Imagine getting from a 210 cm2 cone to a hornthroath of 31.10 cm2 and you will probably see the problem. If this frontchamber is only 3/4 inch thick. I hope this helps, i am not a Hornresp expert.
Ah, I see. I'm creating a throat area that's not easily doable with standard material sizes.
I've gone with a throat chamber of 0.5 inches, so all I'd need to do is cut a 210cm^2 hole in the middle. This is what I end up with:
That doesn't seem to be too bad, as I would have to just cut a hole with a 16.35cm diameter (approximate). Funnily enough, inch measurements are closer, as it comes to as near as 6 7/16" as you can get!
It took me a few moments to understand what I'd done wrong, but I feel like I'm getting there now.
I've gone with a throat chamber of 0.5 inches, so all I'd need to do is cut a 210cm^2 hole in the middle. This is what I end up with:
That doesn't seem to be too bad, as I would have to just cut a hole with a 16.35cm diameter (approximate). Funnily enough, inch measurements are closer, as it comes to as near as 6 7/16" as you can get!
It took me a few moments to understand what I'd done wrong, but I feel like I'm getting there now.
Given the tools I have, I think a square throat chamber is going to be easier 🙂
So, I've exported the data from HornResp, and mocked up the horn in cardboard:
It's nice to see the output of the results. The WAF is a little low, but my wife's wonderful and would accept it if it wasn't in her view all of the time 🙂
Still, I've been wondering about bending or folding it like a Sato. Is there any software out there that's suitable for doing that?
I've seen some debate about bending the horn's path vs. folding the horn with reflectors in the path, and also some concerns when using that approach. Is that completely the case? Do the Sato horns exhibit 'problems'?
So, I've exported the data from HornResp, and mocked up the horn in cardboard:
It's nice to see the output of the results. The WAF is a little low, but my wife's wonderful and would accept it if it wasn't in her view all of the time 🙂
Still, I've been wondering about bending or folding it like a Sato. Is there any software out there that's suitable for doing that?
I've seen some debate about bending the horn's path vs. folding the horn with reflectors in the path, and also some concerns when using that approach. Is that completely the case? Do the Sato horns exhibit 'problems'?
would it be ok to make two of the 'sides' conical(curve linear?), and then curve the two other sides more instead ?
not sure, but might be easier to construct, maybe
but I suppose it also changes 'mouth directivity'
not sure, but might be easier to construct, maybe
but I suppose it also changes 'mouth directivity'
Hi unaHm,
Looks really nice in mock up. I don't think it is that big... I would still be a bit worried about tractrix and the compression rate. You might end up with a horn that is usable 2x Fc. See volvotreter. It's not my intention to rain on your parade with my repeate message of compression and tractrix. I would hate to see you do so much woodwork on a horn an be disapointed 🙂.
Volvotreter uses +- 0.58 x Sd =compression for this 15 inch Conical midbass and 0.4 for his tractrix horn. See his comment about the tractrix and usable range.
Looks really nice in mock up. I don't think it is that big... I would still be a bit worried about tractrix and the compression rate. You might end up with a horn that is usable 2x Fc. See volvotreter. It's not my intention to rain on your parade with my repeate message of compression and tractrix. I would hate to see you do so much woodwork on a horn an be disapointed 🙂.
Volvotreter uses +- 0.58 x Sd =compression for this 15 inch Conical midbass and 0.4 for his tractrix horn. See his comment about the tractrix and usable range.
Last edited:
There are measurements showing the Sato has very smooth on axis response, but I have seen no off axis response curves.
Your horn is very beamy.
Have you listened to it yet?
If you tape the corners completely it will be a very good representation of what you can expect from the horn, though wall vibration will limit it to very low input levels before it "funks out".
You will be probably be surprised at what 1/4 watt will sound like with your cardboard horn.
For as little curve as is going on, this horn could be made as two flat sided conical sections, "breaking" around 1/3 from the mouth end.
There would be virtually no difference in measured on axis or off axis directivity in the pass band between the curved and straight side two part versions.
The straight side version could be made of thick plywood, far easier to build than laminating thin plys.
Art,
Thanks very much for the guidance! I'll export the 'square' profile from HornResp and try again.
From what I remember, the difference in the curve was so minimal (fractions of a mm) that I should be able to get away with two of the pieces I have now, and then two straight pieces.
In relation to the compression 'worry' that Phenoholic Anonymus has, what are your thoughts on that? I suppose with what I have it is higher (6.75:1), so I'll try something with a wider throat to see what HornResp puts out if you think it's necessary.
Thanks everyone for the comments! Sometimes it being 'out in the real world' inspires the next step 🙂
Here's a bigger throat, so the compression is now down to about 3:1:
The response might need some EQ'ing to flatten it out, but it doesn't seem too bad.
Thanks very much for the guidance! I'll export the 'square' profile from HornResp and try again.
From what I remember, the difference in the curve was so minimal (fractions of a mm) that I should be able to get away with two of the pieces I have now, and then two straight pieces.
In relation to the compression 'worry' that Phenoholic Anonymus has, what are your thoughts on that? I suppose with what I have it is higher (6.75:1), so I'll try something with a wider throat to see what HornResp puts out if you think it's necessary.
Thanks everyone for the comments! Sometimes it being 'out in the real world' inspires the next step 🙂
Here's a bigger throat, so the compression is now down to about 3:1:
The response might need some EQ'ing to flatten it out, but it doesn't seem too bad.
Last edited:
IMHO midrange horns aren't practical in the home. It's because you need to be a significant distance away to achieve any type of pattern control. For instance, 2khz is 17cm long. With a wavelength that short, it doesn't take much of a waveguide to achieve pattern control. OTOH, 200hz is 1.7 meters long. At that length, you'd need to listen in a 1200' room to even get far enough from the horn for it to sound smooth.
I found this out the hard way of course; I own a set of Summas and I find that they like a large room, much larger than what I can afford. (I live in one of the ten most expensive cities in the United States, a 3br house is about $750K here.)
My favorite solution, at the moment, is to use dipoles or cardioids for the midrange. That gives you pattern control without requiring a cabinet the size of a refrigerator. For the upper frequencies I'm using waveguides, as high frequency waveguides don't need a ton of space.
If you primary goal with horns is efficiency, there are plenty of high efficiency midranges that are clean and dynamic, like the B&C 8NDL51 or it's ferrite sibling.
John,
Pattern control is determined by the horn type, path length and mouth size, but has nothing to do with the horn mouth to listening position distance.
Smoothness, or the lack of, also has nothing to do with listening distance.
The "pattern control" of a dipole is nothing like that of a large horn.
The OP appears to want the cheapest route for high efficiency, but does not seem concerned about pattern control.
Art
I'm not sure what Phenoholic Anonymus compression 'worry' is, and saw nothing in Volvoreter's horn builds explaining any throat size decisions.
My main concern with high compression ratios is cone deformation or tearing at high power levels.
3/1 is certainly safe with the Alpha 8 driven at or below Xmax.
VTC (volume of the throat chamber) is the volume of the area between the cone and the throat.
That area is roughly cone shaped, with the cones dust cap forming a dome in the center of the volume.
How did you determine the VTC in your simulations?
Have you experimented with different VTC volumes?
Ok here we go again.
When you have two radiators that are spaced apart, the two radiators will interfere with each other. Some of radiation will be in-phase, and some will be out-of-phase. The radiation creates lobes in the vertical response:
One way to minimize the lobes is to move the drivers together; one way to maximize the lobes is to move them apart.
Assuming the OP intends to use a tweeter, the use of a very large midrange horn is going to create some dreadful lobes in the response.
Due to those lobes, the closer you get to the speaker, the more you'll have to listen with your head in a vise. For instance, if the on-axis lobe is 30 degrees wide, then the 'sweet spot' is larger at ten meters than it is at one meter. This is one of the reasons that large two-ways like my Summas sound good in big rooms. I think that a lot of people who haven't listened to speakers with midranges that are 12", 15" or larger aren't aware of how they sound in a regular room. (IE, they *need* a big room. Or better yet, an auditorium.)
I have personally measured this phenomenon with a microphone; as you move the mic you can see the response change dramatically once you're outside of the listening axis. The best illustration of this that I have seen is from Wayne Parham, he has a video that illustrates how these lobes behave. It's named "Vertical_nulls.wmv"
A horn or waveguide is arguably the 'ultimate' solution for directivity control. A dipole or cardioid is not as efficient as a horn or waveguide, not even close. But power is cheap and real estate is expensive. I am suggesting that there are alternatives with high WAF and low space requirements which are very practical in the era of $250 kilowatt amplifiers.
I personally find high output more important than high efficiency; that's why I suggested something like the 8NDL51. An 8NDL51 in a conventional box will not be as efficient as a cheap driver in a midrange horn. But it *will* be easier to cross over due to it's wider bandwidth. And the output of the 8NDL51 may be higher, due to it's low power compression, high power handling, and relatively high efficiency.
Also, please do not use my real name on the forum. I use a pseudonym because I work in software and I want to keep my business and personal life separated. I know that you are an audio professional and that posting on an audio site is consistent with your professional life, but I'm not in the same boat.
Last edited:
"Patrick",
Your preferences noted, but of what import are they in a thread called "midbass horn" in which the OP has already purchased the mid range driver?
Art
To be honest, I experimented 🙂 I was thinking of matching the Sd of the driver, and having something useful fit into a 1/2" thick piece of wood or MDF.
The throat I ended up with obviously had straight 'walls', as I'd not considered that. Thanks for pointing that out, it makes a lot more sense now.
So, for this application, does it make sense to have a throat start at Sd (210cm^2) and then taper down to S1 (63cm^2)?
Hornresp figures the speaker as a flat piston, which it seldom is,and a Alpha 8 definitely is not.
The VTC (volume of the throat chamber) is the volume of the area between the cone and the horn throat.
It is not possible in the real world to have a VTC = 0, as the cone excursion would need to be figured in, or the driver would hammer the face plate.
At 3.2 mm(Xmax) x 210 Sd, the minimum theoretical VTC is 672.
Since you want to allow for more than Xmax without hammering the horn face plate, you need to either put a spacer between the driver and mouth face plate, or rout an excursion ring (you can see that feature in the Paraline cutouts I posted previously) either of which increases VTC more.
Now, fill the cone up with rice flat to the surround, and measure how much volume that is, and add to the above figure.
Even if you sim the exact VTC, the cone breakup will insure the actual response will not match Hornresp, so best start measuring your actual horn 🙂.
Art
Wow, that's nearly 5 times the price of the house we live in. I bet the income of that area is 5 times what I receive too. I've been in the USA since 2009, and in terms of music and audio Colorado's not an easy location.
Thanks 🙂
That was the idea that popped into my head too, where the mass of air between the cone and the plane of the mounting ring would suffice as the Vtc, plus the accommodation for the cone excursion as you'd mentioned.
So, the Xmax is about 1/8", so using a spacer of 1/8" (maybe masonite) would allow for a 1/16" gap between the cone and the mounting plate for the horn if the driver met its Xmax. I think I'd take that route, as it would be easier to calculate the extra volume than routing the excursion rings.
Time to get some rice! 🙂 I'll post the volume I get so that others can use it.
Edit: This is slightly generous as the rice I'm using is a bit 'rough', but I'd say 350ml/cc. To be conservative it would be around 325.
Last edited:
- Status
- Not open for further replies.