Hello ported masters!

I'm trying to build a kick cab for our small cargobike soundsystem! it's my first time designing a cabinet so i have plenty of doubts and would like some feedback on the project before actually going to get the wood.

For now we have a Cubo 15 as sub and a small but very loud top. Since it's all supposed to go on a bike weight and size are best kept small.

i thought about a 6th order bandpass, trying to get the 2 chambers tuned at 100 and 250hz which is the target frequency range of the cab. bouncing back and forth between cad and hornresp i ended up on this (measures in mm):

made it 62cm wide so it would fit perfectly on top of the sub, angled the speaker baffle to save a few cm in height (so the bike driver can see the road!). the driver is an eminence delta 10a.

Simulating it on hornresp gives this result:

The frequency response looks perfect for what we're looking for but maybe there are some beginner mistakes i'm overlooking?

This is it!

does this mean that effectively 100hz are gonna "come out" of the speaker 10ms later?

oh i see from the guide to hornresp post:

A rule of thumb coming from Hifi is: frequency x groupdelay = 400 (max). Less being preferable. At 40 Hz that would be 10 miliseconds. A little bit higher (up to 600) around the tuning frequency can be overcome.

From this rule of thumb you can clearly see that the highed of the groupdelay becomes more important at higher frequencies and less important at lower frequencies. This is because the sensitivity of your ears increases as the frequency rises.

so in this situation freq*groupdelay > 1000 so basically i'd be just getting mud at 100 (which is exactly what we're not looking to) goddammit!

Is there any way around this? maybe a 4th or 8th order bandpass would make the situation better? or should i give up and just make a simple horn?

I would be looking at 6th order BP horns instead of standard 6th BP. The horn increases the radiating surface area (increasing efficiency) whereas standard BP reduces radiating surface area. Alternatively you could go for a simple and small (yet more costly) reflex cabinet with multiple drivers. Personally I would look at modifying a Cubo Kick 15 or Cubo Kick 12, reducing width to fit the T/S-parameters of the 10" used, or start designing your own from one of those as a starting point. That can give + 6 dB per driver used compared to reflex.

For your own design I would focus more on the shape of the frequency response, smoother response for smoother sound, and frequency response becomes smoother if ANG is lowered. For 12V PA it might be near 2,0 Pi for 100 Hz, yet nearer 4,0 Pi for 250 Hz, depending on the exact placement near the ground. Another way of smoothing is reducing band width, yet I would keep at least one octave of band width, to prevent it becoming 'one notish'.

You can also address delay difference between cabinets from a processing point of view (up to a point). In a nutshell, your tops need to be delayed the most, a little less delay added to your kick and none to your subs.

Yes, it seems from that graph that this box will boooooom at 100Hz. The effect is not like a straight time delay, more of a "smearing" of transients and increase in decay time. Or as you put it, mud!

I see your need to have the driver on an angle to keep the box height down. You could try a very short BP4-ish horn? In which case you could have the driver mounted horizontally in a sealed chamber, firing up or down into the horn (port) throat. The horn/port exit is on the front. This will probably simulate a lot like a sealed box but with a little added output at the top of its passband, and you will gain a little directivity from the horn, so it'll be louder in front than a simple sealed design. Group delay should be lower.

As well as GD being overall as low as possible, you also want to avoid significant *changes* in GD over the passband. EG a box with a flat-ish line at about 7-10ms over the passband, is likely to sound better than one with a 5ms baseline but 15ms sharp peaks, or one with 20ms at the lower end and a steep slope to 5ms at the upper end. Constant GD helps a lot with intelligibility and a "dry" or "controlled" sound.