Want to build high frequency horns, help please!

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Administrator
Joined 2004
Paid Member
<snip>
I was told this Beta 12 mid would play out to about 2.5k by a respected individual in the industry. I do understand beaming, and this seems high to me with regards to beaming, but I was going to try it anyhow because I definitely trust this individual's advice. I actually was going to target closer to 2khz from that mid to my tweeter as a starting xover point, but I haven't taken measurements of either the mid or the tweeter yet either, and those will obviously dictate the xover point.

If the Beta 12A curves shown are representative of your drivers I would cross it over much lower, certainly no higher than 700 - 800Hz. Should refer to the Beta12 as a woofer.
 
The two style waveguides I would recommend for home use, SEOS and Geddes, would be very difficult to hand make in wood.

My only reserve is that if I made a Seos style WG, my tweeter's axis will be at about 42" (and that's after I cut that woofer cabinet down 4.5", which I plan on doing). I guess that's not ridiculously tall, and I've always thought a tad high with regards to a tweeter is better than too low. This is also assuming I used something like a ~6" x 12" Seos type WG.

I think the Geddes type (I'm guessing this is referring to the round WG?) will definitely put me too high, but I'm open to suggestions or correction in my assumption here too.

That said... what's the difference between the Seos type waveguide and the the one I have now? Aside from minor differences in their shape, don't they essentially do the same thing? This is the one I have now... Pyle PH612 1" Screw-On Constant Radiation Horn. In addition to that, I also have a larger Eminence that's similar, just bigger and deeper.

Also, with regards to dispersion and sound, what's the difference between the Seos, and something like these? These are what's used in Volti's speakers, and hearing these speakers at AXPONA are part of what made me want to try something like this...

upload_-1.jpg


upload_-1.jpg


And there's this type, which I'm guessing would have a similar dispersion to the Pyle horn I currently have...

upload_-1.jpg


Lot's of good advice here to help you find your way. But how big a horn do you need? What will be your crossover point? I have not herd the Beta 12, but the Beta 12 LTA crossed at 6K works like a charm. Ditto for the Beta 8.

Any idea of your crossover point?

For more clarity... based on hearing that Beta 10 crossed at ~2.2k and loving the way it sounded, plus what I've been told about the Beta 12 working well out to 2.5k, I'm guessing I'll be somewhere in the neighborhood of 2khz for my xover from the Beta to the Selenium tweeter. But that's a total guess on my part, I haven't measured or SIM'd anything yet.
 
Due to the rising inductance and break up, I would cross close to 1k. Also, depending on the waveguide you choose, you may also want to cross lower for a good directivity match.

Based on the fact that I've worked with the Eminence Beta 10 crossed at 2.2k and it sounded great, and the fact that the Beta 10 and the Beta 12 measure fairly similarly, plus the fact that the guy that engineers these speakers told me that he would cross them to the Selenium tweeter above 2k... I think I'll start at 2khz. But thanks for the input.
 
Selenium d220 goes out to 20khz so why do you need a tweeter? What horn are you attaching to the selenium? Is it the Ti version? Its recommended for a 1.5-2khz xover. The 10" may go out to 2khz, but whether it sounds good out there is another story. My 10" RCF is xover at 1200 with good results. If you try active with a minidsp or something similar you can slide the in realtime and listen. Combine that with polar response curves from you mic and it will reveal a good xover point. My guess is 1.5 somewhere.
 
Founder of XSA-Labs
Joined 2012
Paid Member
I don't need full range, I need highs. I already have the bass and midrange.

A ScanSpeak 10F/8424 loaded on a tractrix allows it to reach 22kHz with great authority and low distortion. Fiberglass cones sound so much smoother than a metal diaphragm compression driver. It does highs very well - better than a CD in my opinion. Piano sounds like I have never heard it before.
 
Selenium d220 goes out to 20khz so why do you need a tweeter? What horn are you attaching to the selenium? Is it the Ti version? Its recommended for a 1.5-2khz xover. The 10" may go out to 2khz, but whether it sounds good out there is another story. My 10" RCF is xover at 1200 with good results. If you try active with a minidsp or something similar you can slide the in realtime and listen. Combine that with polar response curves from you mic and it will reveal a good xover point. My guess is 1.5 somewhere.

The whole purpose of this thread is to figure out which horn I want to use. As stated a couple times earlier, I'm using the Selenium D220ti, and I'm aiming for around 2khz.

I understand how to measure speakers to design a crossover very well. The purpose of this thread is to find out how the different tweeter horns effect the sound, and how someone goes about deciding which horn to use, since I've never worked with them before. Once I figured that out, I wanted to make one out wood, rather than using the plastic ones. And (as stated earlier) I'm not using the 10", I'm using the 12". I only said it "plays well out to 2.5k" because that's what the engineer for Eminence told me, and while I haven't seen the polar response measurements, I have seen the on axis response. Based on that, it looks very usable out to at least 2k, but I can't know how far until I measure these myself, and get on and off axis measurements.
 
Last edited:
A ScanSpeak 10F/8424 loaded on a tractrix allows it to reach 22kHz with great authority and low distortion.
xrk971,

The OP already has Beta 12s for mids, capable of around 123 dB at one meter, and Selenium drivers also capable of that kind of output.

Although you have posted graphs showing your ScanSpeak 10F/8424 loaded on a tractrix has low distortion at low volume, what does the distortion and output level look like at full rated power?
I'm guessing an order of magnitude lower output than the Beta 12, and HF distortion exceeding 100%.

Art
 
It's really up to what you want to build first, and what your woodworking skills are. I would recommend a Biradial Yuischi horn. Use a hand planer, or buy them already Cnc'd and do the sanding.
4263459e4a13137c0a5e470a9439e85c.jpg


Or you can do a bent wood tractrix which I'm also building.

I used hornresp to get the dimensions and curve.

Your going to get a million opinions, but at some point you need to just start cutting!

e88a2bde255dfe457b280b65bdb47466.jpg
 
A ScanSpeak 10F/8424 loaded on a tractrix allows it to reach 22kHz with great authority and low distortion. Fiberglass cones sound so much smoother than a metal diaphragm compression driver. It does highs very well - better than a CD in my opinion. Piano sounds like I have never heard it before.

Thanks. While this is definitely interested and might be something I'd like to try in the future, I already have the Beta mid and the Selenium D220 tweeter. That said, I'd love to hear yours one of these days.
 
It's really up to what you want to build first, and what your woodworking skills are. I would recommend a Biradial Yuischi horn. Use a hand planer, or buy them already Cnc'd and do the sanding.
4263459e4a13137c0a5e470a9439e85c.jpg


Or you can do a bent wood tractrix which I'm also building.

I used hornresp to get the dimensions and curve.

Your going to get a million opinions, but at some point you need to just start cutting!

e88a2bde255dfe457b280b65bdb47466.jpg

Thank you, this is EXACTLY what I wanted! Did you read the first post?? I used to build custom kitchens, and I have a fairly extensive workshop, but not one that includes a CNC router. But, I can make that by using the "hillbilly lathe" method (photo taken from the internet)...

upload_-1.jpg


And at this point I think I agree, just measure and cut. I just wanted to check in and see if there was anything I obviously shouldn't try.
 
But, I can make that by using the "hillbilly lathe" method (photo taken from the internet)...

And at this point I think I agree, just measure and cut. I just wanted to check in and see if there was anything I obviously shouldn't try.
The Biradial Yuischi horn and "hillbilly lathe" look great. A horn like that will load well to below 1000 Hz, allowing a wide latitude in crossover points.

Although the on axis response of the 12" looks "usable", when you start measuring the off axis response you will find things don't look anything like that, the driver does not act at all like a piston above the flat region of response, and dispersion is all over the place, varying by 30 degrees (or more) over a narrow bandwidth.

Art
 
One option is to hand jig saw out a series of templated layers then use a hand planer to trim down the angles once glued and stacked. Plus a hand belt sander to finish. Lots of hand work but rewarding. I got half way through a 480hz Yiuschi before moving onto something else. Still might do it though.
730c5a480f2336ac26e207e913e4ed98.jpg
I have a scroll saw, a stationary belt sander, plenty of routers, a couple different bandsaws, and a slew of different sanders (as well as a few different Jasper Jigs for the radius). I'll play and see what I can come up with.

Strictly from aesthetics, the Yuischi type is the mist appealing. And since that'll give a wider FR, that's what I'll start with.

Thanks!!
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.