I want to upgrade from my current speakers (jbl 2225h, some old altec theater HF horns with phenolics, not particularly big, and a jvc zero series ribbon tweeter), but all my research into DIY horn based speakers essentially leaves me with the thought that this is entirely too complicated or too expensive or both. Can't decide a horn/cd combo or how to integrate it.
I like my speakers a lot (they require a lot of dsp EQ), but I know they fall short... but they have certainly at least made me come to know I want the horn magic.
My 2225h's are playing much higher than they should be, since these horns are crossed around 1500hz based on the ear testing. The woofers are certainly sharing the output in that range, audibly.
So I want bigger better newer horns. I'm not too into the idea of vintage hifi, I know people love that stuff, but tech and design has moved on and there exist far better horns with less of the downsides than what existed 30-50 years ago. I'm open to it but I would prefer a slightly more modern CD design, and I'm not super picky about the coverage pattern, as the primary goal is maximum fidelity in the sweet spot... i can deal with other shortcomings. I would prefer a passive crossover but am open to active dsp crossover if necessary... I would need another amp. I've got roughly 125 clean watts/ch to work with.
Am open to any CD design and manufacturer... I don't care if the horn is cheap, as I can tweak or dampen the horn.
I have the 2225h's so I'm open to buying just horns, horns and compression drivers, or full speakers. Large round waveguides, oblate spheroids, any number of CD designs, I'm open to it all.
I'm just an amateur music loving hifi enthusiast at heart who has enjoyed my speakers for 15 years and is ready for something better, but is lacking in time and money. I would like a set of horns that can genuinely play down to 800hz or less. I'd love to keep it 2 way but it seems rare/difficult to be able to cover 500-800hz to even say 16khz well in a single package, and I don't want to tolerate a lot of distortion... if I can just EQ the highs in that's fine, but if it's going to sound bad or be way too beamy, then I'd rather cross at 8khz or higher to my ribbon tweeters which I love the sound of.
The JVC zero ribbons sound great... the slightly modified 2225H's in the custom cabinets sound fantastic... the problem is really my horn... I can't say the model from memory but as an older theater piece it's probably not really a hifi piece of kit, and it's not playing low enough. IMO it EQ's well enough but there are some problem areas... namely the lower crossover region, then the naturally hotter 3-4khz region, then maybe a little peak or distortion in the low to mid 7khz region. My 5 band parametric goes a long way to making the system sound reasonably balanced, but I know I'm just used to it and that there is greener grass somewhere. I'm constantly fiddling with the EQ from album to album out of necessity, when listening critically, especially at higher volumes. BTW I don't require high volumes in my listening environment, 85 or 90db is as high as I would listen, and normally less than that. With the cabinets and modpodged 2225h's (adds rigidity and weight - think mass ring of the 2235h, this isn't quite that, but it's moving in that direction) and EQ, these speakers play COMFORTABLY to 30hz, with authority. So I'm happy with the output of 2 15"s at 30hz, whatever amount of SPL that can produce, that's what I need to balance the mids/highs with, and I like a low bass heavy sound (think gradual boost below 120hz, major boost below 50ish to comp for not having a sub).
So... IDK if this forum is the right place to ask, if not please offer suggestions. I don't even know what to look for on ebay. But if anyone has a vague idea of what I'm looking for and has anything they'd be kind enough to sell me at a decent price, I won't haggle, either I can afford it or I can't... whether full speakers or just horns.... please let me know. If horns, depending on their size, I can always build new cabinets around them and throw in the 2225h's and if needed the ribbons. I wish I knew the crossover specs for my speakers and not just by an ear test, but I don't. They were custom built by some engineer.
Or, if anyone can just help give me direction on what to buy/build, that would also be appreciated. There's far too many options and recommendations out there... if someone is looking to put together some horns/waveguides using a compression driver... NOT expensive vintage hifi or esoteric beryllium stuff.... but under a grand total, preferably under 500 bucks... what do I get and where from? If the goal is 800hz and up as high as reasonably feasible, and the primary goal is smoothness and fidelity, and not SPL, what is readily available? Even though I've said I care mostly about the sweet spot, the reason I want CD in theory is for smooth and balanced room interaction... if dispersion isn't evenly controlled, then the energy/frequency response from reflections isn't going to sound natural and smooth, and won't lend itself to a wide soundstage, which I definitely want.
I like my speakers a lot (they require a lot of dsp EQ), but I know they fall short... but they have certainly at least made me come to know I want the horn magic.
My 2225h's are playing much higher than they should be, since these horns are crossed around 1500hz based on the ear testing. The woofers are certainly sharing the output in that range, audibly.
So I want bigger better newer horns. I'm not too into the idea of vintage hifi, I know people love that stuff, but tech and design has moved on and there exist far better horns with less of the downsides than what existed 30-50 years ago. I'm open to it but I would prefer a slightly more modern CD design, and I'm not super picky about the coverage pattern, as the primary goal is maximum fidelity in the sweet spot... i can deal with other shortcomings. I would prefer a passive crossover but am open to active dsp crossover if necessary... I would need another amp. I've got roughly 125 clean watts/ch to work with.
Am open to any CD design and manufacturer... I don't care if the horn is cheap, as I can tweak or dampen the horn.
I have the 2225h's so I'm open to buying just horns, horns and compression drivers, or full speakers. Large round waveguides, oblate spheroids, any number of CD designs, I'm open to it all.
I'm just an amateur music loving hifi enthusiast at heart who has enjoyed my speakers for 15 years and is ready for something better, but is lacking in time and money. I would like a set of horns that can genuinely play down to 800hz or less. I'd love to keep it 2 way but it seems rare/difficult to be able to cover 500-800hz to even say 16khz well in a single package, and I don't want to tolerate a lot of distortion... if I can just EQ the highs in that's fine, but if it's going to sound bad or be way too beamy, then I'd rather cross at 8khz or higher to my ribbon tweeters which I love the sound of.
The JVC zero ribbons sound great... the slightly modified 2225H's in the custom cabinets sound fantastic... the problem is really my horn... I can't say the model from memory but as an older theater piece it's probably not really a hifi piece of kit, and it's not playing low enough. IMO it EQ's well enough but there are some problem areas... namely the lower crossover region, then the naturally hotter 3-4khz region, then maybe a little peak or distortion in the low to mid 7khz region. My 5 band parametric goes a long way to making the system sound reasonably balanced, but I know I'm just used to it and that there is greener grass somewhere. I'm constantly fiddling with the EQ from album to album out of necessity, when listening critically, especially at higher volumes. BTW I don't require high volumes in my listening environment, 85 or 90db is as high as I would listen, and normally less than that. With the cabinets and modpodged 2225h's (adds rigidity and weight - think mass ring of the 2235h, this isn't quite that, but it's moving in that direction) and EQ, these speakers play COMFORTABLY to 30hz, with authority. So I'm happy with the output of 2 15"s at 30hz, whatever amount of SPL that can produce, that's what I need to balance the mids/highs with, and I like a low bass heavy sound (think gradual boost below 120hz, major boost below 50ish to comp for not having a sub).
So... IDK if this forum is the right place to ask, if not please offer suggestions. I don't even know what to look for on ebay. But if anyone has a vague idea of what I'm looking for and has anything they'd be kind enough to sell me at a decent price, I won't haggle, either I can afford it or I can't... whether full speakers or just horns.... please let me know. If horns, depending on their size, I can always build new cabinets around them and throw in the 2225h's and if needed the ribbons. I wish I knew the crossover specs for my speakers and not just by an ear test, but I don't. They were custom built by some engineer.
Or, if anyone can just help give me direction on what to buy/build, that would also be appreciated. There's far too many options and recommendations out there... if someone is looking to put together some horns/waveguides using a compression driver... NOT expensive vintage hifi or esoteric beryllium stuff.... but under a grand total, preferably under 500 bucks... what do I get and where from? If the goal is 800hz and up as high as reasonably feasible, and the primary goal is smoothness and fidelity, and not SPL, what is readily available? Even though I've said I care mostly about the sweet spot, the reason I want CD in theory is for smooth and balanced room interaction... if dispersion isn't evenly controlled, then the energy/frequency response from reflections isn't going to sound natural and smooth, and won't lend itself to a wide soundstage, which I definitely want.
Last edited:
If you like the 2235H, you could just recone the 2225H to 2235H and start over with the newer smooth-wall waveguide-kind of horns.
Hmm, with such a low budget, recommend DIYing a scaled up pair of the original 1.0 kHz Unity/Synergy conical WGs for a 500-800 Hz XO with whatever 1" CDs best meets your performance goals, budget and use passive CD horn EQ to flatten it out to whatever max HF you want or put a bullet horn tweeter in the horn @ ~ 3-4 kHz depending on the low end XO point and use digital TD to align it, though best to splurge on a pair of one of the new coax CDs.
I think there is something like Hofmann's Iron Law...except for DIY loudspeakers:
"Pick one: low cost...or low effort, because you can't have both."
Even Klipsch Belles, La Scalas, and Klipschorns are in the $2500-$3500 USD/pair range--forty to fifty years after they're new, and these aren't "collector's items" compared to similar vintage JBLs (fully horn loaded Paragon/Metregon) or Altec (A5 or A7, Model 19).
If I were in your shoes and had similar equipment, and knowing what I do now, I'd build a pair of full-range multiple-entry horns (MEHs), and re-use your present woofers in them from your current loudspeakers, then apply one miniDSP 2x4 HD (about $230 USD) to bi-amp them (i.e., adding one more stereo amplifier for the woofers) and buy the best quality 1.4" throat full-range compression drivers you can (usually starting at about $200 per driver) and dial them in using its DSP crossover filters and PEQs. You will also need something like a miniDSP UMIK-1 calibrated microphone (about $100) and Room EQ Wizard (a.k.a., "REW", which is shareware) to plug into your computer and a way to connect your DAC or disc player to your computer (i.e., HDMI, HDMI, or S/PDIF) so you can dial-in the drivers using the 2x4 HD.
Chris
"Pick one: low cost...or low effort, because you can't have both."
Even Klipsch Belles, La Scalas, and Klipschorns are in the $2500-$3500 USD/pair range--forty to fifty years after they're new, and these aren't "collector's items" compared to similar vintage JBLs (fully horn loaded Paragon/Metregon) or Altec (A5 or A7, Model 19).
If I were in your shoes and had similar equipment, and knowing what I do now, I'd build a pair of full-range multiple-entry horns (MEHs), and re-use your present woofers in them from your current loudspeakers, then apply one miniDSP 2x4 HD (about $230 USD) to bi-amp them (i.e., adding one more stereo amplifier for the woofers) and buy the best quality 1.4" throat full-range compression drivers you can (usually starting at about $200 per driver) and dial them in using its DSP crossover filters and PEQs. You will also need something like a miniDSP UMIK-1 calibrated microphone (about $100) and Room EQ Wizard (a.k.a., "REW", which is shareware) to plug into your computer and a way to connect your DAC or disc player to your computer (i.e., HDMI, HDMI, or S/PDIF) so you can dial-in the drivers using the 2x4 HD.
Chris
The spreadsheet enclosed above of Bill Waslo's Synergy Calc template will fit the bill for a two-way full-range MEH that will amaze you with its sound quality. I'd use your present woofers in a one-woofer-per-MEH configuration.
Chris
Chris
The eWave project started out as a low cost alternative and may be useful looking into it.
Hi Lee:
Fifteen years ago I built a pair of speakers that I still really like. Started out listening to a friends 4 pi's but wanted them with upgraded mid/highs. I built basically the 4 pi's but with Radian 475 8 ohm compression drivers which are better than any similar speakers I've heard. Bi-amped them with two different tube amps. The Radians can be crossed at 1,000Hz LR4 and are on sale right now at Parts Express for $200 each. The horns are available from Pi speakers and are not expensive. The 2225 drivers are the same as the the high end 4 pi design uses. With miniDSP for equalization you should be able to get precisely what you want. If you don't have measuring capability, the miniDSP mic system is still within your budget if you go with the Radians. Just a suggestion that may be helpful. Wayne at Pi speakers was very helpful to me years ago as I put this system together.
Fifteen years ago I built a pair of speakers that I still really like. Started out listening to a friends 4 pi's but wanted them with upgraded mid/highs. I built basically the 4 pi's but with Radian 475 8 ohm compression drivers which are better than any similar speakers I've heard. Bi-amped them with two different tube amps. The Radians can be crossed at 1,000Hz LR4 and are on sale right now at Parts Express for $200 each. The horns are available from Pi speakers and are not expensive. The 2225 drivers are the same as the the high end 4 pi design uses. With miniDSP for equalization you should be able to get precisely what you want. If you don't have measuring capability, the miniDSP mic system is still within your budget if you go with the Radians. Just a suggestion that may be helpful. Wayne at Pi speakers was very helpful to me years ago as I put this system together.
Thanks for the recommendation, I'll look into the unity/synergy setup you're referring to, I haven't heard of it.
And omg, dude, one of the new Eminence or Celestion (maybe the B&C option is good too) coax CD's is like my dream speaker, but the cost is just too high. Not only are those CD's super expensive, but to do them justice and mate them with the right horn, I haven't seen any horns/waveguides designed for them that aren't ridiculously expensive too. I mean look at the price for the ME464 from b&c on parts express.... over 700 bucks for a horn? That's absurd considering material cost, I can't justify it.
I have been interested in conical waveguides, as my understanding is, according to some anyway, that being round, they could represent the ideal horn from a fidelity perspective, although that would depend on the flare of course too, so if you don't care about controlling vertical dispersion differently than horizontal it could be a great option. Parts express only has an 8" these days which is too small, but I've seen 10, 12, 15, maybe 18" conical waveguides on ebay I think.
Besides partsexpress or ebay, what good sites/sources exist for buying diy audio gear... horns, CD's, woofers, x-over parts?
Thanks Chris, I hadn't considered multiple entry horns to cover a wider range, but overall you're pretty spot on and getting where I'm coming from. It's not that I couldn't double the budget if I do this over time, but I was just kind of hoping to be able to gradually piece something together in a way that I get to experience some of it now and some later... i.e. perhaps I could get horns/CD's within that price range and build from there, that or get lucky with someone wanting to sell me something of theirs. But I like your thoughts. I'm aware of the mini DSP stuff and REW and I know I need a measurement mic to do things right.
And for the record my ONLY source is and will be a DAC or preamp fed by my computer. Currently I run a Khadas tone board which has ample output and so I don't even use a preamp, though I am curious about what a buffered or active preamp might do to the sound, like if there could be some benefit to better matching the output impedance to the input impedance of my amplifier, idk. The amp I run is a connex electronics TA3020 v3c i think with an SMPS, with all the upgrades and a high supply voltage. The amp and DAC are incredible (and I've run them on multiple speakers), I know my first upgrade and fun project needs to be speakers, I'm quite happy with the amp and source for the time being, though depending on what I end up with speakers wise, I might need either a single higher power amplifier, or another amplifier for bi-amping. I would go TA0105ARB or another TA3020, or an irs2092 or AEM6000 based amplifier from Connex again, as I'm very happy with his products and pricing.
And for the record my ONLY source is and will be a DAC or preamp fed by my computer. Currently I run a Khadas tone board which has ample output and so I don't even use a preamp, though I am curious about what a buffered or active preamp might do to the sound, like if there could be some benefit to better matching the output impedance to the input impedance of my amplifier, idk. The amp I run is a connex electronics TA3020 v3c i think with an SMPS, with all the upgrades and a high supply voltage. The amp and DAC are incredible (and I've run them on multiple speakers), I know my first upgrade and fun project needs to be speakers, I'm quite happy with the amp and source for the time being, though depending on what I end up with speakers wise, I might need either a single higher power amplifier, or another amplifier for bi-amping. I would go TA0105ARB or another TA3020, or an irs2092 or AEM6000 based amplifier from Connex again, as I'm very happy with his products and pricing.
Potentially, I suppose. I've seen the econowave thing. I guess my concerns are: HF crossover point still too high for what I hope to achieve, and the design is not able to go to 30hz. I would have to redesign it at which point I'm on my own again and the existing DIY design will no longer help... once I change the drivers, or make a bigger cabinet, etc., all the values associated with the project/kit/design or whatever will be useless to me, it's just starting from scratch. I want to tune the 15" woofers to around 30hz, and hope to achieve a crossover under 1khz, the lower the better.
I'm even considering a 3-way design now with 18" "subs" if you will, and a smaller pro driver above that... that might open up possibilities. The 18 could cover the bottom 2 octaves let's say, then the smaller driver could cover roughly 80hz to... depending on the size, either 400-800hz, or if it was a 10" for example maybe I'd be comfortable letting it play to 1.2-1.6khz... that would certainly make horn/CD selection easier.
Wow, thank you! This might be the most helpful direction I've received so far! Can you give me some more details or be more specific? I remember hearing about the 4 pi's in the past actually, but I'd never looked at the website. I've pulled it up but i don't know exactly what to look for.
This sounds like a very viable option. I know you like your Radians, but given 15 years have passed, are there any CD's you know of (maybe ones other builders have used in that project or with those horns) that are either better OR as good but cheaper (such as the DE250 BMS4550 or D220Ti for example, or maybe an Eminence offering)? Not that I'm opposed to the Radian's, the reviews look incredible, I'm just curious.
And do you have any thoughts on the B&C ME45 horn? It seems like it might be perfect and it's affordable.
Also, I see some crossing at 1.2k with a 2nd order, some at 1k with a 3rd order or active crossover... given my intentions of staying under 100db (really under 90 most of the time), do you think my lower power requirements open up crossover point and slope options a little bit?
15+Horn/CD = Asathor a 4367 a-like. A search will bring up the thread.
I've used 2225/6 (and AE TD15X/S) in 3 ways with a variety of 8&10s crossing to a WG/CD at about 1k5 to 1k7 for about 25y and am looking at a 2225, 6MD38 and STH100+CD for some bedroom speakers as I have almost all the parts, just not enough time to do them ATM.
2 ways require a lot of horn, and to my experience, so far, too many compromises.
I've used 2225/6 (and AE TD15X/S) in 3 ways with a variety of 8&10s crossing to a WG/CD at about 1k5 to 1k7 for about 25y and am looking at a 2225, 6MD38 and STH100+CD for some bedroom speakers as I have almost all the parts, just not enough time to do them ATM.
2 ways require a lot of horn, and to my experience, so far, too many compromises.
Thanks for sharing your experience. I agree a good 2-way seems difficult and/or expensive to achieve. But when you mention 3-way, and i see the mention of "8&10s", are you talking about a horn tweeter with an 8 or 10" mid-bass/mid-range cone to cover 1.5-1.7k above the woofer? I will look up the Asathor, which sounds like a 2-way design. I guess I'm wondering, even if we assume a 2-way is difficult and likely to contain compromises, and a 3-way is superior... why would you lean towards a woofer, a mid-bass cone driver, and a horn on top, as opposed to a woofer, a mid-range horn, and a tweeter on top (horn tweeter / ribbon / amt / dome+WG... whatever).
I'm open to either kind of design and would have to experience them IRL to know with certainty I suppose, but my assumption is always that getting the horn to cover as low as possible is more ideal, to cover the vast majority of the vocal range especially, and to offer extreme detail, resolution, and dynamics through the mid-range for lifelike instruments as well. Above 6 or 8khz (maybe even 4-5khz) I'm not nearly as concerned about the extreme benefits of a horn in the more sensitive register, and feel the potentially wider dispersion of having this separate "super" tweeter (and potentially lower distortion, depending on the horn geometry) could suffice or even be superior... or of course you could go with a smaller horn tweeter for the last 1-2 octaves. As it stands my set has the ribbon tweeter, which is crossed at ~8-9khz. Test tones appear to produce a roughly similar output at around 8.5k (between my horns and ribbons) I would say, though the dispersion characteristics are so different it's perhaps a little hard to tell when the energy sent to each is equal, vs. the perceived sound on-axis... because the exponential horns I have a certainly beaming a lot at 8+khz, the ribbon less so. I believe there's a 3rd order on the ribbon, and only a 2nd order on the horn, so the horn gradually fades out while the ribbon cuts off fairly sharply. The bottom of the horn I believe also has a 2nd order, and the woofer I suspect has a 1st order, but I don't know/can't recall for sure. I want to avoid this situation too, the horn beaming, which is why I'm more interested in CD/waveguide designs than exponential or other designs that beam up high... either that or I'd want to cross over lower to a dedicated tweeter of whatever kind. Beaming makes EQ'ing difficult, and if the implementation or listening position isn't perfect, having both drivers playing at that upper x-over region leads to lobing/comb filtering.
I like Geddes approach (and I wish I could buy a huge oblate spheroid horn like he designed), and I like Erin's Audio Corner (the YT channel) that looks at klippel nearfield scanner data. I don't expect to be able to build an ideal speaker, but in theory I am interested in objective performance data over subjective anecdotes. A consistent or at least smoothly transitioning coverage pattern, phase coherency, etc. I have no experience to tell me I should be overly concerned with time arrival for its own sake, phase coherency for a smooth crossover region etc. is important but absolute time arrival I suspect is over-hyped by rich audiophiles as I doubt we can perceive it much if at all. But jumping from a 10 degree coverage to 50 degree or vice versa... yeah that's going to affect the sound, either due to off-axis listening, or just because of how it loads the room and the in room frequency response and sound-stage is affected. This is why I would prefer to avoid having a 15" play to 1.5khz, because the coverage narrows a lot by that frequency. It might be ok in the fairly narrow sweet spot but it's not ideal, and doesn't load the room evenly, again leading to a situation that EQ can't completely fix. But an 8 or 10 could be just fine, and perhaps a pro driver of that size is fast and nimble enough to sound detailed and dynamic enough to my ears, even relative to a horn... as compared to my 2225h's playing that high, where the output does not come across as detailed and precise compared to where it transitions to the horn.... there's a mismatch in resolution of voices and instruments that span across both ranges that I want to attempt to eliminate as much as possible.
I will say though, the 2225h sounds amazing overall, for what it's designed for. It's incredibly smooth and balanced and capable... tight and punchy and controlled. My TA3020 based amp has a high damping factor which I think helps.... the kick drum impacts are just so tight and forceful, hence my continued interest in a speaker based on pro drivers with light cones and stiff suspensions and high sensitivities.
I'm interested, in a future build, in making a 3 or maybe even 4 way design that utilizes an 18" pro driver (purely sub duty with a separate if it was a 4-way, maybe crossed higher in a 3-way, but not over 400-500hz). And I'm open to replacing the JBL's with a modern pro driver with higher x-max (and potentially even higher sensitivity), though I know the 2225/2235 is a special and revered driver for good reason.
Since you've used 2225s and 2226s what's your opinion on the differences? From what I can gather, the 2225 is almost universally considered superior in most hi-fi aspects, with the 2226 having the design advantage of higher power handling and maybe greater impact above say 60-100hz, but not as "good" sounding (smooth/natural?). Maybe it plays higher too, I don't recall, though the dispersion isn't going to be any better at the relatively high frequencies they're typically crossed at in PA applications.
Question, if you were filling in mid-bass duty with an 8/10", what driver(s) would you suggest? Or would you say the 6.5" 6MD38 would be superior for that application? How low would you cross it, and can it keep up dynamically/SPL wise with the 15"? It would have wider dispersion at your x-over point. Have you built this design before, and have you compared it to a horn that can play down to say 400 or even 800hz compared to the 1.5-1.7k x-over of your design, and if so how would you describe the character of the midrange from ~500hz to ~1500hz coming from a cone vs a horn?
I'm open to either kind of design and would have to experience them IRL to know with certainty I suppose, but my assumption is always that getting the horn to cover as low as possible is more ideal, to cover the vast majority of the vocal range especially, and to offer extreme detail, resolution, and dynamics through the mid-range for lifelike instruments as well. Above 6 or 8khz (maybe even 4-5khz) I'm not nearly as concerned about the extreme benefits of a horn in the more sensitive register, and feel the potentially wider dispersion of having this separate "super" tweeter (and potentially lower distortion, depending on the horn geometry) could suffice or even be superior... or of course you could go with a smaller horn tweeter for the last 1-2 octaves. As it stands my set has the ribbon tweeter, which is crossed at ~8-9khz. Test tones appear to produce a roughly similar output at around 8.5k (between my horns and ribbons) I would say, though the dispersion characteristics are so different it's perhaps a little hard to tell when the energy sent to each is equal, vs. the perceived sound on-axis... because the exponential horns I have a certainly beaming a lot at 8+khz, the ribbon less so. I believe there's a 3rd order on the ribbon, and only a 2nd order on the horn, so the horn gradually fades out while the ribbon cuts off fairly sharply. The bottom of the horn I believe also has a 2nd order, and the woofer I suspect has a 1st order, but I don't know/can't recall for sure. I want to avoid this situation too, the horn beaming, which is why I'm more interested in CD/waveguide designs than exponential or other designs that beam up high... either that or I'd want to cross over lower to a dedicated tweeter of whatever kind. Beaming makes EQ'ing difficult, and if the implementation or listening position isn't perfect, having both drivers playing at that upper x-over region leads to lobing/comb filtering.
I like Geddes approach (and I wish I could buy a huge oblate spheroid horn like he designed), and I like Erin's Audio Corner (the YT channel) that looks at klippel nearfield scanner data. I don't expect to be able to build an ideal speaker, but in theory I am interested in objective performance data over subjective anecdotes. A consistent or at least smoothly transitioning coverage pattern, phase coherency, etc. I have no experience to tell me I should be overly concerned with time arrival for its own sake, phase coherency for a smooth crossover region etc. is important but absolute time arrival I suspect is over-hyped by rich audiophiles as I doubt we can perceive it much if at all. But jumping from a 10 degree coverage to 50 degree or vice versa... yeah that's going to affect the sound, either due to off-axis listening, or just because of how it loads the room and the in room frequency response and sound-stage is affected. This is why I would prefer to avoid having a 15" play to 1.5khz, because the coverage narrows a lot by that frequency. It might be ok in the fairly narrow sweet spot but it's not ideal, and doesn't load the room evenly, again leading to a situation that EQ can't completely fix. But an 8 or 10 could be just fine, and perhaps a pro driver of that size is fast and nimble enough to sound detailed and dynamic enough to my ears, even relative to a horn... as compared to my 2225h's playing that high, where the output does not come across as detailed and precise compared to where it transitions to the horn.... there's a mismatch in resolution of voices and instruments that span across both ranges that I want to attempt to eliminate as much as possible.
I will say though, the 2225h sounds amazing overall, for what it's designed for. It's incredibly smooth and balanced and capable... tight and punchy and controlled. My TA3020 based amp has a high damping factor which I think helps.... the kick drum impacts are just so tight and forceful, hence my continued interest in a speaker based on pro drivers with light cones and stiff suspensions and high sensitivities.
I'm interested, in a future build, in making a 3 or maybe even 4 way design that utilizes an 18" pro driver (purely sub duty with a separate if it was a 4-way, maybe crossed higher in a 3-way, but not over 400-500hz). And I'm open to replacing the JBL's with a modern pro driver with higher x-max (and potentially even higher sensitivity), though I know the 2225/2235 is a special and revered driver for good reason.
Since you've used 2225s and 2226s what's your opinion on the differences? From what I can gather, the 2225 is almost universally considered superior in most hi-fi aspects, with the 2226 having the design advantage of higher power handling and maybe greater impact above say 60-100hz, but not as "good" sounding (smooth/natural?). Maybe it plays higher too, I don't recall, though the dispersion isn't going to be any better at the relatively high frequencies they're typically crossed at in PA applications.
Question, if you were filling in mid-bass duty with an 8/10", what driver(s) would you suggest? Or would you say the 6.5" 6MD38 would be superior for that application? How low would you cross it, and can it keep up dynamically/SPL wise with the 15"? It would have wider dispersion at your x-over point. Have you built this design before, and have you compared it to a horn that can play down to say 400 or even 800hz compared to the 1.5-1.7k x-over of your design, and if so how would you describe the character of the midrange from ~500hz to ~1500hz coming from a cone vs a horn?
Last edited:
EDIT: Ignore above post
Thanks for sharing your experience. I agree a good 2-way seems difficult and/or expensive to achieve. But when you mention 3-way, and i see the mention of "8&10s", are you talking about a horn tweeter with an 8 or 10" mid-bass/mid-range cone to cover 1.5-1.7k above the woofer? I will look up the Asathor, which sounds like a 2-way design. I guess I'm wondering, even if we assume a 2-way is difficult and likely to contain compromises, and a 3-way is superior... why would you lean towards a woofer, a mid-bass cone driver, and a horn on top, as opposed to a woofer, a mid-range horn, and a tweeter on top (horn tweeter / ribbon / amt / dome+WG... whatever).
I'm open to either kind of design and would have to experience them IRL to know with certainty I suppose, but my assumption is always that getting the horn to cover as low as possible is more ideal, to cover the vast majority of the vocal range especially, and to offer extreme detail, resolution, and dynamics through the mid-range for lifelike instruments as well. Above 6 or 8khz (maybe even 4-5khz) I'm not nearly as concerned about the extreme benefits of a horn in the more sensitive register, and feel the potentially wider dispersion of having this separate "super" tweeter (and potentially lower distortion, depending on the horn geometry) could suffice or even be superior... or of course you could go with a smaller horn tweeter for the last 1-2 octaves. As it stands my set has the ribbon tweeter, which is crossed at ~8-9khz. Test tones appear to produce a roughly similar output at around 8.5k (between my horns and ribbons) I would say, though the dispersion characteristics are so different it's perhaps a little hard to tell when the energy sent to each is equal, vs. the perceived sound on-axis... because the exponential horns I have a certainly beaming a lot at 8+khz, the ribbon less so. I believe there's a 3rd order on the ribbon, and only a 2nd order on the horn, so the horn gradually fades out while the ribbon cuts off fairly sharply. The bottom of the horn I believe also has a 2nd order, and the woofer I suspect has a 1st order, but I don't know/can't recall for sure. I want to avoid this situation too, the horn beaming, which is why I'm more interested in CD/waveguide designs than exponential or other designs that beam up high... either that or I'd want to cross over lower to a dedicated tweeter of whatever kind. Beaming makes EQ'ing difficult, and if the implementation or listening position isn't perfect, having both drivers playing at that upper x-over region leads to lobing/comb filtering.
I like Geddes approach (and I wish I could buy a huge oblate spheroid horn like he designed), and I like Erin's Audio Corner (the YT channel) that looks at klippel nearfield scanner data. I don't expect to be able to build an ideal speaker, but in theory I am interested in objective performance data over subjective anecdotes. A consistent or at least smoothly transitioning coverage pattern, phase coherency, etc. I have no experience to tell me I should be overly concerned with time arrival for its own sake, phase coherency for a smooth crossover region etc. is important but absolute time arrival I suspect is over-hyped by rich audiophiles as I doubt we can perceive it much if at all. But jumping from a 10 degree coverage to 50 degree or vice versa... yeah that's going to affect the sound, either due to off-axis listening, or just because of how it loads the room and the in room frequency response and sound-stage is affected. This is why I would prefer to avoid having a 15" play to 1.5khz, because the coverage narrows a lot by that frequency. It might be ok in the fairly narrow sweet spot but it's not ideal, and doesn't load the room evenly, again leading to a situation that EQ can't completely fix. But an 8 or 10 could be just fine, and perhaps a pro driver of that size is fast and nimble enough to sound detailed and dynamic enough to my ears, even relative to a horn... as compared to my 2225h's playing that high, where the output does not come across as detailed and precise compared to where it transitions to the horn.... there's a mismatch in resolution of voices and instruments that span across both ranges that I want to attempt to eliminate as much as possible.
I will say though, the 2225h sounds amazing overall, for what it's designed for. It's incredibly smooth and balanced and capable... tight and punchy and controlled. My TA3020 based amp has a high damping factor which I think helps.... the kick drum impacts are just so tight and forceful, hence my continued interest in a speaker based on pro drivers with light cones and stiff suspensions and high sensitivities. In my speakers it's covering at least 5.5 octaves, which is the majority of the spectrum, and even though that's too high, it's just amazing that it can cover as wide a range as it does with such a naturalness and smoothness to it, and detailed given it's size. It certainly helps with accuracy that a large driver hardly has to move a millimeter in all but it's lowest octave to achieve good output levels.
I've tripped my amplifier's power supply's protection circuitry before with loud heavy deep bass, and it's rated at 300w. This was around 95db in room. I believe that's actually the limiting factor of my amplifier (considering the inefficiencies in both the SMPS and amplifier, this might be a "real" 250w of total output let's say), because I hear zero distortion in the bass before the protection circuit trips. Based on the specs of my amp and the input voltage and power supply selected, it should be capable of ~125 "clean" watts as I put it... and my standard for saying that is actually ULTRA clean watts, not 10% or even 1% distortion many amps are rated at, but .1% or less. Though with class T/D designs typically distortion stays very low until it doesn't... then it ramps quickly. So the amp should be capable of 205w+/ch at .1% thd into 8 ohms, but at more realistic levels the amp does .02% distortion or less, given my supply voltage, but I don't have a big enough power supply to hit that. I intend to rectify that one way or another too, either with a new smps, or a new amp, or an additional higher power amp just for the woofers if I use DSP x-over.
I'm interested, in a future build, in making a 3 or maybe even 4 way design that utilizes an 18" pro driver (purely sub duty with a separate if it was a 4-way, maybe crossed higher in a 3-way, but not over 400-500hz). And I'm open to replacing the JBL's with a modern pro driver with higher x-max (and potentially even higher sensitivity), though I know the 2225/2235 is a special and revered driver for good reason.
Since you've used 2225s and 2226s what's your opinion on the differences? From what I can gather, the 2225 is almost universally considered superior in most hi-fi aspects, with the 2226 having the design advantage of higher power handling and maybe greater impact above say 60-100hz, but not as "good" sounding (smooth/natural?). Maybe it plays higher too, I don't recall, though the dispersion isn't going to be any better at the relatively high frequencies they're typically crossed at in PA applications.
Question, if you were filling in mid-bass duty with an 8/10", what driver(s) would you suggest? Or would you say the 6.5" 6MD38 would be superior for that application? How low would you cross it, and can it keep up dynamically/SPL wise with the 15"? It would have wider dispersion at your x-over point. Have you built this design before, and have you compared it to a horn that can play down to say 400 or even 800hz compared to the 1.5-1.7k x-over of your design, and if so how would you describe the character of the midrange from ~500hz to ~1500hz coming from a cone vs a horn?
Amp specs for those curious, as I find it interesting:
Specs/Data: https://connexelectronic.com/wp-content/uploads/2017/08/TA3020_Audio_Amplifier_Module_v3c.pdf
Amp module: https://connexelectronic.com/product/ta3020-audio-amplifier-v3c/
The amp has both optional upgrades, and upon customer request the SMPS output voltage is set before shipment. With the upgraded mosfets the dead-time is set to 80ns vs the standard 120ns, and the upgraded capacitors (18000uf ELNA) ensure plenty of juice for transients.
They no longer sell the exact SMPS I have. Highly recommended for anyone interested in a budget hi-fi high output amplifier, though I'd go with the more recent updated v3d version. Mine doesn't have a fan as the heatsink is in open air, and temperature has never been an issue.
All in, for the amp (150) and a 500w or 800w SMPS, and all the upgrades, you're only looking at around 300 dollars, minus whatever you put into a case and connectors/switches, that's not bad for ~200w/ch into 8 ohms and 400w/ch into 4 ohms, or 790/1520w (8ohm/4ohm) in bridged mode as a subwoofer amp, considering the truly audiophile sound quality once it's broken in. For approximately double the power or more, and possibly superior performance in general, they have the TA0105ARB for the same price and an extra 50 bucks for the 2000w SMPS. Very easy to put together. They also have high power+high performance class D and class A/B designs at similarly low prices if you prefer, as well as linear power supplies. www.connexelectronic.com Very good documentation, and communication/help from the seller. He at least used to be a regular on some diy/hifi forums, not sure if he still is, but my experience with him and this product is very good, if you're looking for an alternative to the popular hypex modules.
Thanks for sharing your experience. I agree a good 2-way seems difficult and/or expensive to achieve. But when you mention 3-way, and i see the mention of "8&10s", are you talking about a horn tweeter with an 8 or 10" mid-bass/mid-range cone to cover 1.5-1.7k above the woofer? I will look up the Asathor, which sounds like a 2-way design. I guess I'm wondering, even if we assume a 2-way is difficult and likely to contain compromises, and a 3-way is superior... why would you lean towards a woofer, a mid-bass cone driver, and a horn on top, as opposed to a woofer, a mid-range horn, and a tweeter on top (horn tweeter / ribbon / amt / dome+WG... whatever).
I'm open to either kind of design and would have to experience them IRL to know with certainty I suppose, but my assumption is always that getting the horn to cover as low as possible is more ideal, to cover the vast majority of the vocal range especially, and to offer extreme detail, resolution, and dynamics through the mid-range for lifelike instruments as well. Above 6 or 8khz (maybe even 4-5khz) I'm not nearly as concerned about the extreme benefits of a horn in the more sensitive register, and feel the potentially wider dispersion of having this separate "super" tweeter (and potentially lower distortion, depending on the horn geometry) could suffice or even be superior... or of course you could go with a smaller horn tweeter for the last 1-2 octaves. As it stands my set has the ribbon tweeter, which is crossed at ~8-9khz. Test tones appear to produce a roughly similar output at around 8.5k (between my horns and ribbons) I would say, though the dispersion characteristics are so different it's perhaps a little hard to tell when the energy sent to each is equal, vs. the perceived sound on-axis... because the exponential horns I have a certainly beaming a lot at 8+khz, the ribbon less so. I believe there's a 3rd order on the ribbon, and only a 2nd order on the horn, so the horn gradually fades out while the ribbon cuts off fairly sharply. The bottom of the horn I believe also has a 2nd order, and the woofer I suspect has a 1st order, but I don't know/can't recall for sure. I want to avoid this situation too, the horn beaming, which is why I'm more interested in CD/waveguide designs than exponential or other designs that beam up high... either that or I'd want to cross over lower to a dedicated tweeter of whatever kind. Beaming makes EQ'ing difficult, and if the implementation or listening position isn't perfect, having both drivers playing at that upper x-over region leads to lobing/comb filtering.
I like Geddes approach (and I wish I could buy a huge oblate spheroid horn like he designed), and I like Erin's Audio Corner (the YT channel) that looks at klippel nearfield scanner data. I don't expect to be able to build an ideal speaker, but in theory I am interested in objective performance data over subjective anecdotes. A consistent or at least smoothly transitioning coverage pattern, phase coherency, etc. I have no experience to tell me I should be overly concerned with time arrival for its own sake, phase coherency for a smooth crossover region etc. is important but absolute time arrival I suspect is over-hyped by rich audiophiles as I doubt we can perceive it much if at all. But jumping from a 10 degree coverage to 50 degree or vice versa... yeah that's going to affect the sound, either due to off-axis listening, or just because of how it loads the room and the in room frequency response and sound-stage is affected. This is why I would prefer to avoid having a 15" play to 1.5khz, because the coverage narrows a lot by that frequency. It might be ok in the fairly narrow sweet spot but it's not ideal, and doesn't load the room evenly, again leading to a situation that EQ can't completely fix. But an 8 or 10 could be just fine, and perhaps a pro driver of that size is fast and nimble enough to sound detailed and dynamic enough to my ears, even relative to a horn... as compared to my 2225h's playing that high, where the output does not come across as detailed and precise compared to where it transitions to the horn.... there's a mismatch in resolution of voices and instruments that span across both ranges that I want to attempt to eliminate as much as possible.
I will say though, the 2225h sounds amazing overall, for what it's designed for. It's incredibly smooth and balanced and capable... tight and punchy and controlled. My TA3020 based amp has a high damping factor which I think helps.... the kick drum impacts are just so tight and forceful, hence my continued interest in a speaker based on pro drivers with light cones and stiff suspensions and high sensitivities. In my speakers it's covering at least 5.5 octaves, which is the majority of the spectrum, and even though that's too high, it's just amazing that it can cover as wide a range as it does with such a naturalness and smoothness to it, and detailed given it's size. It certainly helps with accuracy that a large driver hardly has to move a millimeter in all but it's lowest octave to achieve good output levels.
I've tripped my amplifier's power supply's protection circuitry before with loud heavy deep bass, and it's rated at 300w. This was around 95db in room. I believe that's actually the limiting factor of my amplifier (considering the inefficiencies in both the SMPS and amplifier, this might be a "real" 250w of total output let's say), because I hear zero distortion in the bass before the protection circuit trips. Based on the specs of my amp and the input voltage and power supply selected, it should be capable of ~125 "clean" watts as I put it... and my standard for saying that is actually ULTRA clean watts, not 10% or even 1% distortion many amps are rated at, but .1% or less. Though with class T/D designs typically distortion stays very low until it doesn't... then it ramps quickly. So the amp should be capable of 205w+/ch at .1% thd into 8 ohms, but at more realistic levels the amp does .02% distortion or less, given my supply voltage, but I don't have a big enough power supply to hit that. I intend to rectify that one way or another too, either with a new smps, or a new amp, or an additional higher power amp just for the woofers if I use DSP x-over.
I'm interested, in a future build, in making a 3 or maybe even 4 way design that utilizes an 18" pro driver (purely sub duty with a separate if it was a 4-way, maybe crossed higher in a 3-way, but not over 400-500hz). And I'm open to replacing the JBL's with a modern pro driver with higher x-max (and potentially even higher sensitivity), though I know the 2225/2235 is a special and revered driver for good reason.
Since you've used 2225s and 2226s what's your opinion on the differences? From what I can gather, the 2225 is almost universally considered superior in most hi-fi aspects, with the 2226 having the design advantage of higher power handling and maybe greater impact above say 60-100hz, but not as "good" sounding (smooth/natural?). Maybe it plays higher too, I don't recall, though the dispersion isn't going to be any better at the relatively high frequencies they're typically crossed at in PA applications.
Question, if you were filling in mid-bass duty with an 8/10", what driver(s) would you suggest? Or would you say the 6.5" 6MD38 would be superior for that application? How low would you cross it, and can it keep up dynamically/SPL wise with the 15"? It would have wider dispersion at your x-over point. Have you built this design before, and have you compared it to a horn that can play down to say 400 or even 800hz compared to the 1.5-1.7k x-over of your design, and if so how would you describe the character of the midrange from ~500hz to ~1500hz coming from a cone vs a horn?
Amp specs for those curious, as I find it interesting:
Specs/Data: https://connexelectronic.com/wp-content/uploads/2017/08/TA3020_Audio_Amplifier_Module_v3c.pdf
Amp module: https://connexelectronic.com/product/ta3020-audio-amplifier-v3c/
The amp has both optional upgrades, and upon customer request the SMPS output voltage is set before shipment. With the upgraded mosfets the dead-time is set to 80ns vs the standard 120ns, and the upgraded capacitors (18000uf ELNA) ensure plenty of juice for transients.
They no longer sell the exact SMPS I have. Highly recommended for anyone interested in a budget hi-fi high output amplifier, though I'd go with the more recent updated v3d version. Mine doesn't have a fan as the heatsink is in open air, and temperature has never been an issue.
All in, for the amp (150) and a 500w or 800w SMPS, and all the upgrades, you're only looking at around 300 dollars, minus whatever you put into a case and connectors/switches, that's not bad for ~200w/ch into 8 ohms and 400w/ch into 4 ohms, or 790/1520w (8ohm/4ohm) in bridged mode as a subwoofer amp, considering the truly audiophile sound quality once it's broken in. For approximately double the power or more, and possibly superior performance in general, they have the TA0105ARB for the same price and an extra 50 bucks for the 2000w SMPS. Very easy to put together. They also have high power+high performance class D and class A/B designs at similarly low prices if you prefer, as well as linear power supplies. www.connexelectronic.com Very good documentation, and communication/help from the seller. He at least used to be a regular on some diy/hifi forums, not sure if he still is, but my experience with him and this product is very good, if you're looking for an alternative to the popular hypex modules.
Hi Lee:
The horn for the 4 Pi speakers is here: https://www.pispeakers.com/catalog/...ucts_id=276&osCsid=27n1vs56e129b16h3bec47tbf5
A real bargain at $50 each. Send an email to Wayne asking for the design package for the 4 Pi's. If there is not a lot a diffenence between what you have for your JBL's and the 4Pi just build a simple box for the horns to set on top or if you are lucky, incorporate the driver/horn into your current cabinet and start measuring. The tutorials on the miniDSP site will get you started on a new crossover. The bottom end miniDSP is all you need. Buy it and the mic if you don't have a measuring system and start listening, measuring and tweaking until you get what you want. I am not current on any newer drivers so you are on your own with that. Have fun.
The horn for the 4 Pi speakers is here: https://www.pispeakers.com/catalog/...ucts_id=276&osCsid=27n1vs56e129b16h3bec47tbf5
A real bargain at $50 each. Send an email to Wayne asking for the design package for the 4 Pi's. If there is not a lot a diffenence between what you have for your JBL's and the 4Pi just build a simple box for the horns to set on top or if you are lucky, incorporate the driver/horn into your current cabinet and start measuring. The tutorials on the miniDSP site will get you started on a new crossover. The bottom end miniDSP is all you need. Buy it and the mic if you don't have a measuring system and start listening, measuring and tweaking until you get what you want. I am not current on any newer drivers so you are on your own with that. Have fun.
Since those are designed for open air concerts, and you know how much tickets cost nowadays...
Thanks, I've sent him an email with some questions. His product listings look quite intriguing. I do suspect by default the bass extension might possibly be lacking with his box design, but I've asked him about that. It's interesting he doesn't offer the Radian's you (and others) claim are superior CD's in those horns.
I do have enough space to incorporate those horns into my baffles if I wanted to. If I do any modifications to my cabinets, that would also include removing the outer frame protrusion... the front baffles are set back like 2" from a rounded decorative frame of sorts that goes around the whole baffle, which is also what holds the grills in place via simple press-fit. It looks nice but I suspect it could cause diffraction issues. My cabinets are pretty large in volume compared to a lot of 15" 2-way designs similar to the 4-pi or some JBL's/Klipsch, offering a lower tuning frequency hence the effortless and powerful 30hz, though I don't know the exact tuning frequency I've never figured exact volume numbers or port length/diameter; they're rear-ported, stained birch ply on casters, but I have them set up on two end tables that are a little over 2 feet tall, and the casters are all located atop special vibration isolation pads... this height offers far superior soundstaging and imaging and matches much better with my TV height, such that the perfect phantom center aligns where it should, and it sounds like a huge wall of sound. Probably in part because of the x-over point of the woofers/horns, they sound best when your ears are roughly in line with the middle of the front baffles, around the mid-point between the horn and woofer dust-cap.
If his pricing is per pair I'm very intrigued... but I'm guessing it's per speaker, so perhaps buying the horns and CD's, and maybe even a flat pack, may be the way to go.
Basically, yes, with a 15 below them crossing at 250-300Hz or so. 15 midbass, 8 or 10 midrange and a CD/horn above. I have made many variations on the same basic formula.
I just picked up a pair of Altec Vott speakers I am restoring for a song and I am putting the 2225's I have in them (I just listened to them roughed in last night). I played around with no crossover but a diy l-pad and only a few watts while listening to them with a SEOS 10, a JBL horn setup I have as well as the Altec 811's that they came with. While these are far from being done, the early playing around with them gave me some ideas on what they will sound like when completed (aka awesome - I could literally hear the difference between mp3s on the old ipod I was using - at first I thought something was wrong with the drivers until I switched songs and the problem went away).
From skimming your post, it would be really important for you to find out what horns and drivers you currently have or else everything would be speculation. One option would be to get your hands on a large wave guide like the SEOS 24 or maybe 15 and get them playing down to the 500hz area, where it would match the 2225 (811's would also work).
If you put up pictures of what you have maybe we could help deduce what the parts are so that you aren't just asking in the dark.
From skimming your post, it would be really important for you to find out what horns and drivers you currently have or else everything would be speculation. One option would be to get your hands on a large wave guide like the SEOS 24 or maybe 15 and get them playing down to the 500hz area, where it would match the 2225 (811's would also work).
If you put up pictures of what you have maybe we could help deduce what the parts are so that you aren't just asking in the dark.
BL-409 will work
And just use pretty straight forward 800 Hz crossover
With whatever 2 bolt mylar driver fits your budget.
Moving up higher at 3k
Another small horn with whatever titanium driver fits your budget
And just use pretty straight forward 800 Hz crossover
With whatever 2 bolt mylar driver fits your budget.
Moving up higher at 3k
Another small horn with whatever titanium driver fits your budget