FLH design Sensitivity compared to Quake, Psycho .

Out of curiosity, I also modelled a B&C 18DS115. I was able to achieve a roughly 2db sensitivity improvement from 35 to 80hz, but there are a few caveats:

The idea is effectively to have two horns, the main one from the front of the driver as normal, and a smaller separate horn originating from the back of the chamber.

Throwing some quick numbers together I was able to come up with http://https://imgur.com/a/HvXqngO - this. Black line = CH, grey line is my previous FLH design. Sensitivity is 1w/1m in half space (2Pi). This is using the same B&C 18DS115 driver. You can see f0 is about 4hz higher (37 vs 33), however sensitivity is vastly improved ~4dB, and this enclosure is around 110L smaller (670L vs 780L); so the f0 could likely be dropped to match in a comparably sized enclosure. The compression ratio is also a much more reasonable 3.13:1, versus 5:1 in the FLH design.

One caveat is you can see there is a sharp dip at 70hz, I assume this is some form of cancellation from the two horn paths. There is also a bit more of a "sag" between 43-50hz than I'd like (most dnb basslines sit around 43hz), however this does smooth out http://https://imgur.com/a/BGxIm4q - once you scale up to 4 boxes. In a stack of 4 the f0 point also drops to around 35hz with f3 around 31. 

The other thing is I'm not at all sure how this type of enclosure sounds. The main reason I am a big fan of sealed FLH (and less of a fan of TH) is they have a very "tight" but aggressive sound (I assume this stems from the fact that the driver sits in a sealed chamber). On the other hand, I find TH to have a less punchy/more "loose" sound -- I'm not sure if this compound horn design would suffer the same issue. With that said, power handling can potentially be increased, as there would be some airflow into the driver chamber - sealed FLH tent to get preeeeety toasty. 

Does anyone know of any designs based around this CH principle? I saw some early stuff from the paraflex guys, but nothing concrete, and it seems those sketches don't use a flared horn.

I dropped a little the B&C 18DS115 since the max SPL is anyway limited by the excursion. Because of large excursions, I decided to continue with the BMS 18N862 but also the RCF 18N405 as in the HESSBH seems to work nicely with almost no modifications. As psychotea wrote, I don't really see the described low frequency enhancement from the HESSBH design in hornresp. Maybe I am not getting it, but anyway, with the classical design I can see 30Hz now. I want to stay below 700l in volume. Following just a documentation of the recent hornresp output. 

The max SPL is wideband above 150dB (1xpi, 4 speakers) and is higher compared to the older PD 1850 mainly due to higher excursion as seems to be possible with the BMS driver. 

Please correct me if I am wrong. 

 

Why Sims in 1 x Pi Do you have a massive wall that you will be using directly behind the stack? If just on the ground 2 X Pi then you are simulating the equivalent of 8 cabinets which is helping the LF extension somewhat with a FLH design.

The red line on your Sim, being a displacement limit I guess, suggests a lack of a High Pass filter?

Hopefully this post doesn't violate the forum rules regarding commercial designs. If so I'd be happy to remove it. This is purely for hypothetic research purposes with regards to designing something new & improved. To be very clear, all the information is gleamed from freely available sources online, combined with a bit of brainpower to extract some data.

I came across a rough sketch online which outlines the folding of a Quake. Using this alongside the measurements EM Acoustics provide, I was able to trace over this in SketchUp and create what I think is quite an accurate representation of the horn path:

(My sketch is actually about 40mm too wide, but I realised this too late)

I unfolded the horn into a straight line, and was able to make it accurate to within <1sqcm:

I split the first segment of the horn into two parts, a 11.6L throat, and a 31.39cm horn segment. I'm not sure exactly the best way to model this in Hornresp, since the throat still appears to be flared (as if its part of the horn), but with the driver entering on the side, rather than at S0 which Hornresp assumes. I don't know if there's an option to specify the driver entering between S1 & S2.

Hornresp inputs:

Using PD1850/2, the box sims really well:

As a single unit we see an f3 point of 35hz, which matches EM's claimed specs.

The response in a stack of 4x also looks great, with f3 around 33hz. The thing I don't get is EM claims 27hz here, which just doesn't add up. The 2-3hz drop is about what I'd expect from horn loading, so I don't understand how they are claiming 8hz lower. Their frequency response graph is also much flatter down to 30hz, with a slight bump at 30 (https://emacoustics.co.uk/wp-content/uploads/2021/12/MSE-118_gb_2015.pdf)

If anyone knows what they might be doing here to get this result, I'm all ears. My first guess would be the driver/throat/S1 & S2 entry sim.

Just for fun, 4 boxes @ 80v (800w). A nice 140dB @ 40hz: