Midhorn for JBL2123?

[electroaudio]

I have played around with hornresp trying to make a good midhorn for jbl2123, and i always fail miserably.Well i do get much better results than hornresps own wizard..

So i wonder, is it possible to use that driver in a horn at all?
If so.. HOW??
(250-1750Hz)

Or are there something wrong with hornresp?
The reason i start to suspect hornresp is also the fact that my new basshornplans was too nice...


But while am at it..
Has anyone built a basshorn that has a TL to fill in the lower frequencies, and what considerations should be taken if doing so?

[Magnetar]

Horn response does not indicate the HF limits, and only approximates the low and mid response. In a proper sized horn/throat/back chamber the 2123 can cover the 100 to 1k range.

I would not use a TL or any direct radiator with a horn loaded 2123. It simply won't be able to keep up.

[electroaudio]

Originally posted by Magnetar
Horn response does not indicate the HF limits, and only approximates the low and mid response. In a proper sized horn/throat/back chamber the 2123 can cover the 100 to 1k range.

I would not use a TL or any direct radiator with a horn loaded 2123. It simply won't be able to keep up.

So, you think that its okay to use 2123 in a horn and just ignore hornresps tagged responsepredictions?
Also, i want a tractrixhorn similar to the basshorn but the program just doesnt accept that contour with 2123, the horn always ends up too short or some other errors.
Would it be okay to make a scaled model of the basshorn and just build it?


The picture shows a different horn, with a 15 inch driver that should be below the 2123 midrange horn, and it is that horn i was wondering if TL is an approach suitable to fill in at the subbass region.

(It doesnt matter if TL will give a lower response, a shelf in the graph, as long as there arent any inherent phase problems and such stuff by mixing TL and Horn, its just that i guess TL is a much better alternative than those bassreflex constructions in for example the classic W horn)

[Magnetar]

---So, you think that its okay to use 2123 in a horn and just ignore hornresps tagged responsepredictions?
Also, i want a tractrixhorn similar to the basshorn but the program just doesnt accept that contour with 2123, the horn always ends up too short or some other errors.
Would it be okay to make a scaled model of the basshorn and just build it?

A full scale model will be fine. I have no problem modeling a tractrix midhorn that will go down to 250 cycles like you mention. I use one with the 2123 that models down to 200 cycles and rolls off at 450 cycles but in the real world it is -3 db at 160 and 1K. You must pay particular attention to the back chamber. Around one liter works best.

---The picture shows a different horn, with a 15 inch driver that should be below the 2123 midrange horn, and it is that horn i was wondering if TL is an approach suitable to fill in at the subbass region.

It appears to be fine -- what is the sensitivity and how will you allign it with the low midrange horn? You will need a much bigger horn to do the sub bass region though. You modelled a basshorn, not a subhorn.

--(It doesnt matter if TL will give a lower response, a shelf in the graph, as long as there arent any inherent phase problems and such stuff by mixing TL and Horn, its just that i guess TL is a much better alternative than those bassreflex constructions in for example the classic W horn) [/B][/QUOTE]

Why would a TL be better? It will have higher distortion, lower sensitivity and probably more phase problems. Mixing transmission line bass with an otherwise horn loaded system is like putting a Chevette motor in a top fuel dragster.

[electroaudio]

A full scale model will be fine. I have no problem modeling a tractrix midhorn that will go down to 250 cycles like you mention. I use one with the 2123 that models down to 200 cycles and rolls off at 450 cycles but in the real world it is -3 db at 160 and 1K. You must pay particular attention to the back chamber. Around one liter works best.

The problem is that i dont want any frontchamber and i also want the 2123 to be inserted from the front, which gives a rather big throat, that is the part that gives trouble with hornresp.
The 2123 will work together with a (modified) JBL 2460 1" horn and i dont want to cut that lower than 1750Hz either.
Well, actually i do need more of a waveguide than a horn just to limit the disperson of the 2123 to 90 degrees.
Maybe i should drop the tractrixpart and go for a conical horn instead?
Or what will happen if the horn is too short compared to its width? I have a calculation for a 18cm long tractrixhorn with the right dimensions on both ends, which hornresp refuses to take.
Lot of questions here.. Im not sure which way to go yet either.

It appears to be fine -- what is the sensitivity and how will you allign it with the low midrange horn? You will need a much bigger horn to do the sub bass region though. You modelled a basshorn, not a subhorn.

Yes i know, but thats a tradeoff since i cant go bigger than 120x60cm = 7200cm2
the sensitivity was the graph in the picture, i think.. db/1watt
(Also, will 120x60 really be the response modelled in hornresp?)

Why would a TL be better? It will have higher distortion, lower sensitivity and probably more phase problems. Mixing transmission line bass with an otherwise horn loaded system is like putting a Chevette motor in a top fuel dragster.

Isnt it better to make a delay on the bassreflex then to have it in opposite phase like an ordinary bassreflexport would be? or??
Anyway, i usually cut everything on the eq below 60hz so it isnt that necessary to reproduce those frequencies.

[electroaudio]

I have no problem modeling a tractrix midhorn that will go down to 250 cycles like you mention. I use one with the 2123 that models down to 200 cycles and rolls off at 450 cycles but in the real world it is -3 db at 160 and 1K. You must pay particular attention to the back chamber. Around one liter works best.

I forgot to tell you that i tried the "one litre solution" and i didnt like the 10db rolloff at 450 and 800 hz, i think i stay with >50litres, it gave a much smoother response over the whole range.

But the last horn in the box was built without any calculations att all, i didnt know how to calculate stuff in those days but it has proven to give a very good freqresponse over these 15years.
But as i am rebuilding the bassbins, which wasnt that good, so i thought why not rebuild the midhorns too?

So, with some scepsis against relying on theoretical models, what will happen if the horn is too short compared to its with?
Like it would be with that 18cm long tractrixflare.
(18cm seems to be long enough within its intended freqrange, though. 1/8 of 250hz= 17cm and the last piece of that flare is 90 degrees, or more of a baffle than a horn)

Also, what is the best way to simulate a 120x60cm horn as it was round, is it with 60x120=7200cm2 or to calculate on the maximum resonant dsitance aviable; radius=60cm