I've recently gotten interested in getting my feet wet in speaker building. What got me interested was Morgan Jones' article on designing a small sealed box speaker for flea power single-ended amps - the Arpeggio. Everything was nicely demonstrated as to the design principles and the Small equations that must be used to get your target results. The whole article really appealed to my "less is more" sensibility of using correct engineering principles to simplify the final product.
When I got into audio as a young fart, (now i'm an old fart), my first good sounding speakers were Spica TC-50s. They imaged like crazy, and were built from the start with one of the first computer designed crossovers ever designed. I never liked them on stands set 2 to 4 feet from the front wall. Just too thin sounding with no bass. So they were a very convenient speaker to have sitting on top of furniture where they disappeared into the room. I hate speakers that define a room because for a guy like me house real estate is just too precious to waste with big speakers that stand out several feet from the walls. So I'm really big on small speakers with smooth, but lower than usual bass response. I use bass traps to smooth out any boominess created by that speaker placement.
For equivalent speakers a sealed enclosure will be the smallest volume, another thing that appealed to me about MJ's article. So I'm thinking about building a speaker using a full range driver and using that article as a baseline. Thirty years after acquiring those original Spicas I'm only now appreciating one part of the design element that made made them so special. Apart from the time aligned nature of the woofer and tweeter what also contributed to their special appeal (I think) was the fact that the tonal balance didn't shift much as one moved away from the sweet spot. This was due to the slanted baffle.
In most normal speakers there is a certain amount of beaming in the high frequencies, even worse in many full range drivers. If you move from the listening sweet spot most speakers tonal balance changes. I think that is part of the reason the TC-50s were good. They used a rising response tweeter and but then directed it about 30 degrees up so one just got a smoother response high frequency response as one moved in the room. Think of an analogy where the speakers were directed straight up but with a FR designed for that. As long as you are moving in the horizontal plane FR will not change. Some people advise using a rectangular box speaker with a rising response driver and then just use toe in or out to correct for that at the listening spot. That is an inferior way of building a speaker because as you move laterally you will
decrease the high frequency response on one speaker but increase it on the other - terrible idea if you like to boogie to the music.
So I'm planning on building a small bookshelf speaker using MJ's ideas. But I'm going to select a driver with a rising response on the high end. Then tilt the baffle, ala Spica. With any luck it should image even better than the old TC 50s because i won't be using two drivers.
When I got into audio as a young fart, (now i'm an old fart), my first good sounding speakers were Spica TC-50s. They imaged like crazy, and were built from the start with one of the first computer designed crossovers ever designed. I never liked them on stands set 2 to 4 feet from the front wall. Just too thin sounding with no bass. So they were a very convenient speaker to have sitting on top of furniture where they disappeared into the room. I hate speakers that define a room because for a guy like me house real estate is just too precious to waste with big speakers that stand out several feet from the walls. So I'm really big on small speakers with smooth, but lower than usual bass response. I use bass traps to smooth out any boominess created by that speaker placement.
For equivalent speakers a sealed enclosure will be the smallest volume, another thing that appealed to me about MJ's article. So I'm thinking about building a speaker using a full range driver and using that article as a baseline. Thirty years after acquiring those original Spicas I'm only now appreciating one part of the design element that made made them so special. Apart from the time aligned nature of the woofer and tweeter what also contributed to their special appeal (I think) was the fact that the tonal balance didn't shift much as one moved away from the sweet spot. This was due to the slanted baffle.
In most normal speakers there is a certain amount of beaming in the high frequencies, even worse in many full range drivers. If you move from the listening sweet spot most speakers tonal balance changes. I think that is part of the reason the TC-50s were good. They used a rising response tweeter and but then directed it about 30 degrees up so one just got a smoother response high frequency response as one moved in the room. Think of an analogy where the speakers were directed straight up but with a FR designed for that. As long as you are moving in the horizontal plane FR will not change. Some people advise using a rectangular box speaker with a rising response driver and then just use toe in or out to correct for that at the listening spot. That is an inferior way of building a speaker because as you move laterally you will
decrease the high frequency response on one speaker but increase it on the other - terrible idea if you like to boogie to the music.
So I'm planning on building a small bookshelf speaker using MJ's ideas. But I'm going to select a driver with a rising response on the high end. Then tilt the baffle, ala Spica. With any luck it should image even better than the old TC 50s because i won't be using two drivers.
Since one would be dissipating most of the high frequency energy if the driver was point straight up it is not practical to have a full-ranger pointed straight up. But I think 30 degrees up baffle angle is practical. I've seen lots of full range drivers with FR charts going up around 2000 hz. But the beaming characteristics start as low as 500 hz at about 1db and from some charts I've seen can peak around 15k hz at over 15 db difference on axis vs 30 degrees off axis. That's huge.
I think full range driver designers should start thinking about designing in a high frequency tilt starting about 500hz and increasing to around maybe 15 db at 15k Hz. That would be the baseline flat frequency response with the baffle tilted up 30 degrees. If you were looking to more closely align with a B&K response curve then a single driver designer could modify it so the final 30 degree baffle tilt curve would match that modified frequency response.
It makes no sense at all to engineer high frequency lift when listening to a traditional vertically aligned baffle. In that condition, especially in a small
room, any change in lateral position of the listener with respect to the speakers will maximize the frequency response change as one moves from the center position. If the speakers were towed in originally to compensate for the rising response then as you move "right" the left speaker would peak 15 db at 15k hz at a 30 degree lateral angle while the right would dip say 5 to 10 db at that freq. If the speakers were towed out originally then then as you move right at some point the right speaker would increase 15db at 15k Hz at 30 degee angle while the left would further dip 5 to 10 db at that freq. So at a particular listener position you could have a maximum change in high frequency content of 20 to 25 db. That ain't the way to design a
system.
While just tilting the baffle up won't get rid of all lateral off axis frequency change it is a good happy medium between beaming loss versus efficiency loss.
I think full range driver designers should start thinking about designing in a high frequency tilt starting about 500hz and increasing to around maybe 15 db at 15k Hz. That would be the baseline flat frequency response with the baffle tilted up 30 degrees. If you were looking to more closely align with a B&K response curve then a single driver designer could modify it so the final 30 degree baffle tilt curve would match that modified frequency response.
It makes no sense at all to engineer high frequency lift when listening to a traditional vertically aligned baffle. In that condition, especially in a small
room, any change in lateral position of the listener with respect to the speakers will maximize the frequency response change as one moves from the center position. If the speakers were towed in originally to compensate for the rising response then as you move "right" the left speaker would peak 15 db at 15k hz at a 30 degree lateral angle while the right would dip say 5 to 10 db at that freq. If the speakers were towed out originally then then as you move right at some point the right speaker would increase 15db at 15k Hz at 30 degee angle while the left would further dip 5 to 10 db at that freq. So at a particular listener position you could have a maximum change in high frequency content of 20 to 25 db. That ain't the way to design a
system.
While just tilting the baffle up won't get rid of all lateral off axis frequency change it is a good happy medium between beaming loss versus efficiency loss.
no beaming with TB w3-1797, but you need bass helpers, because TB is a midhigh unit for use from about 600Hz (12dB), or 1000-1500Hz (6dB)... btw. it's a magical driver
Well, I looked at the response charts for it, and you're right that it looks like what I'm looking for. Unfortunately I'm looking for a little more efficiency so I can dip my foot into the single ended pond. The Alpair 12 mk2 looked real good but they don't seem to be available anymore.
then take two per side
6moons audio reviews: Boenicke Audio SLS
Boenicke Audiomanufacture SLS ))) Raindrop Audio (((
AUDIOMANUFACTURE ULTIMATE GRADE MUSICAL PLAYBACK INSTRUMENTS AND RECORDINGS
6moons audio reviews: Boenicke Audio B10
here is my opinion about those little gems
http://www.diyaudio.com/forums/full-range/185427-3d-sound-2.html
6moons audio reviews: Boenicke Audio SLS
Boenicke Audiomanufacture SLS ))) Raindrop Audio (((
AUDIOMANUFACTURE ULTIMATE GRADE MUSICAL PLAYBACK INSTRUMENTS AND RECORDINGS
6moons audio reviews: Boenicke Audio B10
here is my opinion about those little gems
http://www.diyaudio.com/forums/full-range/185427-3d-sound-2.html
I heard the Boenicke Audiomanfacture SLS once. The shop owner played the live version of "Calling Elvis" by Dire Straits. Sound was very live and the sound stage was created beautifully. It sounded like a proper performance vs his small showroom. But that was not only the TB W3-1797 playing, the whole speaker and electronics playing together.
But Dynablastertuners, you sound very impressed and have made me very curious. The FAST thing is bit of an inconvenience though... more stuff to put together.
Driver seems to be out of stock from PE (discontinued?) - are they coming out with a ferrite magnet version at lower cost?
But Dynablastertuners, you sound very impressed and have made me very curious. The FAST thing is bit of an inconvenience though... more stuff to put together.
Driver seems to be out of stock from PE (discontinued?) - are they coming out with a ferrite magnet version at lower cost?
The Alpair 12 mk2 looked real good but they don't seem to be available anymore.
Madisound are out of stock - have you tried CSS?
FAST is the way to go, active crossover with DSP and two amps is all you need, and today that stuff is cheap, and you are very flexible with such a solution
on TB site there are still those flatcones, nothing seems to be discontinued
I must say it again, for electronic music I claim there is nothing better in the universe, period. I think that same for home theater setup
PS: after listening to flatcone 3", all fullrangers that I have sounds flat in dimensionality, flat cone has flattened them all lol
on TB site there are still those flatcones, nothing seems to be discontinued
I must say it again, for electronic music I claim there is nothing better in the universe, period. I think that same for home theater setup
PS: after listening to flatcone 3", all fullrangers that I have sounds flat in dimensionality, flat cone has flattened them all lol
Last edited:
Dynablaster, I think perhaps you have a different philosophy than me. I hate adding new amplifiers, additional speaker drivers to a system if not required. More expense that seems to often just be the bill of goods the merchant is trying to sell you. Plus, how many diyers have access to CNC machines for carving horns. Sure, there are backloaded horn kits available but the whole premise of my original post was to use MJ's original Arpeggio project and perhaps update it and improve it.
The way I was trying to do it was to increase structural rigidity by creating a scalene triangle front to back box instead of a rectangular box. Plus that eliminates parallel front and back that reinforce standing waves. Simultaneously that angled front eliminates one of the biggest drawbacks of full range drivers - excessive beaming. There are a million projects out there where someone says "well the old one was good so let's just add this to it". It takes a lot more thinking to reorder things so you not only accomplish the original goal with less parts and complexity, but actually the improve the original in the process. I'm not saying my idea accomplishes that goal since I haven't built it to see, but it has a chance to. Plus its a lot more fun thinking about simple, but neat ideas. Of course YMMV
The way I was trying to do it was to increase structural rigidity by creating a scalene triangle front to back box instead of a rectangular box. Plus that eliminates parallel front and back that reinforce standing waves. Simultaneously that angled front eliminates one of the biggest drawbacks of full range drivers - excessive beaming. There are a million projects out there where someone says "well the old one was good so let's just add this to it". It takes a lot more thinking to reorder things so you not only accomplish the original goal with less parts and complexity, but actually the improve the original in the process. I'm not saying my idea accomplishes that goal since I haven't built it to see, but it has a chance to. Plus its a lot more fun thinking about simple, but neat ideas. Of course YMMV
Zman, thanks for the suggestion. I looked but it says they are out of stock on their site. I'm starting to realize their freq response profile isn't the best for my tentative plans. The best freq profile I've seen for a tilted baffle seems to be the 8" Tang Bands. Plus they have sufficient sensitivity for my needs. I wish they had more excusion so they could push more air like the Alpairs but it seems you can't get everything when you have very specific needs. I'll probably end up getting them as they seem to come the closest.
The Alpair 12 mk2 looked real good but they don't seem to be available anymore.
I have 2. The are odd-eyed (1 Silver, 1 Copper), fully treated and will go cheap.
dave
Another slightly different way to go about this can be Alp 7.3 - their mid and tops are superb and bass is acceptable if volumes are not high. In order to get more headroom and efficiency, and reduce excursion you can go for a 2 driver/cabinet top firing design - something like this:
http://www.diyaudio.com/forums/planet-10-hifi/205802-twin-chr-ken70-3-castle-teaser.html
It's a P10 design.
http://www.diyaudio.com/forums/planet-10-hifi/205802-twin-chr-ken70-3-castle-teaser.html
It's a P10 design.
Hi Zman and Dyna,
I hope I didn't come off as dismissive of your suggestion of adding more complexity to get better sound. Actually I understand how difficult it is to tailor a driver's frequency response to get good sound in an unorthodox application, like what I'm intending. There is no way to know, for instance, what tilt of the baffle will give the smoothest and most natural response for any given driver. It may be that even with a full ranger with the most aggressive top end that it only requires a 15 degree baffle tilt up. That would hardly counteract the high frequency beaming on most drivers.
So yes, i'm planning to have DSP on it but it will be incorporated into room correction software. That way I can be assured that I can tilt the drivers enough, say 35 to 40 degrees, that it will really make a difference in reduction in off lateral axis beaming. I also already have a subwoofer and will probably get another and they will be incorporated in the room correction software. I don't see any necessary reason to include woofers close to the main speakers since most low frequencies below about 80hz are omnidirectional. Doing it that way also agrees with my preference for keeping speakers against the wall and as inconspicuous as is humanly possible. (not always possible anyway)
But doing all this still requires that you get full rangers with the smoothest high freq tilt to begin with. That let's out most of the Alpairs that I understand are designed with a sonic signature for a natural sounding listening experience but not necessarily a smooth measured response. So I ordered a pair of Tang Band w8-1808.
Everyone has their own ideas how to do things and there certainly is no single right way to do things, just inclinations that we already have that alters how we arrive at our final destination. I'm just hoping my destination isn't audio hell as this is just an experiment. But I'm a risk taker so I'm not too worried (yet)!
I hope I didn't come off as dismissive of your suggestion of adding more complexity to get better sound. Actually I understand how difficult it is to tailor a driver's frequency response to get good sound in an unorthodox application, like what I'm intending. There is no way to know, for instance, what tilt of the baffle will give the smoothest and most natural response for any given driver. It may be that even with a full ranger with the most aggressive top end that it only requires a 15 degree baffle tilt up. That would hardly counteract the high frequency beaming on most drivers.
So yes, i'm planning to have DSP on it but it will be incorporated into room correction software. That way I can be assured that I can tilt the drivers enough, say 35 to 40 degrees, that it will really make a difference in reduction in off lateral axis beaming. I also already have a subwoofer and will probably get another and they will be incorporated in the room correction software. I don't see any necessary reason to include woofers close to the main speakers since most low frequencies below about 80hz are omnidirectional. Doing it that way also agrees with my preference for keeping speakers against the wall and as inconspicuous as is humanly possible. (not always possible anyway)
But doing all this still requires that you get full rangers with the smoothest high freq tilt to begin with. That let's out most of the Alpairs that I understand are designed with a sonic signature for a natural sounding listening experience but not necessarily a smooth measured response. So I ordered a pair of Tang Band w8-1808.
Everyone has their own ideas how to do things and there certainly is no single right way to do things, just inclinations that we already have that alters how we arrive at our final destination. I'm just hoping my destination isn't audio hell as this is just an experiment. But I'm a risk taker so I'm not too worried (yet)!
Last edited:
Not at all. DIY is all about trying out new things and following ideas/convictions - and that leads to your last paragraph:
Everyone has their own ideas how to do things and there certainly is no single right way to do things, just inclinations that we already have that alters how we arrive at our final destination. I'm just hoping my destination isn't audio hell as this is just an experiment. But I'm a risk taker so I'm not too worried (yet)!
Very true, and it will be good to learn from your experiences. Keep us posted on your progress and results.
Have fun!
With the phase plugs (and voice coil open to the air) pointing them upward is an invitation for early failure.
I was going to suggest Visaton B200, which if not modified with a phase plug, have a nice consisitent on axis rising from low to high.
dave
I just looked at the chart for it and the trend is right but is too extreme below 500hz where there isn't much beaming in a speaker that you would need to compensate for to start with. I'd just end up boosting the hell out of the lower midrange. Also, it looks like on the high end it's already down 10db at 15khz and 20 db down at 20khz. That's just guestimating a moving average from the response peak at 8khz. So I'll probably just stick with what I ordered. If it looks like it will be a problem I'll try to find an acoustically transparent cloth to cover it. Nothing is perfect in life and the interesting part is how one negotiates that fact.
Update: I decided that, at least initially, it would make more sense to create a rectangular sealed box mule speaker with the W8-1808. That way I can see what baffle angle is even possible without creating an overly ambient soundstage or running into limits with that would make the speaker too low efficiency. I can just tilt the speaker at different angles and see how the speaker responds with different digital room correction software settings. Even DRC has limits to how wide a sweet spot you can create in your room. I think a lot of that, but not all of it, has to do with how beamy your high speaker high freqs are to begin with.
Unfortunately, I have a lot on my plate at the moment as I'm doing a whole house renovation. So as not to disappoint anyone who is interested in my results, I should say up front that it might make more sense to try some experimenting on your own by tilting your own speaker up, rather than wait for my results. It's easy enough to do. Most likely you will hear some opening up in the narrowness of the sweet spot by doing that. However it os also possible, and probable, that you might hear added deficiencies in other areas, such as lower volume for the same watts and duller high frequency sound. That is where there might be a symbiotic effect with the normal peaking response of many full rangers. If you are already interested in room correction software that would be the natural correction for the problems you add by tilting them up. Remove the baffle step correction and let the room correction software correct make the changes for any given baffle tilt. Experiment to your heart's content. Fun, Fun, Fun!
Unfortunately, I have a lot on my plate at the moment as I'm doing a whole house renovation. So as not to disappoint anyone who is interested in my results, I should say up front that it might make more sense to try some experimenting on your own by tilting your own speaker up, rather than wait for my results. It's easy enough to do. Most likely you will hear some opening up in the narrowness of the sweet spot by doing that. However it os also possible, and probable, that you might hear added deficiencies in other areas, such as lower volume for the same watts and duller high frequency sound. That is where there might be a symbiotic effect with the normal peaking response of many full rangers. If you are already interested in room correction software that would be the natural correction for the problems you add by tilting them up. Remove the baffle step correction and let the room correction software correct make the changes for any given baffle tilt. Experiment to your heart's content. Fun, Fun, Fun!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.