A boost at 20Hz and a -5dB cut at 40Hz to flatten the hump. That was all that was required for the result below:
Interestingly(?) there is a phase anomaly between the two sets of drivers. The new version of REW can now display phase though I omitted to record it. The effect of this anomaly is that the older drivers suck out the bottom end of the newer drivers unless their phase is reversed. I chose not to reverse the phase because of the unwanted effects higher up the frequency range.
Here is the combined effect with the drivers level matched and in parallel:
Not a bad result from just two BFD filters on the older set of drivers and leaving the new ones quite untouched. No cheating here. This is exactly as measured at the listening position at ear height.
My next objective is to better match the IB to the speakers. The images above are with a much higher crossover point setting on the CX2310 to match the different needs of the "sealed" boxes. I shall have to experiment with a lower crossover point to ensure there is no bass source localisation. Though none was noticed upon listening to a few samples of music. The crossover is probably somewhere up around 200Hz, at the moment, but there was no more time to play after dinner.
Below is shown the awful "raw" result from having no BFD but with the speakers playing as well as both sets of IB drivers:
The deep trough, centred on 180Hz, has returned with a vengeance. I have now decreased the IB level relative to the speakers but need much more time to play. It still sounds amazing compared with the previous "sealed" boxes. These were intended as mini IBs but I was never allowed to make new holes in the walls for heat loss reasons.
I have had another idea though. The ceiling below the AV stage needs to be redone. I have already boarded the other half of the ceiling in the sitting room with pine. I could use the stage area as an OB (open baffle) with the manifolds opening into the room below. The distance to the edge of the baffle would only be about five feet to the open stairwell at my feet. Still enough to be well worthwhile if deep infrasonics are not required.
The older 32Hz drivers don't do very deep bass anyway without massive boost. The best thing about this idea is no connection to the outside world. So no heat loss. The area below the stage is a transition area to the main sitting room area so a couple of ceiling grills will pass completely unnoticed.