These are my choices for speakers . Tower is 4 channel
woofer 21w/8555-01
m/w 15w/4531/g00
mid 15m/4531/k00
tweeter d3004/660000
all scan disk . I still haven't decided on an av unit to drive it all .
towers will have one of each
satellites will have tw / m /mw each
center will have 2 mids and 2 tweeters .
any thoughts , suggestions ?
woofer 21w/8555-01
m/w 15w/4531/g00
mid 15m/4531/k00
tweeter d3004/660000
all scan disk . I still haven't decided on an av unit to drive it all .
towers will have one of each
satellites will have tw / m /mw each
center will have 2 mids and 2 tweeters .
any thoughts , suggestions ?
Active or passive? Wwhat size room? what about room treatment?
It all sounds rather complicated. have you seen the 3 way kit from madisound?
https://www.madisoundspeakerstore.com/3-way-speaker-kits/b741-scan-speak-kit-pair-by-peter-noerbaek/
It all sounds rather complicated. have you seen the 3 way kit from madisound?
https://www.madisoundspeakerstore.com/3-way-speaker-kits/b741-scan-speak-kit-pair-by-peter-noerbaek/
Last comment, consider a complimentary kit for the surrounds to be the Klang-Tong 2 way kit also available from Madisound, with cabinets available from Taylor Speakers.
The ScanSpeak 18W drivers are pretty incredible and easily capable of being full-range in a surround system so long as your cabinet is up for it. Unless you are space constrained it's a much more capable woofer for a 2-way.
The ScanSpeak 18W drivers are pretty incredible and easily capable of being full-range in a surround system so long as your cabinet is up for it. Unless you are space constrained it's a much more capable woofer for a 2-way.
Last edited:
Thank you for the reply Erik . Yes it sure is to complicated as a 4 way . So I will reduce it to three . I appreciate the kit link but I would like to create something unique . I was looking at those Eton 7 inch , they are sweet . Does the spl rating need to be close when you pick speakers ? or within a range ?
Cirrus57,
In that case, I really think you should start with a 2 way. You'll learn not just the technology but also your own tastes in sound as well as where you want to spend your hobby time. This is a much faster and cheaper way to get to where you want to go than to try to design the entire thing all at once.
Things which appear very important right now in pure theory land become less important once you've built your first pair, and vice versa. You'll learn things you didn't know before about what you like or don't like. Lots easier to learn that with a simple system and just 2 speakers.
Usually I go for tweeters that have a higher SPL than the woofer. You'll end up padding the tweeter down and loosing some power in heat but since the treble energy is so much less than the rest musically this ends up being a small loss.
I like the Eton looks, but there's precious little to go on to justify the cost. They may be worth it, may not. A spec sheet won't really help me get there.
Even if you don't want to use some one else's kit, at least make sure you understand all of the design it took to get there.
What measurement tools will you use? Of course I'm a big proponent of OmniMic + DATS as well as XSim, but there's lots out there. I also use WinISD for box calculations.
Best,
Erik
In that case, I really think you should start with a 2 way. You'll learn not just the technology but also your own tastes in sound as well as where you want to spend your hobby time. This is a much faster and cheaper way to get to where you want to go than to try to design the entire thing all at once.
Things which appear very important right now in pure theory land become less important once you've built your first pair, and vice versa. You'll learn things you didn't know before about what you like or don't like. Lots easier to learn that with a simple system and just 2 speakers.
Usually I go for tweeters that have a higher SPL than the woofer. You'll end up padding the tweeter down and loosing some power in heat but since the treble energy is so much less than the rest musically this ends up being a small loss.
I like the Eton looks, but there's precious little to go on to justify the cost. They may be worth it, may not. A spec sheet won't really help me get there.
Even if you don't want to use some one else's kit, at least make sure you understand all of the design it took to get there.
What measurement tools will you use? Of course I'm a big proponent of OmniMic + DATS as well as XSim, but there's lots out there. I also use WinISD for box calculations.
Best,
Erik
When you design a speaker the overall sensitivity generally becomes dependant on the lowest frequency driver.
This is because it is difficult/expensive to attenuate (reduce sensitivity) of a speaker with a low resonant frequency, as creating a filter which attenuates the speaker evenly over it's frequency range requires flatting the woofers resonant impedance peak, and that requires very large value inductors/capacitors which need to handle significant power - it all becomes very expensive to do.
For that reason unless you are sure that the driver levels will match perfectly for your intended design you are better off picking a driver combination where the mid/tweeter is more sensitive than the woofer and attenuating the mid/tweeter in the crossover to match the level of the woofer.
You can also use baffle step to your advantage by crossing from the woofer to the mid at the baffle step frequency. This makes your mid effectively 6dB more sensitive than the woofer which may allow you to match a large (sensitive) woofer to a smaller (less sensitive) mid without having to level match too much in the crossover.
Another thing you can do is double up on the less sensitive driver. I.e. A 12" woofer will be too sensitive for a 5" midrange, but if you use two 5" midranges it may work out. This comes at a compromise though and that is directivity - off axis the two mids will cancel at certain frequencies. It becomes worse at higher frequencies - for that reason you'd never use two tweeters to boost efficiency.
I'll agree that if you've never designed a crossover before, I would highly recommend you start with a 2-way (and cheaper drivers) before attempting anything more complicated, otherwise you may disappoint yourself with the result.
If you're still keen using using high end drivers perhaps use the same drivers and enclosure as an existing design but attempt the crossover yourself. That way if designing the crossover becomes too hard you can just use someone elses design to complete them. It also allows you to study someone elses crossover design and learn the 'tricks of the trade'.
This is because it is difficult/expensive to attenuate (reduce sensitivity) of a speaker with a low resonant frequency, as creating a filter which attenuates the speaker evenly over it's frequency range requires flatting the woofers resonant impedance peak, and that requires very large value inductors/capacitors which need to handle significant power - it all becomes very expensive to do.
For that reason unless you are sure that the driver levels will match perfectly for your intended design you are better off picking a driver combination where the mid/tweeter is more sensitive than the woofer and attenuating the mid/tweeter in the crossover to match the level of the woofer.
You can also use baffle step to your advantage by crossing from the woofer to the mid at the baffle step frequency. This makes your mid effectively 6dB more sensitive than the woofer which may allow you to match a large (sensitive) woofer to a smaller (less sensitive) mid without having to level match too much in the crossover.
Another thing you can do is double up on the less sensitive driver. I.e. A 12" woofer will be too sensitive for a 5" midrange, but if you use two 5" midranges it may work out. This comes at a compromise though and that is directivity - off axis the two mids will cancel at certain frequencies. It becomes worse at higher frequencies - for that reason you'd never use two tweeters to boost efficiency.
I'll agree that if you've never designed a crossover before, I would highly recommend you start with a 2-way (and cheaper drivers) before attempting anything more complicated, otherwise you may disappoint yourself with the result.
If you're still keen using using high end drivers perhaps use the same drivers and enclosure as an existing design but attempt the crossover yourself. That way if designing the crossover becomes too hard you can just use someone elses design to complete them. It also allows you to study someone elses crossover design and learn the 'tricks of the trade'.
Last edited:
Learning
I picked the higher drivers because I know I'm going to learn a lot in the next few months and I can change the crossovers/enclosure to improve performance . The scan-speak d3004/6060-10 and the scan-speak 15m/4531/k00 sit at 90db . The Eton 7-308 sits at 89 . Cost is high but I think it will pay off . I used visaton's box sim crossover optimizer . Looks ok on paper ( lol ) . I do appreciate the advice . I was thinking of reclaimed hardwood for the enclosure with an exotic front . I would like to flatten out the woofers graph , it has a small spike and then a dip . Going to play with it a bit . ty again
I picked the higher drivers because I know I'm going to learn a lot in the next few months and I can change the crossovers/enclosure to improve performance . The scan-speak d3004/6060-10 and the scan-speak 15m/4531/k00 sit at 90db . The Eton 7-308 sits at 89 . Cost is high but I think it will pay off . I used visaton's box sim crossover optimizer . Looks ok on paper ( lol ) . I do appreciate the advice . I was thinking of reclaimed hardwood for the enclosure with an exotic front . I would like to flatten out the woofers graph , it has a small spike and then a dip . Going to play with it a bit . ty again
Cirrus57, Is there a commercial speaker which has the physical size + construction style + performance range(SPL/W, freq_range, polar_pattern, etc..) that comes close to your diy-vision? It seems you favor traditional drivers...e.g. no dipoles, no horns, no ribbons, no AMTs, no ESL,
A few slightly-over-the-top >90db/watt drivers:
--Satori TW29RN Ring Dome Tweeter with Neodymium Motor $145. (high 95db sensitivity, flat + extended SPL)
https://www.madisoundspeakerstore.c...w29rn-ring-dome-tweeter-with-neodymium-motor/
------
Faital 6PR122 95db sensitivity $140 93db sensitivity from NdFeB motor
http://www.faitalpro.com/products/files/6PR122/8/6PR122_datasheet_8.pdf
Woofer selection requires cabinet selection.
-------- PROVEN design ~85db sensitivity
The Kairos Three-Ways
SATORI-TW29R-dome
SATORI MW16PR-8 midbass
SB Acoustics SB29NRX-75 woofer
A few slightly-over-the-top >90db/watt drivers:
--Satori TW29RN Ring Dome Tweeter with Neodymium Motor $145. (high 95db sensitivity, flat + extended SPL)
https://www.madisoundspeakerstore.c...w29rn-ring-dome-tweeter-with-neodymium-motor/
------
Faital 6PR122 95db sensitivity $140 93db sensitivity from NdFeB motor
http://www.faitalpro.com/products/files/6PR122/8/6PR122_datasheet_8.pdf
Woofer selection requires cabinet selection.
-------- PROVEN design ~85db sensitivity
The Kairos Three-Ways
SATORI-TW29R-dome
SATORI MW16PR-8 midbass
SB Acoustics SB29NRX-75 woofer
Attachments
MDF is a great choice unless they will be in a humid environment. I believe the Eton's are hard cone drivers. Any hard cone drivers virtually always has a severe resonance in the upper midrange, right where the ear is most sensitive. Those drivers have better resolution, but require active crossovers with 24dB/oct cutoff rates, so the resonance is very attenuated. I'd go 3 way if you are going to use a hard cone woofer.
If you're new at this or are going to use passive crossovers, I'd recommend using paper or even better, polypropylene cones for woofer and midrange drivers. they often have little or no resonance issue. Seas makes a very good dome tweeter for around $50 last time I looked. Scanspeaks are top notch too. The Fountek 1.5 inch ribbon tweeter is another one I really like (about $50 ea.). For woofer and mid I like high end Peerless drivers. They compete well with the most expensive, and are much cheaper. An important spec for woofers is Xmax. Always look at the frequency response graphs for any driver. Flat is what you want. Ports and passive crossovers can take a lot of time to get right. If you build an active crossover, you can also add active EQ to pump up the low end of a woofer, without having to build another chassis and power supply. Low bass makes a BIG difference in overall enjoyment IMO.
If you're new at this or are going to use passive crossovers, I'd recommend using paper or even better, polypropylene cones for woofer and midrange drivers. they often have little or no resonance issue. Seas makes a very good dome tweeter for around $50 last time I looked. Scanspeaks are top notch too. The Fountek 1.5 inch ribbon tweeter is another one I really like (about $50 ea.). For woofer and mid I like high end Peerless drivers. They compete well with the most expensive, and are much cheaper. An important spec for woofers is Xmax. Always look at the frequency response graphs for any driver. Flat is what you want. Ports and passive crossovers can take a lot of time to get right. If you build an active crossover, you can also add active EQ to pump up the low end of a woofer, without having to build another chassis and power supply. Low bass makes a BIG difference in overall enjoyment IMO.
The bottom-right black Sony SS-AR1 speaker has won several major awards for sound quality and style. Attached photos will give you cabinet construction details.
Sony uses custom made ScanSpeak drivers. You could choose to use similar ScanSpeak production drivers for ~87db sensitivity,
--D3004/660000 tweeter
--Midrange ScanSpeak split cone Revelator 15W/4531G00
--Two AL-cone woofers ScanSpeak Revelator 22W/8857t00
OR use newer tech drivers like the (92db Faital 6PR122 mid or 86db Satori MW16P-8 mid) + (94db Satori TW29R tweet) + (two 8" SB23NRXS45-8 woofers = 92db).
Sony uses custom made ScanSpeak drivers. You could choose to use similar ScanSpeak production drivers for ~87db sensitivity,
--D3004/660000 tweeter
--Midrange ScanSpeak split cone Revelator 15W/4531G00
--Two AL-cone woofers ScanSpeak Revelator 22W/8857t00
OR use newer tech drivers like the (92db Faital 6PR122 mid or 86db Satori MW16P-8 mid) + (94db Satori TW29R tweet) + (two 8" SB23NRXS45-8 woofers = 92db).
Attachments
You can also look at Home - Solen Electronique - World leading producer of high-end crossover components for everything you need to design and build a set of speakers. They will even design the crossover for you as long as you buy the drivers from them (perfectly reasonable.). I have bought from this company for 20 yrs +. They have always been excellent.
The carry Peerless and several other high quality brands of speaker. They might even be able to tailor a tower box for you if you send them a picture of something like what you want.
Here's the best part. Retail speakers are always smaller than the optimal size for the woofer drivers. Thin wood as well. After all, they are shipping boxes of air around. If you build your own, the construction will be more solid, and the internal volume will be correct. This is where your efficiency and damping come together for an excellent sounding speaker system.
How small are retail boxes? Often less than 1/2 to proper size. Crippled before they make a single noise.
-Chris
The carry Peerless and several other high quality brands of speaker. They might even be able to tailor a tower box for you if you send them a picture of something like what you want.
Here's the best part. Retail speakers are always smaller than the optimal size for the woofer drivers. Thin wood as well. After all, they are shipping boxes of air around. If you build your own, the construction will be more solid, and the internal volume will be correct. This is where your efficiency and damping come together for an excellent sounding speaker system.
How small are retail boxes? Often less than 1/2 to proper size. Crippled before they make a single noise.
-Chris
The Canadian Chris Conspiracy (CCC) is a real thing.
AudioHobby.EU has much better prices on drivers, if you don't mind waiting. Solen's website is slow as molasses in winter and yeah, my experience with their customer service is iffy.
Best,
Erik
Last edited:
If we are talking "clones" or at least using production models to think about here's my $2 worth (inflation!).
The Sony SS AR-1 could have used the Be tweeter, which is a drop in replacement electrically, but for $20 large they felt it was too expensive. Retail to driver cost: 20:1
The Wilson Sabrina is another possible model for using off-the-shelf parts. Retail to driver cost: 27:1 or more.
Wilson used the Vifa D27 which is pretty rough at the top end. Overall Wilson is balancing cost with output. That is, he's matching the $$$ spent with how much output each driver will put out, so in this case most of the driver $$ is in the woofer, and goes down to the tweeter which is a Vifa D27. $30 off the shelf. I don't buy the claims of custom work doing much there. Not a terrible tweeter but personally in the same price point I'd have gone with the DX25 or XT25.
In the middle range I really want to give the low-end Be tweets a try around $280. Of course if money is no object and you want amazing linearity and frequency response, the Mundorf AMTs are very difficult to beat and run around the same as the top end Be tweeters.
I guess my point is, I think the Sabrina is a good example (in terms of design) of a large floor-standing 3-way, but if I was building it I'd have to seriously consider the tweeter. Of course, nowhere in my head can I justify spending that kind of money, nor would I design such a big bass bump. Also, for $16K they still use the cheapest possible floor mounting spikes? Give me a freaking break.
Best,
Erik
The Sony SS AR-1 could have used the Be tweeter, which is a drop in replacement electrically, but for $20 large they felt it was too expensive. Retail to driver cost: 20:1
The Wilson Sabrina is another possible model for using off-the-shelf parts. Retail to driver cost: 27:1 or more.
Wilson used the Vifa D27 which is pretty rough at the top end. Overall Wilson is balancing cost with output. That is, he's matching the $$$ spent with how much output each driver will put out, so in this case most of the driver $$ is in the woofer, and goes down to the tweeter which is a Vifa D27. $30 off the shelf. I don't buy the claims of custom work doing much there. Not a terrible tweeter but personally in the same price point I'd have gone with the DX25 or XT25.
In the middle range I really want to give the low-end Be tweets a try around $280. Of course if money is no object and you want amazing linearity and frequency response, the Mundorf AMTs are very difficult to beat and run around the same as the top end Be tweeters.
I guess my point is, I think the Sabrina is a good example (in terms of design) of a large floor-standing 3-way, but if I was building it I'd have to seriously consider the tweeter. Of course, nowhere in my head can I justify spending that kind of money, nor would I design such a big bass bump. Also, for $16K they still use the cheapest possible floor mounting spikes? Give me a freaking break.
Best,
Erik
Last edited:
Sony wrote a white paper on the SS-AR1 which you can download for more technical details, measurements, and dimensions. Sony had ScanSpeak manufacture a 5" midrange with custom T/S parameters optimized for Sony's rear-port midrange box. The SS-AR1 is a ~4-ohm speaker.
You have to decide if you want a ~4ohm or ~8ohm speaker.
I would recommend a sealed midrange box.
4-ohm---RUN a sealed box sim for the 15W/4531K01 "K" specs Qts=0.23 in 0.25cuft sealed box and see if you agree it looks good. You should also tweek box size for Qtc ~0.6.
8-ohm---RUN a sealed box sim for the 15W/8531K00 "K" specs Qts=0.27 in 0.25cuft sealed box and see if you agree it looks good. You should also tweek box size for Qtc ~0.6.
------
ScanSpeak offers the 15W/4531K01 4-ohm with Qts=0.23, 15W/8531K00 uncoated cone 8-ohm with Qts=0.31, Scanspeak 15W/8530K-00 8-ohm with Qts=0.27, and 15W/4531G00 4-ohm with Qtx=0.34(good for ported box). Confusing.
OT...take a look at the Satori line of drivers. They offer better value but will take more work to pull together a full design.
You have to decide if you want a ~4ohm or ~8ohm speaker.
I would recommend a sealed midrange box.
4-ohm---RUN a sealed box sim for the 15W/4531K01 "K" specs Qts=0.23 in 0.25cuft sealed box and see if you agree it looks good. You should also tweek box size for Qtc ~0.6.
8-ohm---RUN a sealed box sim for the 15W/8531K00 "K" specs Qts=0.27 in 0.25cuft sealed box and see if you agree it looks good. You should also tweek box size for Qtc ~0.6.
------
ScanSpeak offers the 15W/4531K01 4-ohm with Qts=0.23, 15W/8531K00 uncoated cone 8-ohm with Qts=0.31, Scanspeak 15W/8530K-00 8-ohm with Qts=0.27, and 15W/4531G00 4-ohm with Qtx=0.34(good for ported box). Confusing.
OT...take a look at the Satori line of drivers. They offer better value but will take more work to pull together a full design.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.