Double horns gentlemen. There's a contentious subject if ever there was one. What happens in them? Worth the effort? Problems? Benefits? How to enter parameters properly in MathCad (and other sim programs if you really must for some odd reason)? What other changes would be needed?
Thought I'd start a thread on the topic as there seems to be some interest, and I'd be interested in finding out some stuff myself, as I'm not a horn specialist. I gather a double horn should boost efficiency -I seem to recall 6db or something of the ilk, but I could be out there. However, you also need to account for the fact that the two mouths would be coupled differently, right? One to the floor, one effectively in free space, unless it's very close to the ceiling. I was going to ask Martin this in a private email, but I figured there's be room for a wider discussion as a forum thread. Should be interesting...
Best
Scott
Thought I'd start a thread on the topic as there seems to be some interest, and I'd be interested in finding out some stuff myself, as I'm not a horn specialist. I gather a double horn should boost efficiency -I seem to recall 6db or something of the ilk, but I could be out there. However, you also need to account for the fact that the two mouths would be coupled differently, right? One to the floor, one effectively in free space, unless it's very close to the ceiling. I was going to ask Martin this in a private email, but I figured there's be room for a wider discussion as a forum thread. Should be interesting...
Best
Scott
Greets!
Why? I mean 'W'/whatever folded layout bass bins are well understood. A 'FR' BLH is basically a BR with a huge horn vent and as best I can tell, both design theories peaked circa 1935.
?? The output travels down two paths instead of one.
No can do, though you could sim them both and do your own summing.
Depends on the app. A vertical layout will control vertical directivity to a lower frequency at the expense of some off-axis lobing just like a dual driver or MTM layout, so best to follow the same XO point limitations of each.
?? No efficiency gains over a single pathlength design terminating into an equal area mouth unless you design it in.
Yes, the floor coupled mouth will load to a lower frequency than the top unless the pathlength and mouth area are reduced to offset it: http://www.cain-cain.com/I-BEN/index.html Note the smaller lower mouth...............
GM
Why? I mean 'W'/whatever folded layout bass bins are well understood. A 'FR' BLH is basically a BR with a huge horn vent and as best I can tell, both design theories peaked circa 1935.
?? The output travels down two paths instead of one.
No can do, though you could sim them both and do your own summing.
Depends on the app. A vertical layout will control vertical directivity to a lower frequency at the expense of some off-axis lobing just like a dual driver or MTM layout, so best to follow the same XO point limitations of each.
?? No efficiency gains over a single pathlength design terminating into an equal area mouth unless you design it in.
Yes, the floor coupled mouth will load to a lower frequency than the top unless the pathlength and mouth area are reduced to offset it: http://www.cain-cain.com/I-BEN/index.html Note the smaller lower mouth...............
GM
Interesting. Thanks for that Greg! I take it you have a fairly low opinion of 'progress' in speaker design over the last 70 years or so then... 
You know, I never noticed in Terry's horns before that the lower mouth was smaller. Cunning. 'Can't remember where I heard the efficiency boost thing something I noticed on AA a while back I think -however, if you say it's wrong, that's good enough for me!
Hope you don't mid if I ask some basic questions here as horns are really not my speciality -I'm only just starting to get interested in the things. So: which elements would require summing for a sim? Throat area I assume would need it -mouth area too and general cros-section? I assume the horn-length would not though, providing they remained the same length. Compared to a single path horn, I also assume they would keep the same coupling chamber volume / dimensions, right?
Sorry for the basic questions!
Scott
You know, I never noticed in Terry's horns before that the lower mouth was smaller. Cunning. 'Can't remember where I heard the efficiency boost thing something I noticed on AA a while back I think -however, if you say it's wrong, that's good enough for me!
Hope you don't mid if I ask some basic questions here as horns are really not my speciality -I'm only just starting to get interested in the things. So: which elements would require summing for a sim? Throat area I assume would need it -mouth area too and general cros-section? I assume the horn-length would not though, providing they remained the same length. Compared to a single path horn, I also assume they would keep the same coupling chamber volume / dimensions, right?
Sorry for the basic questions!
Scott
I was having a look at some of the old threads on the FE166ES-R a few minutes ago since the new FE206ES-R is slated fro release soon, and ran across this comment by Terry Cain in response to a query regarding double-horns like his own BENs:
"You can double the factory design, will be very heavy. Keep the rear chamber the same volume or slightly larger (you can always add mass to make smaller). Stunning system like no other."
Interesting. I know he rates the Fostex Factory designsvery highly indeed, and whether or not people agree with all his design methodologies etc, Terry's no fool (to put it mildly). So it looks like it's possible to double the factory designs. However -and this is the question finally -how to go about it? Two throats, I assume, but you'd need to modify the driver position somewhat to achieve that properly I imagine to make things symmetrical, and that would mean they would be directly behind the driver, which doesn't strike me as being such a good idea given the shape of the Factory Design's compression chamber. Perhaps a couple of additional pieces extending into the CC, angling the throats away from the driver somewhat? Greg, Dave, Martin, (Terry if you're out there!), any other horn experts (I'm terrible on horns): any help you can provide on this will be much appreciated!
Best
Scott
"You can double the factory design, will be very heavy. Keep the rear chamber the same volume or slightly larger (you can always add mass to make smaller). Stunning system like no other."
Interesting. I know he rates the Fostex Factory designsvery highly indeed, and whether or not people agree with all his design methodologies etc, Terry's no fool (to put it mildly). So it looks like it's possible to double the factory designs. However -and this is the question finally -how to go about it? Two throats, I assume, but you'd need to modify the driver position somewhat to achieve that properly I imagine to make things symmetrical, and that would mean they would be directly behind the driver, which doesn't strike me as being such a good idea given the shape of the Factory Design's compression chamber. Perhaps a couple of additional pieces extending into the CC, angling the throats away from the driver somewhat? Greg, Dave, Martin, (Terry if you're out there!), any other horn experts (I'm terrible on horns): any help you can provide on this will be much appreciated!
Best
Scott
Greets!
You're welcome!
Yes, and no. There's been tremendous improvements in materials and manufacturing technology, and of course computers have allowed us to fine tune designs to a fare-thee-well if we want to, but there's been virtually no advancement in speaker design per se. As the inventor/horn designer Tom Danley lamented, "the ancients keep 'stealing' all my inventions".
The whole system would need summing since you'll have two complete horns driven by identical drivers except they will each have only 1/2 the Vas of the actual one, ergo its filter chamber will be half size. Once you have the two sims then you'll need to sum them. It still won't be all that accurate though since you can't account for the distance between them and the top mouth's loss of some floor loading that the sim will automatically add, so best IMO to just sim the complete horn and then split it up as required.
GM
You're welcome!
Yes, and no. There's been tremendous improvements in materials and manufacturing technology, and of course computers have allowed us to fine tune designs to a fare-thee-well if we want to, but there's been virtually no advancement in speaker design per se. As the inventor/horn designer Tom Danley lamented, "the ancients keep 'stealing' all my inventions".
The whole system would need summing since you'll have two complete horns driven by identical drivers except they will each have only 1/2 the Vas of the actual one, ergo its filter chamber will be half size. Once you have the two sims then you'll need to sum them. It still won't be all that accurate though since you can't account for the distance between them and the top mouth's loss of some floor loading that the sim will automatically add, so best IMO to just sim the complete horn and then split it up as required.
GM
question
if i'm correct, a horn's performance is related to the taper and also to a degree, the ratio between the driver area (Sd) and the throat area.
if we then implement a double horn by morphing a horn and its mirror image into one, what does this do to the horn's preformance, given my thoughts above?
for simplicty's sake, assume that there is no compression chamber.
another way of stating my question:
assume a horn profile A,
assume we then construct two horns B1 and B2 by splitting A through its center into two halfs (each has half the cross sectional area of A at an identical point along their length).
what would the difference in performance be between A driven by driver D
vs.
B1+B2 each driven by a seperate identical driver D and situated next to each other
thanks,
vin
if i'm correct, a horn's performance is related to the taper and also to a degree, the ratio between the driver area (Sd) and the throat area.
if we then implement a double horn by morphing a horn and its mirror image into one, what does this do to the horn's preformance, given my thoughts above?
for simplicty's sake, assume that there is no compression chamber.
another way of stating my question:
assume a horn profile A,
assume we then construct two horns B1 and B2 by splitting A through its center into two halfs (each has half the cross sectional area of A at an identical point along their length).
what would the difference in performance be between A driven by driver D
vs.
B1+B2 each driven by a seperate identical driver D and situated next to each other
thanks,
vin
More useful info. Cheers for that as always Greg, and sorry for my asking what are I imagine some very daft questions! Horns. I must spend more time on the theories. In the meantime, when I work on my own designs in MathCad, I'll follow your advice and simply model one, and then split the thing in half when drawing it up, if I fancied doubling the thing. One last question if I may? Hopefully it'll make what I was talking about in the last post a bit clearer -looking at it, it's almost unintelligible!
'Found a picture. I didn't draw this myself -this is from that AA discussion I mentioned that I'd run across some time ago. I'd forgotten that I'd saved it -no idea who it was who drew them though (apologies if it's someone here!). There's two ideas here for a double horn -doubled versions of the Fostex Factory FE108ESigma enclosure in this case. I woudn't choose this layout myself I I was designing one obviously, but for doubling the existing horns as Terry suggests is possible, there isn't a while lot of choice.
Now, I'm leaning toward the right-hand design as being the more logical. However, you can see what bothers me a bit on it quite clearly; the throats are almost directly behind the driver. And with the CC being a shallow, vertical type, I'm thinking this might affect things a bit. Or am I worrying about nothing in that regard and it'll work just the same?
Regards
Scott
'Found a picture. I didn't draw this myself -this is from that AA discussion I mentioned that I'd run across some time ago. I'd forgotten that I'd saved it -no idea who it was who drew them though (apologies if it's someone here!). There's two ideas here for a double horn -doubled versions of the Fostex Factory FE108ESigma enclosure in this case. I woudn't choose this layout myself I I was designing one obviously, but for doubling the existing horns as Terry suggests is possible, there isn't a while lot of choice.
Now, I'm leaning toward the right-hand design as being the more logical. However, you can see what bothers me a bit on it quite clearly; the throats are almost directly behind the driver. And with the CC being a shallow, vertical type, I'm thinking this might affect things a bit. Or am I worrying about nothing in that regard and it'll work just the same?
Regards
Scott
Attachments
Based upon what you said in your two responses Greg, now I've re-read them a few times and done some thinking, I think I've got the idea. (I'm now keeping my fingers firmly crossed!)
Say you were modelling the enclosure in MathCad, and were going for a TC Double BEN type design, I assume that you would double the CC length, the horn throat area and the horn mouth area, whilst holding all the other parameters as per a single horn. (I'm thinking that the horn length will also remain unaltered because the overall path-length hasn't changed: there's simply two of them being fed by the driver, rather than 1 that's twice the cross-section.) Am I on the right lines with that?
Cheers for that, and I won't ask any more questions!
Scott
Say you were modelling the enclosure in MathCad, and were going for a TC Double BEN type design, I assume that you would double the CC length, the horn throat area and the horn mouth area, whilst holding all the other parameters as per a single horn. (I'm thinking that the horn length will also remain unaltered because the overall path-length hasn't changed: there's simply two of them being fed by the driver, rather than 1 that's twice the cross-section.) Am I on the right lines with that?
Cheers for that, and I won't ask any more questions!
Scott
have you had a look at Onur Ilkorur's horn for a Fostex fe206e? Check it out at this url. http://www.yildiz.edu.tr/~ilkorur/speaker/fostex_fe206e.htm
Regards Moray James.
Regards Moray James.
zobsky said:
Greets!
Right, increasing its compression ratio (Sd:St) increases its BW, acoustic power and decreases its efficiency.
Since the driver didn't change, ergo each horn's CR did, the two will sum to a different FR and power response than the single horn driven by only one of the drivers since the summed horn is being driven by a summed driver with 2x Vas at a 2x CR.
GM
Scottmoose said:
Greets!
Neither is the correct way to make a DBLH, at least WRT splitting up a single path design. Once you move away from doing that, then either will work depending on the gain/BW desired.
WRT the throat behind the driver, this is no different than in a BR or FLH, i.e. it will exhibit max loading on the vent/throat. My attitude is that for 'FR' driver BLH designs, the CC should be calc'd as a MLTL for theoretically best performance.
GM
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.