horn loading midrange

[JasonP]

So I was thinking about horn loading a midrange driver (5" or so) to get a boost of maybe 6-10 db in the 1khz and up range. What kind of profile and size would I be looking at to do that? I'd use hornresp to model it and play, but since I can't load the .hlp file (windows tells me its corrupted) its like reading some unknown foreign language.

[winslow]

How high/low are you intending to play this speaker? What's your coverage angles?

You need to see if your speaker can play as high as you intend on playing it...it might have a low mass roll off and not be able to make it to how high you want to play it.

[JasonP]

My thought process is like so: I am going to take the Zaph Audio driver and horn load it from 1khz (or so) and up to increase efficiency and quite possibly lower distortion numbers, which rise from 1 to 3 khz. The plan is to then cross the driver at 3 khz to a waveguide or horn tweeter. I don't think the playing high will be the issue, I just wanted to predict the rise in the response so I don't have to create a bunch of horns and test them out to see.

[winslow]

You need to check the mass roll off.

2Fs/Qes= upper mass roll off frequency.

65.8*65.8/.56= 7731 hertz.

Should be good in that regard.

http://volvotreter.de/downloads/Edgar-Midrange-Horn.pdf

[winslow]

Learn akabak...hornresp isn't the most accurate thing that high up in frequency.

[billfitzmaurice]

Learn akabak...hornresp isn't the most accurate thing that high up in frequency.

Neither is AkAbak, because a great deal of what goes on that high up involves the horn geometry, not the flare, and fine details like phase plugs.

You need to check the mass roll off.
2Fs/Qes= upper mass roll off frequency.

Don Keele missed the mark with that one. If true the roll off here would be 232Hz (you didn't double the the Fs, you squared it) and virtually no cone driver would be any good above 800Hz.

I just wanted to predict the rise in the response so I don't have to create a bunch of horns and test them out to see.

For the reasons above you will have to do just that, unless you get really lucky with the first effort, and IME that's not likely.

[winslow]

I did square it...opps.

[billfitzmaurice]

I did square it...opps.

To make it easier you can use 2 x EBP, but the result is still worthless. I don't know what possessed Don to come up with the Fhm calc, probably the mad rush forty years ago to apply T/S specs to anything and everything. It didn't help when designers like Bruce accepted it as gospel without question.