With my latest project complete, I have been thinking about what to do next. I want to work with drivers I have not worked with before, in a speaker system that is different from my recent projects.
The new Dayton signature series is intriguing. They seem to offer a lot of performance at a bargain price.
The SB26STWGC is also an interesting driver. It incorporates a small waveguide into the excellent SB soft dome tweeter. I have been quite impressed with the SB26STAC soft dome (flat faced) used in another project. The small waveguide of this tweeter provides directivity control above 3k, so it would seem to be most useful when combined with a midrange driver in a 3-way system.
SB26STWGC-4 / Fabric - Sbacoustics
I am also interested in doing something I promised myself I would not do… a 3-way passive crossover. The fact it is challenging and possibly frustrating also makes it interesting.
Normally my projects use sealed box woofers with DSP to extend the bass response. A passive speaker, on the other hand, should stand on its own and have sufficient bass response without external EQ. I have never worked with a passive radiator before, and this kind of system is intriguing.
I also want to build something which is modest sized and relatively uncomplicated to build. At this point I was envisioning a gross cabinet volume of 35 to 40 liters. I considered several form factors, including a a large stand mount and a very wide thing cabinet, but in the end, a small tower shape is what I decided upon. This will be another open source fully-documented design (like the LCCAM-10.3), so I want it to be speaker which has a wide appeal to many people with a variety of needs and rooms.
So the initial concept is a 3-way speaker using the SB26STWGC tweeter, a Dayton signature woofer, a midrange to be determined later, a passive radiator, and a passive crossover, in a floor standing tower cabinet. It should be easy to build, and it should be easy to integrate into a room, both sonically and aesthetically.
j.
The new Dayton signature series is intriguing. They seem to offer a lot of performance at a bargain price.
The SB26STWGC is also an interesting driver. It incorporates a small waveguide into the excellent SB soft dome tweeter. I have been quite impressed with the SB26STAC soft dome (flat faced) used in another project. The small waveguide of this tweeter provides directivity control above 3k, so it would seem to be most useful when combined with a midrange driver in a 3-way system.
SB26STWGC-4 / Fabric - Sbacoustics
I am also interested in doing something I promised myself I would not do… a 3-way passive crossover. The fact it is challenging and possibly frustrating also makes it interesting.
Normally my projects use sealed box woofers with DSP to extend the bass response. A passive speaker, on the other hand, should stand on its own and have sufficient bass response without external EQ. I have never worked with a passive radiator before, and this kind of system is intriguing.
I also want to build something which is modest sized and relatively uncomplicated to build. At this point I was envisioning a gross cabinet volume of 35 to 40 liters. I considered several form factors, including a a large stand mount and a very wide thing cabinet, but in the end, a small tower shape is what I decided upon. This will be another open source fully-documented design (like the LCCAM-10.3), so I want it to be speaker which has a wide appeal to many people with a variety of needs and rooms.
So the initial concept is a 3-way speaker using the SB26STWGC tweeter, a Dayton signature woofer, a midrange to be determined later, a passive radiator, and a passive crossover, in a floor standing tower cabinet. It should be easy to build, and it should be easy to integrate into a room, both sonically and aesthetically.
j.
When I looked at the Dayton signature drivers, the Sig225-4 seemed the best choice. Since I decided at the outset that this would be a passive radiator system, I had to evaluate the driver and PR together.
The Sig270-4 seems to be a great 10” woofer for the price. But it needs about 55 liters to work well in a vented or PR box. It would need either a pair of 10” PRs, or a single 12” PR. The resulting size and weight is beyond what I had in mind.
The Sig180-4 seems to be a very nice 7” mid-woofer, and it would be a great woofer for a 2-way. But it seems small for a 3-way speaker. If there was an 8 Ohm version, then a pair of them might be interesting to try, but for now it is 4 Ohm only.
After some experimenting with many options, I found the Sig225-4 works very nicely with the Dayton DSA270-PR. A woofer volume of 32 to 40 liters works nicely, and this is right in the range of size that I want. In a 35 liter box, F3 is 40 Hz, and both the woofer and PR hit their Xmax limits at 105 dB SPL.
The Sig270-4 seems to be a great 10” woofer for the price. But it needs about 55 liters to work well in a vented or PR box. It would need either a pair of 10” PRs, or a single 12” PR. The resulting size and weight is beyond what I had in mind.
The Sig180-4 seems to be a very nice 7” mid-woofer, and it would be a great woofer for a 2-way. But it seems small for a 3-way speaker. If there was an 8 Ohm version, then a pair of them might be interesting to try, but for now it is 4 Ohm only.
After some experimenting with many options, I found the Sig225-4 works very nicely with the Dayton DSA270-PR. A woofer volume of 32 to 40 liters works nicely, and this is right in the range of size that I want. In a 35 liter box, F3 is 40 Hz, and both the woofer and PR hit their Xmax limits at 105 dB SPL.
Just my couple of cents, but a PR for a 3-way system, especially with that much volume, doesn't make an awful lot of sense from a practical point of view.
In that case you might as well go for two smaller woofers in a ported cabinet for almost the same price.
Or just use the same woofer in a ported cabinet.
With a 3-way system we don't have to worry about the port resonance and with about 40-60 liter we can pick a port that is big enough that it will be far from creating any significant issues.
In that case you might as well go for two smaller woofers in a ported cabinet for almost the same price.
Or just use the same woofer in a ported cabinet.
With a 3-way system we don't have to worry about the port resonance and with about 40-60 liter we can pick a port that is big enough that it will be far from creating any significant issues.
I appreciate your thoughts, @b_force. Assuming a 40 l box, a vent of 7.5cm x 17.4 cm would provide a good response. This would give me a fundamental vent resonance of 814 Hz. Assuming a 400 crossover from woofer to mid, would you be comfortable in assuming the vent resonance would not be energized?
j.
j.
The next step in the conceptual design process is to consider cabinet shape. The largest element to design around is the passive radiator at 270 mm diameter. And to be clear, I am not wedded to the idea of a PR, but it would be interesting to use one. So we will see how the design shapes up...
I also need to consider the baffle edge treatment. Several of my recent projects have used tapered bevels to reduce the baffle area around the tweeter. This is a substantial construction hurdle for many people. If possible, I want this design to avoid a tapered bevel. I am hoping the use of a waveguide on the tweeter would lessen the need for tapered bevels.
Using the VituixCad diffraction simulation tool, I looked at a variety of tweeter placements on several cabinet widths from 8 to 14” wide (200 – 360 mm), and with various edge radius. It is not possible with this tool to fully simulate how a waveguide tweeter will respond in a baffle, but is possible to make some informed estimates. I came to the tentative conclusion that there was no penalty for a wider baffle (12” +) compared to a minimal baffle (8”). There were differences, yes, but nothing that jumped out as a disadvantage. I also tentatively concluded that an edge radius of 20 mm was a good compromise between ease of construction and good diffraction performance.
A driver hole has a significant negative impact on panel rigidity. This is why I and many others make our baffles at least twice as thick as the other cabinet walls, the extra thickness will partially compensate for the substantial loss in stiffness of the baffle. When I design a bracing layout, I am primarily focused on stiffening the baffle, and reacting the baffle out-of-plane loading and spreading that load into the other walls. If driver holes are present in other cabinet walls, those walls become less stiff as well, a more complicated bracing layout will be needed. With that in mind, I am disinclined to consider mounting the PR on a side wall or rear wall.
After evaluating many cabinet dimensions, I settled on a preliminary design that is 940 mm tall and 310 mm wide. This has enough width for the 270 mm PR and 20 mm bevels (or radius), although the PR will probably need to be surface mounted.
I also need to consider the baffle edge treatment. Several of my recent projects have used tapered bevels to reduce the baffle area around the tweeter. This is a substantial construction hurdle for many people. If possible, I want this design to avoid a tapered bevel. I am hoping the use of a waveguide on the tweeter would lessen the need for tapered bevels.
Using the VituixCad diffraction simulation tool, I looked at a variety of tweeter placements on several cabinet widths from 8 to 14” wide (200 – 360 mm), and with various edge radius. It is not possible with this tool to fully simulate how a waveguide tweeter will respond in a baffle, but is possible to make some informed estimates. I came to the tentative conclusion that there was no penalty for a wider baffle (12” +) compared to a minimal baffle (8”). There were differences, yes, but nothing that jumped out as a disadvantage. I also tentatively concluded that an edge radius of 20 mm was a good compromise between ease of construction and good diffraction performance.
A driver hole has a significant negative impact on panel rigidity. This is why I and many others make our baffles at least twice as thick as the other cabinet walls, the extra thickness will partially compensate for the substantial loss in stiffness of the baffle. When I design a bracing layout, I am primarily focused on stiffening the baffle, and reacting the baffle out-of-plane loading and spreading that load into the other walls. If driver holes are present in other cabinet walls, those walls become less stiff as well, a more complicated bracing layout will be needed. With that in mind, I am disinclined to consider mounting the PR on a side wall or rear wall.
After evaluating many cabinet dimensions, I settled on a preliminary design that is 940 mm tall and 310 mm wide. This has enough width for the 270 mm PR and 20 mm bevels (or radius), although the PR will probably need to be surface mounted.
Assessing the tweeter:
The next step is to see if my estimates of tweeter performance are accurate. I constructed a prototype from XPS foam board, 310 mm wide, with a 20 mm bevel. I made polar frequency response scans in 15 degree increments from 0 to 180. The results are shown below.
As I had hoped, the small waveguide begins working at about 3k. Most 25 mm dome tweeters have a constant DI of about 4 dB up to around 5k, and then they steadily become more directive. This small waveguide begins controlling directivity at 3k. It is a small, but important difference. It suggest that the tweeter is a good match for a 4” – 5” midrange driver crossed at about 3k. It also means that it is not necessary to minimize the baffle area around the tweeter with large tapered bevels.
The next step is to see if my estimates of tweeter performance are accurate. I constructed a prototype from XPS foam board, 310 mm wide, with a 20 mm bevel. I made polar frequency response scans in 15 degree increments from 0 to 180. The results are shown below.
As I had hoped, the small waveguide begins working at about 3k. Most 25 mm dome tweeters have a constant DI of about 4 dB up to around 5k, and then they steadily become more directive. This small waveguide begins controlling directivity at 3k. It is a small, but important difference. It suggest that the tweeter is a good match for a 4” – 5” midrange driver crossed at about 3k. It also means that it is not necessary to minimize the baffle area around the tweeter with large tapered bevels.
@hifijim Looking forward to see your work on this. The Sig225, with such a smooth response up to 2KHz, you may be able get away with just a single inductor and capacitor for acoustic LR2, for the W-M XO of around 500Hz, or even (or even higher, if necessary).
With a single aluminium demodulation ring, it will be interesting to see how the lower midrange distortion fares.
And a vented 8” should have as much output as a sealed 10”, as your previous design. I like passive crossovers, for the convenience of plug and play.
Can I ask you- do you have any data on what happens when the Bebe on the top edge is omitted?
And for people who would rather NOT do a bevel (for a variety of reasons) what is the roundover equivalent? 3/4”? 1”?
With a single aluminium demodulation ring, it will be interesting to see how the lower midrange distortion fares.
And a vented 8” should have as much output as a sealed 10”, as your previous design. I like passive crossovers, for the convenience of plug and play.
Can I ask you- do you have any data on what happens when the Bebe on the top edge is omitted?
And for people who would rather NOT do a bevel (for a variety of reasons) what is the roundover equivalent? 3/4”? 1”?
Last edited:
I would be happy (and honoured) to provide a vent suggestion based on my latest port geometry tests. Such a vent could even be installed temporarily using an adapter board that fits into the PR opening.
Following my findings an optimized port could provide the same or even better characteristics but with roughly half the length and thus double the resonance frequency, far outside woofer pass band. In addition the resonance peak would be very much reduced.
I could provide a 3d-printable file or even send a prototype to you. Let me know what you think.
I will update my port resonance+geometry thread shortly, by the way!
@stv I don't wanna ruin your suggestion, because you're doing beyond incredible work with those ports.
But what I was gonna say to @hifijim , is to just use a simple notch.
Which is perfectly doable passively as well as actively, since it's out of passband of the woofer.
Simple, cheap and very effective.
I don't know what what you're gonna use as a mid-range yet, but moving the crossover down obviously also helps.
Although lower than say 300Hz is not practical anymore with passive filter.
With a 2nd order filter, the response is already down at least 12dB at 800Hz.
Speaking of which. I believe all SIG woofers are 4 ohm?
So I would suggest first making a preliminary filter with the data from Dayton (or trace the data), to see if the impedance doesn't go to low.
But what I was gonna say to @hifijim , is to just use a simple notch.
Which is perfectly doable passively as well as actively, since it's out of passband of the woofer.
Simple, cheap and very effective.
I don't know what what you're gonna use as a mid-range yet, but moving the crossover down obviously also helps.
Although lower than say 300Hz is not practical anymore with passive filter.
With a 2nd order filter, the response is already down at least 12dB at 800Hz.
Speaking of which. I believe all SIG woofers are 4 ohm?
So I would suggest first making a preliminary filter with the data from Dayton (or trace the data), to see if the impedance doesn't go to low.
@b_force , using "my" port not only mostly eliminates the resonance problem of the port itself without additional resistance of a coil, but it also potentially
- reduces chuffing
- increases bass output
- reduces transmission of cabinet resonance effects via port
- and might even look nicer
I know you were gonna say that lol. 😄@b_force , using "my" port not only mostly eliminates the resonance problem of the port itself without additional resistance of a coil, but it also potentially
- reduces chuffing
- increases bass output
- reduces transmission of cabinet resonance effects via port
- and might even look nicer
Yes, I know.
But with these large cabinets, the most of these things aren't really an issue anymore, since we have so much wiggle room for size, surface area and length of the port.
Or in other words, we will be very comfortable with the low air velocity.
Leaking or resonances won't be that much of a problem, since we crossover at around 300-400Hz.
It's one of the benefits of a large 3-way system.
Very different than a 2-way system.
For internal cabinet resonances (standing waves), I would suggest going for a mass loaded transmission line (MLTL).
Although the name suggests a TL, it actually acts more like a BR.
With the benefits of being able to get rid of the vertical standing wave (since the wave length will be quiet high) of the cabinet as well as a solution to dampen a bit easier plus adding adequate mechanical bracing.
I couldn't resist to quickly simulate the SIG225 as well.
I wouldn't go as big as 40-50 liter with it, it seems quite happy around 30-35 liter tuned around 35Hz.
Which gives a nice smooth response and a f3 of around 35-38Hz or so.
Personally I wouldn't push a single regular 8 inch woofer lower than that.
The downside of this, is that the port will be much longer, with having the first port resonance roughly around 550Hz or so.
In that case we might want to use @stv wonderful and magical tricks!
Although this was simulated as a plain BR system, not a MLTL.
You'll loose a bit of effective volume/size in that case.
Very rough rule of thumb is around 20-30% more space needed for a MLTL.
(or any TL for that matter)
I wouldn't go as big as 40-50 liter with it, it seems quite happy around 30-35 liter tuned around 35Hz.
Which gives a nice smooth response and a f3 of around 35-38Hz or so.
Personally I wouldn't push a single regular 8 inch woofer lower than that.
The downside of this, is that the port will be much longer, with having the first port resonance roughly around 550Hz or so.
In that case we might want to use @stv wonderful and magical tricks!
Although this was simulated as a plain BR system, not a MLTL.
You'll loose a bit of effective volume/size in that case.
Very rough rule of thumb is around 20-30% more space needed for a MLTL.
(or any TL for that matter)
Jim, I have been designing something similar for the last couple of weeks. The only difference is I don't think I will ever build it. I have been planning a thread on what to do to feed my DIY habit when I don't need any more speakers. My design uses an Dayton DSA270-8 in a ported 68L cabinet and a Morel CAT378 horn loaded tweeter. I tried about 5 or 6 different 5 inch mid-range driver before finding the Seas MCA15RCY. I put this in a sealed 12L enclosure. The 12L enclosure create an acoustic high pass that forced me to cross at 500Hz to the woofer. This pushed my crossover point to the tweeter up to 4kHz due to the reality of passive band pass filters. Because you plan on going passive you may forced to crossing higher than you had with in the past with active crossovers. Of course this was 100% simulated design so you may be able to better with a real design. Take a look at the Seas MCA15RCY, it seems to be an easy to integrate mid.
Thanks for all the discussion. I appreciate everyone's thoughts.
@stv - thanks for the generous offer to design/build a vent for me. I have not committed to a type of reflex enclosure, and that decision point is still quite a bit in the future. If I decide to go with a vent instead of a passive radiator, I will discuss with you the best path forward... thanks !
In my prototype, the width of the bevel when viewed from the front is 20 mm. My bevel is cut at an angle of 35 degrees relative to the front baffle. Based on the preponderance of evidence, I believe that any angle from 30 to 45 degrees would be acceptable. I also believe that a roundover of 20mm radius would be equivalent.
I do have data for the box with a hard edge. All 3 sides had a sharp 90 degree edge
j.
@stv - thanks for the generous offer to design/build a vent for me. I have not committed to a type of reflex enclosure, and that decision point is still quite a bit in the future. If I decide to go with a vent instead of a passive radiator, I will discuss with you the best path forward... thanks !
In my prototype, the width of the bevel when viewed from the front is 20 mm. My bevel is cut at an angle of 35 degrees relative to the front baffle. Based on the preponderance of evidence, I believe that any angle from 30 to 45 degrees would be acceptable. I also believe that a roundover of 20mm radius would be equivalent.
I do have data for the box with a hard edge. All 3 sides had a sharp 90 degree edge
j.
Hopefully you mean today, I'm sure you know PE has a 12% off memorial day sale.
@hifijim
It's not totally clear to me what mid-woofer you're planning to use?
But just to give a bit of very general and generic advise.
With a 5 inch mid-woofer, you need at least a 5 inch waveguide.
Otherwise you will never get the directivity right.
3-3,5kHz is to high to be able to combine this well with a 5 inch woofer.
You will always see a discrepancy in the directivity curve.
With a 4 inch mid, we actually have two choices.
Both will work an will give you a different directivity and sound stage.
Without waveguide, the speaker will be a lot more omni-directional to about roughly 10kHz or so.
Above 8-10kHz the directivity collapses compared to a waveguided tweeter.
It's debatable how important this is, in my personal experience, it isn't really.
However, I think it's obvious that these approaches will sound VERY different.
From a more practical and visual point of view, I personally don't like the aesthetics of a 4 inch mid combined with a 8 inch (or bigger) woofer.
But obviously this is super personal and subjective.
Also the system could be entirely limited/held back by the 4 inch mid-woofer.
(I haven't done the calculations yet)
Anyway, I thought I just put it out there, because once you decide on certain things, there isn't really a way back anymore.
The choices are obviously 100% up to you, depending on the goals etc.
It's not totally clear to me what mid-woofer you're planning to use?
But just to give a bit of very general and generic advise.
With a 5 inch mid-woofer, you need at least a 5 inch waveguide.
Otherwise you will never get the directivity right.
3-3,5kHz is to high to be able to combine this well with a 5 inch woofer.
You will always see a discrepancy in the directivity curve.
With a 4 inch mid, we actually have two choices.
- a tweeter with a waveguide, like you have now
- tweeter with no waveguide.
Both will work an will give you a different directivity and sound stage.
Without waveguide, the speaker will be a lot more omni-directional to about roughly 10kHz or so.
Above 8-10kHz the directivity collapses compared to a waveguided tweeter.
It's debatable how important this is, in my personal experience, it isn't really.
However, I think it's obvious that these approaches will sound VERY different.
From a more practical and visual point of view, I personally don't like the aesthetics of a 4 inch mid combined with a 8 inch (or bigger) woofer.
But obviously this is super personal and subjective.
Also the system could be entirely limited/held back by the 4 inch mid-woofer.
(I haven't done the calculations yet)
Anyway, I thought I just put it out there, because once you decide on certain things, there isn't really a way back anymore.
The choices are obviously 100% up to you, depending on the goals etc.
I have a taller stand made of wood, but it was too tall, so I used these stacked milk crates. I'm not sure if "milk crates" are used outside of North America? they are heavily built plastic boxes, open on one end, and when stacked they lock together. I normally use them to transport tools in my vehicle. With large heavy speakers, I am always concerned about tipping...
I appreciate your comments and thoughts. I have not decided on a midrange yet. Based on simulation with ideal piston drivers, it looks like a 4" to 5" mid would ideally cross to the tweeter between 2.8k and 3.8k. I came to the same conclusion when I substituted data from the SB15NBAC30-4 driver from another project... I was able to get a good response with a 3k crossover.
The midrange will almost have to be an 8 Ohm driver. I worked up several crossover sims with 4 Ohm mids, and the overall impedance can get alarmingly low at some frequencies. It makes sense, because I am using a 4 Ohm woofer and tweeter, the mid should be 8 Ohm. That is just good design practice. The possible exception would be if I used a 4 Ohm mid which had a high sensitivity, something like 93 dB/2.83V@1m. Then I would be adding at least 3 Ohm resistance in series with the mid to pad it down... But I don't think a mid like this exists in my price range.
The Seas MCA15RCY is a good driver. I will make sure it is on the candidate list.
j.