1.4" or 2" throat large constant directivity horns you can actually buy!

Member
Joined 2007
Paid Member
It's probably easiest if I start this off by explaining what I am aiming for. The speaker I'm designing is going to cover the range 80 - 18 kHz and ideally have a directivity of 90x40 (HxV). At the low end of the range I'm am willing to accept the speaker becoming a mono pole but I would like the directivity to extend as low as possible in frequency, however I would also like broad horizontal directivty even at low frequency. The speaker also needs to be able to output 125dB/1M peak AES SPL. The speaker is to be used at home where I will be listening 17 degrees off axis vertically (hence a speaker with nominal 50 degree directivity would be preferred over 40 degree) due to positioning constraints and also as a PA top where it will not have to array. I will be running different crossover settings depending on application, ideally I would like to get down to ~600 Hz if possible.

What I'm having trouble with is the lack of good quality data on horn/waveguide combinations. I'm aware of the following issues:
1. Large diaphragms only having high frequency output by operating beyond the pistonic mode but manufacturer data is very smoothed and as such I can't see if this is well damped or not.
2. Often I read about some wonder wave guide from a few years ago but the manufacturer doesn't exist anymore or has moved out of the market. Or the manufacturer dosen't sell to the public. (E.G the K402 waveguide looked very interesting).
3. The expected exit angle isn't listed for the waveguides and most manufacturers are not giving the compresion driver exit angle so matching a waveguide with a driver seems to involve expensive trial and error.
4. Many waveguides are not constant directivity at all and yet are listed as such by the manufacturer! (I.E I can see from the shape that its exponential)
5. The high frequency dispersion of a waveguide can be limited by the dispersion of the compression driver but no data is supplied on this.
6. Vast differences in price between various options with no clear data on why.

So taking on this difficult landscape I have come up with a few combinations and I wonder what you think of them and if you know of any more information I have missed:
1. P-AUDIO PH-4525 This is the largest P-Audio wave guide and is only £25 and has a mouth width of 425 mm with a throat of 2". I have read that a lot of P-Audio products are clones of JBL, if so does anyone know what model this would be? In terms of driver pairings they measure it with this P-Audio BM-D750 (£85) or there is the huge P-Audio BM-D950 (not sure where to buy) a 4" diaphragm seems to be pushing it in terms of possible breakup problems! P-Audio claim it can go lower than the D750 though...

There are some measurements of the D750 combo here seems that at least one person likes it.

2. The same horn but assuming the dispersion of these large diaphragm drivers is poor using the very expensive (for me) BMS 4592 ND (£430). This is a coaxial and carry's a number of risks that make me uncomfortable. Firstly I'm not sure if it will work well with this horn and secondly if using it I would like to have a passive crossover between the 1" and 2" driver which will add a lot of development time. Ideally as this driver can go super low (300Hz) I would like to mate it with a much larger horn but I don't know where to get one for sane cost. I would be willing to try and extend the mouth of an existing flare but building the throat area I'm not very confident I could do well. This is also more money than I would like to spend considering I will also need to add woofers, electronics, box etc.

3. An in-between driver not as expensive as the BMS but with better data than the P-Audio such as the HF200. This seems to be quite well damped as the frequency response isn't as smoothed (£187).

3. 1.4" exit drivers. As my SPL requirements aren't crazy this may actually make a lot of sense, as the horns will be smaller and dispersion better? and result in lower cost. For this the RCF HF950 looked quite attractive (especially as it has 50 degree vertical dispersion and claims 'perfect' constant directivity) (£55). For a matching driver the RCF ND840 ooked quite interesting and as its from the same manufacturer it may match up with the horn well? (£145)

*Yes I am aware that I should build a synergy horn type thing and I will but this will take a while and in the meantime I need better speakers...
 
Regarding issue number 3; I think that in a one off situation, trying to match the exit angle to the horn would be very low on my list of concerns. Very low, like last on the list. If you had two horns that met every criteria, and one matched to driver’s exit angle and one didn’t, then I would go with the one that matched. Otherwise, I would not even ponder it.
 
Your speaker goals match the Danley Synergy klones in this forum. Several wood Synergy builds have been documented.

The Klipsch K402 waveguide would be ideal if you can find one on eBay UK.
Have you tried to buy an SEOS24 or SEOS30 from Poland?

I have been impressed with the BMS 4594Nd coaxial CD on a SEOS24 waveguide, and this combination would simplify the required wood work.
 
See Heritage Klipsch - that big "theater" sound and "HOM" ? - Technical/Modifications - The Klipsch Audio Community

If you're wanting a cheap approximation of the K-510 to try out (with suitable EQ applied to tame the 1-4 kHz hump in response, and a little HF boosting ramp filter on the 5-16+ kHz...just like the K-510 horn needs), here is a very good horn that's also very inexpensive and that tests very well in terms of its polars relative to the K-510: 

2" Throat Horn Bolt-On 18"x10"For Assorted Bolt On 2"Exit Drivers 90degx 40deg(398) | eBay

First horizontal:
1545964018_NewHornK-69-Ahorizontalpolarsonogram.jpg.78481fc917cc15bd4a090e879f99f9e2.jpg


Then vertical:
411800026_NewHornK-69-Averticalpolarsonogram.jpg.6f879c0b3dcf916403d58388e00bc670.jpg


The polars of the BMS 4592ND will be the even better up high (10-20 kHz) due to the absence of diaphragm breakup that the Klipsch K-69-A exhibits in the plots linked just above.

The K-402 used with a BMS 4592ND is superb. Here is an on-axis SPL/phase plot of the MEH version of that horn with dual 15" Crites cast frame woofers. Pay particular attention to the difference between the minimum phase trace and the total phase trace. (A hint: it's the best I've seen in a full-range horn.)

2080353062_K-402-MEH(onaxismid-wall)frequencyresponseandphasewithminphase.jpg.62d5d8add00e49ed06d45474dc2fcd86.jpg


Chris
 
Member
Joined 2007
Paid Member
Re: solution 3. The BMS 4594 would be a great solution for that RCF 1.4” horn, since you already expressed interest in a BMS coaxial.

Apart from the cost there seems to be few downsides to this setup. The driver seems to be basically the same as the 2" throat version? Do you know why this driver can play so much lower than most compression drivers?

Take a look at Faital Pro and Eighteen sound. You might have to toe the bov in a bit.
Looking at the 18sound horns I don't see anything that goes very low? Constant Directivity High Frequency Horns

Regarding issue number 3; I think that in a one off situation, trying to match the exit angle to the horn would be very low on my list of concerns. Very low, like last on the list. If you had two horns that met every criteria, and one matched to driver’s exit angle and one didn’t, then I would go with the one that matched. Otherwise, I would not even ponder it.
So this has only a minor effect on the systems response? is the effect of the mismatch to alter the on axis response differently to the off axis?

Regarding solution 3, SPL has nothing to do with driver exit size. EG, the BMS 1.4” driver will have the same SPL as the BMS 2” with the same characteristics, if the horn matches up the same.
Doesn't the ratio of the exit to the surface area of the diaphragm define the compression ratio? so a larger diaphragm is usually going to have a larger throat and when running low like I would like to the SPL is going to be excursion limited so drivers that are meeting my requirements are more likely to have a larger throat.
 
diyAudio Moderator
Joined 2008
Paid Member
Matching the driver to the throat is going to ensure the wavefront remains intact rather than diverting to higher order modes. The size of the throat in comparison to the wavelengths can be considered. Also there are some waveguides where the shape asks for compensation or correction, especially when not round like the driver. Fixing throat issues can be simple in some cases and unreasonable in others. Fixing the angle can sometimes be done by extending the throat back with a radius.
 
Last edited:
Quote:
Originally Posted by Jack Arnott
Re: solution 3. The BMS 4594 would be a great solution for that RCF 1.4” horn, since you already expressed interest in a BMS coaxial.

<Apart from the cost there seems to be few downsides to this setup. The driver seems to be basically the same as the 2" throat version? Do you know why this driver can play so much lower than most compression drivers?>


The BMS has a very large diaphragm area. It has "only" a 3.5" voice coil on the midrange, but it is a ring radiator, so it has a lot more of surface area outside the voice coil than inside it, and more surface area than a full dome with a 4" VC.



Quote:
Originally Posted by Jack Arnott
Regarding issue number 3; I think that in a one off situation, trying to match the exit angle to the horn would be very low on my list of concerns. Very low, like last on the list. If you had two horns that met every criteria, and one matched to driver’s exit angle and one didn’t, then I would go with the one that matched. Otherwise, I would not even ponder it.

<So this has only a minor effect on the systems response? is the effect of the mismatch to alter the on axis response differently to the off axis?>


I would think that if the driver has a different angle than the horn that this will cause diffraction, and differences mostly in the high frequencies. I think this will have less effect than many other considerations. If you were manufacturing your own horn, you would make sure this matched. It irks me when people put too much emphasis on this, and not on other considerations.



Quote:
Originally Posted by Jack Arnott
Regarding solution 3, SPL has nothing to do with driver exit size. EG, the BMS 1.4” driver will have the same SPL as the BMS 2” with the same characteristics, if the horn matches up the same.

<Doesn't the ratio of the exit to the surface area of the diaphragm define the compression ratio? so a larger diaphragm is usually going to have a larger throat and when running low like I would like to the SPL is going to be excursion limited so drivers that are meeting my requirements are more likely to have a larger throat.>


I should have qualified this by saying that if you have otherwise identical drivers, like the BMS 4592 vs the 4594, the smaller throat will not restrict the frequency response. The two drivers use the same diaphragms, and all the compression is done in the diaphragm exit, which is anular, so when it gets to the throat of the driver, all things are equal. The only difference is that the 1.4" exit driver is truncated, shorter. This means that you get into your horn sooner, and all the loading and dispersion is handled by the horn, and less by the driver. This gives the opportunity for better HF dispersion, and if the throat is loaded the same, and same final size requirements are met, the resulting frequencies that the combination can go down to will be the same.



Other drivers that are compressing into a phase plug might be limited by the horn being a smaller entrance.
 
I'm working with Don Keele and Bjorn Kolbrek on a horn design that have several of the same specifications. A 2-way design that requires a separate bass/subwoofer solution. May also get a horn subwoofer designed with it.

- 80x50 coverage angle
- Frequency range 90 Hz-18KHz or higher
- Crossover at 600 Hz or lower if the driver permits it
- Very uniform directivity low in frequency
- Exit angle of 1.4" (Radian 951BePB driver) but could also be made for other drivers including those with an 1.5" exit
- Non resonant enclosure

Downside is the size the horn requires for constant directivty low in frequency and it's not cheap to build either. I have a Klipsch K-402 horn which I will use for comparison to the top horn. The top horn needs to be better than Klipsch for it to be worth it.

Already have measured and tested the midbass horn. While we'll may be making some adjustments to it, it does work tremendously well today. The horizontal directivty doesn't change much before you get to 100-150 Hz. Picture of it below with the Klipsch K-402 horn on top.
 

Attachments

  • Midbass horn 002 (Large).JPG
    Midbass horn 002 (Large).JPG
    255.3 KB · Views: 1,453
Last edited:
The RCF HF94 gets a thumbs-up from me. I use it in my main PA system, and it does the job well. The HF950 appears to have a slight "pinch" before the final expansion, which isn't great for HOMs. It does load to a lower frequency, though.

I also use the 18Sound ND1460 (not -A. I use the titanium version - better for PA use), and that gets another thumbs-up. Titanium diaphragms are incredibly tolerant to abuse - you'll hear the diaphragm hit the phase plug in response to LF and little (if any) damage will occur. Aluminium diaphragms aren't so tolerant.

The thing to remember here is that larger exit diameters will beam at a lower frequency. A 1" exit driver will have reasonably good off-axis response past 10kHz, while a 2" exit driver will not. A 1.4" exit diameter is a decent compromise for a lot of applications.

Rod Elliott has a bit here: PA Systems
About compression driver exit diameters vs output vs LF cutoff.

The BMS coaxial compression drivers are really nice, but still constrained by the off-axis response of the exit diameter. ie, they might well get past 20kHz, but it could still be a laserbeam up there. Horns will constrain wavelengths that are trying to be omnidirectional, but when the compression driver itself is beaming, the horn becomes irrelevant.

My main PA system* will do what you're after with a bit of headroom, and sound good doing it. I wouldn't want to be in any domestic-sized space while it's run up to full power, though. Hopefully you're not planning on doing that.

*HF - 18Sound ND1460 on RCF HF94 horns
MF - 2x Faital Pro 10FH520
LF - Up to 4x Faital Pro 15HP1060
^ per side. You'd probably manage your SPL goals with 2x 15"s per side.

It'd be around 100 miles round-trip for you, but I'd be happy to demo that setup.
I've crossed over as low as 750Hz for PA use, which gives a really smooth off-axis response that matches to the 10"s well, but I usually run at about 1kHz out of concern for the HF driver's long-term health. No point in working them harder than necessary if the next band down can cover that range easily.

I also have a pair of 6.5" 2-way boxes that have 1" exit drivers. They'd probably manage 125dB peaks between them, but it'd be a couple of cycles "burst", while the 2x10"s can do >125dB sine waves.
If it's of interest, I wrote an article about SPL ratings in the PA world. It ain't pretty.
Spec Wars: Looking Inside Loudspeaker SPL Specifications - ProSoundWeb

Chris
 
Thank you for posting the URL (Grimshaw). I see an article on "Home HiFi and Measurements", so something even for the home hobbyist. I don't "do" PA, but I am among the converts to using "PA" systems in the home. In my case, I got a trashed pair of Yorkville Unity U15 at a good price. Good for me, since my construction skills are limited.
 
Is your listening room a typical home living room 5 Meters by 5-10 meters? IF so you may not need constant directivity below 700hz where the XT1464 will match well to a 15 and deliver awesome results.


Another used option would be a JBL PD 5322 or another JBL compression driver cone mid combination. That's a hard to beat combo to 300hz.

Otherwise the SEOS 24 with BMS should be studied.
 
I've used the P Audio BM D750's with some JBL 2385A horns crossed to a 15" driver at home and they do sound good. Only thing is the top octave doesn't sound as good as my beyma 380's. However the lower xo point gives better midrange compared to running the 15's up to ~1.5kHz.

The JBL's can be picked up on ebay, and P Audio do a copy I think.

edit: Think this is the P Audio copy
 
Last edited: