Mid range horns - DIY

[Kwikas]

I've got a couple of JBL 2123h drivers that I'd like to use in mid range horns. They would slot in between a JBL 2226h L/F and a 1" HF CD/horn.

 

I could use them as direct radiators in a 3 way set up but ideally would like to get some directivity and pattern control as I intend to use them as 'wides' in my HT.....although I'm just not sure how important pattern control is for wides.

 

Anyway, I have looked around the internet and about the only thing I can find is this:

 

http://www.volvotreter.de/dl-section.htmhttp://www.volvotreter.de/dl-section.htm

 

.....it's the 10" scaled down Lascala mid horn.

 

Problem is, I'm not sure whether the 2123h is suitable for this build as I can't see any specific drivers mentioned.

 

Does anyone have any ideas......??

 

Thanks

 

 

[tuyen]

gday kwikas,

 

I can recommend also asking for ideas on diyaudio forum if you haven't already done so.  much much bigger diy crowd over there.

 

re the 2123h driver, i've only ever seen it used as direct radiator in small (10-15L) sealed box. 

 

sorry if i'm not much help.

[Kwikas]

Thanks tuyen,

 

I'll try the diyaudio forum as well.

 

Yes, you're right. Generally the 2123h is used in a small sealed enclosure but I have seen them loaded into a community horn.....well, once anyway. 

 

For a build, I would prefer a direct radiator mid for simplicity sake. However, I was also looking for comments around horn loading vs direct radiator in front wides. I can see the advantages of good pattern/direction control (and hence horns) in the L/C/R speakers but wonder whether this is so critical for wides given their closer proximity to the listening position....and since they are also blending in the surrounds to the fronts.

 

Cheers

 

 

 

 

[davewantsmoore]

I think pattern control is good for every speaker no matter what the location.

 

I've never heard of this driver in a horn ...  You'd be practically on your own.   Could be a hike.

[Paul Spencer]

It may surprise you that a lot of horns don't have real pattern control. They can actually be quite wide in their pattern like a direct radiator and they can also beam. To optimise this kind of thing you need extensive measurements, done like this:

 

 

and the outcome is a sonogram like this:

 

You can easily see that this one is beaming before it loses pattern control below 1.5k.

 

The simplest way to do this is with the 10" as a direct mid, using a horn that can match dispersion of the mid at around 1.5k. If you horn load the midrange, then you want to at least match dispersion at some point where they can hand over so that you can get some kind of smooth transition.

[Kwikas]

Okay thanks Paul.

Now it's getting a little clearer.

My wide speakers L/F (JBL 2226H) and HF CD/horn is exactly the same as those used in my L/C/R speakers....only the midrange is going to be different. My 3 way JBL L/C/R speakers all have horn loaded mid ranges and the crossover points are 350Hz and 1.5kHz.

So Paul, what you are saying is that my 'proposed' 2123h mid horn in my wides will need to match the dispersion of my HF CD/horn at the 1.5kHz crossover point.....

How do I achieve that? It sounds like a lot of work will need to be done to design a suitable mid horn/wave guide....

[davewantsmoore]

my 'proposed' 2123h mid horn in my wides will need to match the dispersion of my HF CD/horn at the 1.5kHz crossover point

 

 

Yes

 

How do I achieve that? It sounds like a lot of work will need to be done to design a suitable mid horn/wave guide....

 

Correct...   the design of the HF horn will also contribute to the overall system design.

[Paul Spencer]

What Dave said ...

[Kwikas]

Hmmmm....just as you said. Sounds like a hike!!

Okay, so what about using the 2123h as a direct radiator for mid duties rather than horn loaded? That could also be problematical since the whole dispersion thing comes into play at the crossover point too - right?

[davewantsmoore]

That could also be problematical since the whole dispersion thing comes into play at the crossover point too - right?

 

Right.   This is the basics of choosing drivers and designing a crossover.   Matching the response at the crossover point over different angles of radiation.

 

So... Regardless of what you do for the mid (in a horn or in a box) ... You will need to take measurements of the drivers.

 

We could guess at the directivity for an average 10" driver... and conclude something circa 1.5khz could work ... but this will depend on the response of your tweeter+horn.

Edited May 15, 2013 by davewantsmoore

[Paul Spencer]

If you have a horn/waveguide with constant directivity, then you will have a midrange driver that is narrowing in directivity with an increase in frequency. So you pick the point where the midrange has a beamwidth that matches the waveguide/horn. The actual point will vary depending on the driver and the horn/waveguide. Of course, you still have to consider things like cone breakup and the limits of the tweeter in question, but it's not difficult to get a 1" compression driver to meet a 10" mid. Often you can go up to a 15" mid with a crossover closer to 1k, but that does count out some compression drivers.

[svenr]

You don't have to guess the directivity of a 10" driver.  The 2123H has quite a well dampened membrane which stays close to piston operation up to the upper frequency limit.  Look in any textbook such as Beranek or Kinsler & Frey for the directivity plot of a piston radiator.  But you will have to obtain directivity data for the CD horn to determine the useful crossover point.

 

And depending on the size and design of the CD horn, 1.5kHz may actually be low.  The first step is to measure the electrical impedance of the driver when mounted to the horn (use ARTA Steps), look at the resonance frequency, and decide whether your crossover frequency will result in excessive excursion of the diaphragm.  Alternatively, you can estimate the input voltage you require to achieve your design acoustic output power, measure the distortion characteristic at that input voltage above the resonance frequency (keep transient plus steady state time low to damaging the driver), and then check for the frequency at which the distortion drops - this is the lowest possible crossover frequency at that input voltage.

[Kwikas]

svenr and Paul,

OK, it seems that the best way to do this is try the direct radiator approach first. Making a mid range horn will be a lot more complicated and as it would be my first horn attempt, so it's probably a stretch.

A 2123h as a mid in a sealed enclosure should be a lot more do-able.

Svenr, I mentioned the 1.5kHz crossover from mid horn to the HF horn (and it's incorrect...it should have read 1.2kHz) because this is what JBL do in their 3732/3632 screen arrays.

Edited May 16, 2013 by Kwikas

[svenr]

Have a look at the incredible size of the HF horns in those cinema arrays.  They will provide a very high resistive load to the HF driver and thus will allow the low crossover frequency without excessive excursion and distortion.  In comparosion, your horn will most likely be no larger than 200..300mm at most.  II haven't come across a small constant directivity horn (i.e. short length and limited mouth size as typically used in clustered sound reinforcement systems) that would allow such a low crossover frequency.  Thus my advice to find the lowest crossover frequency for your required voltage level.

 

On the other hand - you could just give it a crack and find out the difference by changing the crossover frequency.