I'm not really feeling the love in this thread
If it wasn't for JimmyB I'd be quite depressed. Maybe people don't understand the significance of what I've done? Or maybe it's the thread title (does sound a bit cocky, although that wasn't my intention). Or maybe I come across as the village idiot, who can't possibly know anything about loudspeaker design because I'm dyslexic, useless with grammar and punctuation etc?
When I modified my speaker cabinet and achieved the frequency response that I did, I was really pleased with myself (To be honest I was surprised at the lack of feedback, as I was expecting to hear people say, nice one Matt, you nailed that etc, but apart from Jimmy, I was met with silence
). I've Googled "open baffle" and "hybrid open baffle" etc, and I couldn't see anything out there that looks like what I did. The 2" x 8" slots weren't just a random thing, I realised there was a problem centred around 1khz, and where there should have been a peak due to baffle diffraction, there was a dip. It could have ONLY been caused by something inside the cabinet (or been inherent in the driver, but a quick and dirty open baffle experiment proved that wrong), and due to the size of the wavelengths, had to be near the baffle edges (hence the slots near the baffle edges).
Right, I'm going to ask the questions that I hoped others would ask - Why do I want a peak at 1khz? Two reasons, One, That's the presence region. Look at the Falcon LS3/5a frequency response (
here). A peak here makes voices sound alive, a dip here is my worst nightmare (dull, undefined midrange).... Two, to get good phase with a midrange bandpass filter, it needs to look like this target curve below (grey line). As you can see, the frequency response of the Faital 6RS140 in the 8.5" wide cabinet (black line) has a nasty dip around 1khz and doesn't follow the target curve in LspCAD (you can also see the breakup problems at 5 and 7khz). I don't know why I said I needed two notch filters to reduce the breakup (post #52), I only actually needed two parts, but i still think it's better to fix the problem at source, rather than try to hammer the problem into shape with extra parts (hence the Scanspeak style slits to the driver).
So what do the 2" x 8" cabinet slots actually achieve?
The red line is open baffle, but with no slots. The blue line below is the same, but with the 2" x 8" slots (both have the rear filled with wool) (EDIT: not sure if Giffs work, so I'm uploading the Blue line separately) (EDIT, EDIT: NO GIFF DON'T WORK)
As you can see, the slots fill in the 1khz dip and flatten the frequency response below 600hz, but not only that, I get all the benefits of an open baffle speaker, but without the drawbacks. If you've ever tried to fit a door in a very small space (cupboard door for example) with no ventilation, you will struggle to close the door because the air pressure is so high (you really have to push the door hard). Imagine a speaker in a small, sealed enclosure. I'm convinced this is why most sealed speakers sound dead at low volumes. They are fighting the air pressure inside the box. I once built a sealed 3-way - and as an experiment I fitted an exhaust vent (not a port). It sounded so much more alive and free with the vent.
I don't believe that the benefits of an open baffle speaker are due to the sound bouncing around all over the place causing dips and peaks due to sounds being in and out of phase. I believe it's because the motor (magnet power) isn't fighting against the air pressure inside the enclosure. To make things worse, most speakers designed for sealed enclosures have a high Qes (weak motor).