Help Me Troubleshoot My Horn Design

[BAM]

I've got a bit of a problem. I'm trying to design this horn for the Eminence HL-10 using Hornresp 7. I'm running into a weird problem here, where the response has this really dramatic undulation. It's less apparent when simulated in half pi space, but that's not realistic for me because I want this to be a pro sound sub that I can set in the middle of the floor and get good output from.

Here's the Hornresp parameters window


Here's the output simulated in 2 pi space:


Here's the output simulated in half pi space:


This looks like a symptom of a problem that can be easily diagnosed, like a rear chamber that's too large, or a horn expansion that doesn't occur at exactly the same rate in the middle as at the beginning and end. If you could help me figure out what is going wrong with my horn design that is causing the undulating response, I'd really appreciate this.

[mike.e]

low low low throat area for this driver!

YOur not gona get good output in 2pi space with a horn of thise size. Almost all horns are corner horns ie 1/8th horns and ul need 2 against a wall or 4 in open space

Or you could make one HUGe horn thats not portable but works in 2pi !

Youl have to have a higher F3..


Use the Lab12! If your gona use a couple of hl10s why not use lab12,it models well even on a 50hz cutoff horn

Bill has little horns for this driver
Also your Vtc is unrealistic- try 3000cc or similar.
Also you can use the HYPEX tool and make the horn smaller.

Typically HL10
St 100cm^2 Mouth=2400cm^2 expo 2.8m Vrc 16L Vtc 2000cc

Cheers!

[Centauri]

I haven't been successful modelling this driver either. The speaker's Fs of 32Hz would appear to be one of the main problems - seems to work better in a 20Hz horn and in multiples of 4 or 6 (which you would have to do for 2pi loading anyway).

There was supposed to be plans for a horn box for this driver on the Eminence web site some time, but it hasn't appeared yet. I have seen the plots and they are not too bad, but have no idea on the input parameters to achieve this.

Cheers

[BAM]

In Hornresp, the VTC and VRC are in liters, or 1000s of cubic centimeters. So it's actually 31,000 cubic centimeters. I wonder if this is why it doesn't seem to be working. Should a folded horn have a larger rear chamber or a larger compression chamber?

On the ProSoundWeb, there are people showing graphs of HL-10 horns that have very smooth responses, not the junky one I have. I wanted to design something where I could get away with only two per side.

[BillFitzmaurice]

Your rear chamber and throat chamber sizes are both way off; play with them. My HornResp has VRC in liters but VTC in CC- make sure what your's is. The predicted undulations probably won't actually be there in the final product. To operate with any SPL below Fc you need adequate chamber volumes to support direct radiator mode below Fc. Your Fc is as good as you're going to get from this size box without going to a smaller driver.

The only graphs I've seen with very flat predictions are Cowan's, and he has a history of playing fast and loose with his results; everything is in eighth space, always predicted. He once tried to pass off a chart in 1/16 space. I've never seen him publish an actual measured half-space anechoic response. I've never published a predicted response, only actual measured, as that's ultimately the only chart that counts.