***************************************** Jacek asked this: define mouth size, throat size (driver model), vertical and horizontal coverage, midrange throat location and area/diameter, number of drivers. ***************************************** The idea is to only reproduce the section of the horn with the compression driver and midranges. Then extend about 1" past the midranges so the plastic horn can be mounted to a larger wooden horn. I know the throat size will be 1". I'm not sure if it can incorporate the SEOS shape at the beginning and then straighten it out so there is a smooth transition from plastic to the larger wooden horn. For the first model, should we only use two midranges instead of four? I think the most obvious choice to start with is the 4" sealed back Celestion. Is there any reason to have a different vertical/horizontal ratio or should they be the same? If the base starts as a SEOS shape, maybe the ration needs to be different, and maybe we can only use two midranges?
Erich -- the h/v ratio doesn't affect the SEOS shape, and I think the SEOS shape (actually it's just the EOS shape until you get further out from the throat) would be good to have if only because of the brand. I kind of like the idea of having the Horizontal and Vertical angles being the same (say 80 degrees?), though that means the woofers pretty much also have to output through the horn walls to get them close enough to the other drivers. I personally think two of those mids would be more than enough, though probably other people would go with the "too much isn't enough" philosophy. Maybe the way to do it is to put mounting areas and screw attachments on all four walls for midranges, with indentations or other markings to indicate where output tap holes can be drilled. Then the builder can go either way, or even with just one or even no midranges. The "frustrums" could even be molded into the outside walls, but the actual holes to be drilled through by the customer (or as a pre-shipping step if you are sending kits of various types). How about if I worked up the dimensions for an 80x80 horn, with likely-to-work-ok tap points for midranges, and you can figure out the throat shape, how to attach the midranges, and where and how the walls transition from plastic to wood? The woofers should of course mount on the wood portion (or primarily on the wood portion at least). And how big is this beast to be? I guess with the plastic throat and wooden walls you could make different sizes with the same plastic part. I think the big seller would be about 30 to 36 inches per side, should get the testosterone crowd drooling. But a 20 or 24 inch per side version might be preferred by people who don't have a theater space. Another thing to consider is how does this horn mount to its enclosure? The woofers need to have an enclosure behind them, so there has to be something fabricated onto the wooden horn mouth so it can mount, maybe from the front with a mounting ring like a normal driver would.
The H/V ratio would affect the plastic part after a few inches from the throat, though, so that has to be decided. If you have lots of cash you want to invest, you could make two versions, on an 80x80 and another the usual 90x60.
Sorry Bill, when I was talking about the horizontal/vertical ratios and the SEOS shape, I was just wondering if it should look similar to the current SEOS waveguides. I do realize the SEOS curve can be used regardless of the ratio. But should it be something like 80x80 or 90x60? I'm not sure. Personally I would like 90x60 (or similar) just so it looked a little like the other waveguides. But if 80x80 is easier or better, that's fine. I think we should go ahead and do 4 midranges because it's not going to change the cost enough to worry about and the midranges are cheap anyway. I think we should just go ahead and have the holes already drilled out for the midranges and even a method to mount them on the back side. If smaller midranges are used in the future, I'm guessing we can just make an adapter or possibly just make a new model. The overall size (including the wood section) will be 24" and larger. Probably 30" and larger. The plastic section might only be 8" x 8". Mounting the smaller plastic section to the wood horn should be pretty easy to figure out once the design gets started. I have a few ideas for that, but need to draw them out somehow. Are you able to describe the frustrums for Jacek? I'll post some pictures that might help.
I guess the main thing we need to figure out is the ratio and then the approximate distance the mids will be from the compression driver intake.
The frustrums aren't just something generic like that. The volume between the front of the midrange cone, the length of the port into the horn, and the port opening size are all variables you adjust in the HornResponse model. A frustrum plays into the effective value of all three of those. A frustrum will vary with how gradual its walls are (more gradual also changes the volume in front of the midrange cone, but shortens the effective length of the port) so those need to be determined ("estimated" is more like it, since the HornResponse model isn't exact of course). Also, the port opening is usually near the edge of the cone (otherwise the ports would come out too far from the tweeter) so the frustrum isn't just a simple cone-shaped hole, but is usually steeper near the edges of the cone and more gradual toward the center. Let me run models of a 90x60 (probably that would interest most users) and of a 80x80 (which would probably be better acoustically but kind of klunky looking) and see what the parameters look like. I still strongly suggest you make all but two of the midrange ports be kept "optional" (still blocked-over inside the horn unless fully later drilled out afterwards) as I'm very convinced that fewer port holes on the horn walls near the tweeter give better performance for the tweeter. The HF response of the Synergy-ish "SmallSyns" I made is MUCH smoother (even when using the same tweeter driver) than earlier versions that had 4 ports rather than 2, and much smoother than my bigger Cosine horns (that also use more and bigger ports). In the SmallSyns, there aren't weird kinks in the tweeter frequency response like when unnecessary ports are used. More ports are only needed for more midrange drivers for bigger midrange SPLs such as for public address sound. But not for home hi-fi or even home theater. Fewer ports is a nicer arrangement for the tweeter driver, which is also important. Undoubtedly some customers will focus on the midranges and decide "too much is not enough", but a good design optimizes tradeoffs and reducing port edges when not making speakers intended to be installed in Yankee Stadium is a good tradeoff!
Here are result of a Horn Response diesign for the 90x60 horn -- __________________________________________________________________________________ SynROS 90x60, modeled using midranges = Celestion TF0410MR Sensitivity using 2 midrange drivers (in parallel) or with 4 midrange drivers (series-parallel) would be about 100dB SPL/1m/2.83V. It would be about 95dB using a single midrange driver. Either would match up well with usual compression tweeters. Ports distance from the CD diaphragm should be 8.4cm, measured on-axis of horn Port opening cross-sectional area: 2.5 square cm for EACH MIDRANGE DRIVER ( = 2x ports of 1.26cm diameter per driver; OR 1x port of 1.78cm per driver.) Approximate length of ports: 1cm (up to about 2cm if in a frustrum). Frustrums should be relatively narrow and long, leaning outward toward the center of the midrange driver, since the air volume in front of the cones (see below) needs to be kept kind of small. The ports should be kept close to each other -- so if ports are on the top and bottom of the horn, then the ports on the top (or bottom) should be about the same distance from the other ports on the same wall as they are from the matching port on the opposite wall. *DON'T put them in the corners of the horn (because that puts them too far from each other and can cause polar issues near the midrange-tweeter crossover frequencies). The air volume (compression chamber) in front of EACH midrange driver, before the port = 35 cubic centimeters (including about ~70% of the volumes of the frustrums) NOTE: THIS CAN BE TRADED OFF AGAINST PORT LENGTH -->a SHORTER PORT LENGTH allows MORE AIR VOLUME Woofers would crossover to midranges at a little under 400Hz (because of the midrange driver), so the centers of the woofer ports need to be within 19cm of the tweeter diaphragm. Assuming the woofers will be mounting on the wooden parts of the walls, then the plastic part needs to not be deeper (distance from the tweeter diaphragm, on the horn axis) than about 17cm. -Bill
Here's a similar design, but for a 80x80 horn (only the size of the ports changes from the above 90x60 design): __________________________________________ SynROS 80x80, modeled using midranges = Celestion TF0410MR (design 1) Sensitivity using 2 midrange drivers (in parallel) or with 4 midrange drivers (series-parallel) would be about 99dB SPL/1m/2.83V. It would be about 94dB using a single midrange driver. Either would match up well with usual compression tweeters. Ports distance from the CD diaphragm should be 8.4cm, measured on-axis of horn Port opening cross-sectional area: ~2.25 square cm for EACH MIDRANGE DRIVER ( = 2x ports of 1.2cm diameter per driver; OR 1x port of 1.7cm per driver.) Approximate length of ports: 1cm (up to about 2cm if in a frustrum). Frustrums should be relatively narrow and long, leaning outward toward the center of the midrange driver, since the air volume in front of the cones (see below) needs to be kept kind of small. The ports should be kept close to each other -- so if ports are on the top and bottom of the horn, then the ports on the top (or bottom) should be about the same distance from the other ports on the same wall as they are from the matching port on the opposite wall. *DON'T put them in the corners of the horn (because that puts them too far from each other and can cause polar issues near the midrange-tweeter crossover frequencies). The air volume (compression chamber) in front of EACH midrange driver, before the port = 35 cubic centimeters (including about ~70% of the volumes of the frustrums). NOTE: THIS CAN BE TRADED OFF AGAINST PORT LENGTH -->a SHORTER PORT LENGTH allows MORE AIR VOLUME Woofers would crossover to midranges at a little under 400Hz (because of the midrange driver), so the centers of the woofer ports need to be within 19cm of the tweeter diaphragm. Assuming the woofers will be mounting on the wooden parts of the walls, then the plastic part needs to not be deeper (distance from the tweeter diaphragm, on the horn axis) than about 17cm.
And an alternate 80x80 design, in case there are problems getting the ports within 8.4cm of the compression driver diaphram or with getting the air volume small enough: ____________________________________________ SynROS 80x80, modeled using midranges = Celestion TF0410MR (design #2) Sensitivity using 2 midrange drivers (in parallel) or with 4 midrange drivers (series-parallel) would be about 99dB SPL/1m/2.83V. It would be about 94dB using a single midrange driver. Either would match up well with usual compression tweeters. Ports distance from the CD diaphragm should be 10.2cm, measured on-axis of horn Port opening cross-sectional area: ~2.1 square cm for EACH MIDRANGE DRIVER ( = 2x ports of 1.16cm diameter per driver; OR 1x port of 1.63cm per driver.) Approximate length of ports: 0.7cm (up to about 1.4cm if in a frustrum). Frustrums should be leaning outward toward the center of the midrange driver. The ports should be kept close to each other -- so if ports are on the top and bottom of the horn, then the ports on the top (or bottom) should be about the same distance from the other ports on the same wall as they are from the matching port on the opposite wall. *DON'T put them in the corners of the horn (because that puts them too far from each other and can cause polar issues near the midrange-tweeter crossover frequencies). The air volume (compression chamber) in front of EACH midrange driver, before the port = 60 cubic centimeters (including about ~70% of the volumes of the driver's frustrums). NOTE: THIS CAN BE TRADED OFF AGAINST PORT LENGTH -->a SHORTER PORT LENGTH allows MORE AIR VOLUME Woofers would crossover to midranges at a little under 400Hz (because of the midrange driver), so the centers of the woofer ports need to be within 19cm of the tweeter diaphragm. Assuming the woofers will be mounting on the wooden parts of the walls, then the plastic part needs to not be deeper (distance from the tweeter diaphragm, on the horn axis) than about 17cm. ______________________________________________ If needed, I can probably alter the 90x60 horn design like this if port distance or compression chamber volumes are a problem in the first design. The disadvantage of that is that the midrange-tweeter crossover frequency may need to get pushed down a little (which means smaller compression drivers probably couldn't be used with it). -Bill
Sorry Bill, I'm working on the new website and looking back at some responses for other Synergy's. There was an audio GTG a few weeks ago and someone brought a Synergy speaker. The response did look a big rough and I'm guessing some of that could be because 4 midranges are used versus 2. So maybe we should just stick with 2 mids right now and not even worry about 4 at all. My only concern is that they might not be efficient enough for some dual 12" or 15" Kappalites. But The bass could always be run hot like we did with the 1899. Actually, I guess it wouldn't be off by too much anyway. I think Matt said dual 15" Kappalites would come in around 104db, so it's no big deal. Unless you see any real issues, I think we should lean more towards trying the 90x60 just so it looks a little different. But it sounds like you might think the 80x80 is the easier way to go? It would likely be easier to design and with less chances for mistakes. Can you explain this a little more for the compression chamber: "35 cubic centimeters (including about ~70% of the volumes of the frustrums)" Do you mean we need to design a 'stand off' for mounting the woofer that is tall enough to get a total volume of 35 cubic centimeters? Did you already take into account the slight concave portion of he woofer cone?
Hi Mr. Eadmin! Keep in mind that the Danley speakers are designed with clean SPL as a priority over smoothness. They aren't hifi, they just happen to still work pretty well that way too, but could be better if designed for it. Something else to keep in mind: the sensitivity of the compression tweeters. A conical speaker (such as Synergy or SEOS or Geddes) doesn't peak up the sensitivity at the upper end of the tweeter frequencies (because it doesn't beam), so you only have the naturally rolled off response of the tweeter driver. So your tweeter sensitivity, after EQing it flat, is really only going to only be about 95dB. Maybe 99dB tops (if you squint while looking at it and also count the rolloff within the +/-3dB spec). So unless you make the speaker REALLY bass-hot, or have a heck of a sloped room curve, or use active crossover and multiple amplifiers, you don't want your woofers to be much hotter than 100dBSPL sensitivity. The compression chamber is the amount of air trapped between the front of the cone and the port hole. The "natural" compression chamber would be the volume (the 'slight concave portion') that would be inside the cone if you set a board over the speaker -- that's the easy way to figure out the volume, in fact, fill the cone with rice, flatten off the top, and then pour that rice into a measuring cup or measuring spoons. In the SmallSyns, I had to fill up the volume between the cone and the horn wall (since the wall wasn't flat) with putty to reduce the compression chamber volume. I think the compression chamber volume is NOT to include the volume in the port holes (at least that's the way it seems to work out on the few synergy horns I've modeled and built), but when you make frustrums the volume in the chamber and the ports isn't so clearly separated. I've estimated that a good frustruming reduces the port length by about half and about 70% of the frustrum volume counts for compression chamber. I scientifically determined that correction via a wild-ass guess.