Still fiddling with the design. I'm really liking where it's going.
Best part about it is the insides are asymmetrical so less internal resonance.
Feel free to chime in on this one. Would a wave guide like this work for a push-pull configuration?
Best part about it is the insides are asymmetrical so less internal resonance.
Feel free to chime in on this one. Would a wave guide like this work for a push-pull configuration?
You have depicted a push-push configuration with a small plenum, push-pull reverses one of the drivers.
It will "guide" sound waves out of the cabinet, where they will diffract around the plenum exit.
The edges should be rounded to avoid chuffing noises at high levels.
It would work OK, other than it's pipe resonance(s) and response null which may be in the passband of your low/mid crossover or woofers.
Earthquake's PUMP-12 passive radiator module looks similar, but it's pipe resonance of ~ 245Hz is well above the usual subwoofer range it is used in.
http://hyperphysics.phy-astr.gsu.edu/hbase/Waves/opecol.html
Modeling the effects of the stubby offset horn you are considering in Hornresp would be prudent.
Art
I actually kind of have a rule to my self about hornresp: The second I get to it, it means my design stopped being a design and started to become a science fare project. While I would like maximum performance, I also recognize that I am jumping through A LOT of hoops just to try and include two woofers, when in reality I could always just pick a larger bass driver. While a push-pull configuration would be nice, I could always just choose a higher performing woofer. A 10" driver will more than likely perform better than two 8" and so on. Especially because of the fact that those larger drivers are meant to go that low as opposed to the 8" which will inherently have limited capabilities when compared to the former. Most of all, I bet a 10" or even a 12" would be far easier to integrate into a design. The original idea was to do push-push for an inert chassis, but I can do the same thing by lining and bracing it well. I guess I'm off to find a nice sub.You have depicted a push-push configuration with a small plenum, push-pull reverses one of the drivers.
It will "guide" sound waves out of the cabinet, where they will diffract around the plenum exit.
The edges should be rounded to avoid chuffing noises at high levels.
It would work OK, other than it's pipe resonance(s) and response null which may be in the passband of your low/mid crossover or woofers.
Earthquake's PUMP-12 passive radiator module looks similar, but it's pipe resonance of ~ 245Hz is well above the usual subwoofer range it is used in.
http://hyperphysics.phy-astr.gsu.edu/hbase/Waves/opecol.html
Modeling the effects of the stubby offset horn you are considering in Hornresp would be prudent.
Art
The manifold out if the push-push driver pair should be oblong "V" shape out the front baffle. That guarantees no significant standing waves in the mids. The V shape slot is not extreme enough to create abnormal coupling out of the front baffle (potential horn loading). It should only act as air motion transformer at lower frequencies below the crossover to the mids.
In a previous 4th order bandpass design I've used elliptical coupling openings to a large common manifold using 4 x 12" neo drivers discharging into elliptical ducts out of a common V shape slot (similar to an MEH but without the WG loading). The shape and geometry of the port opening pushed the secondary resonance mode past 500 hz and rendered the primary mode dampened enough to not affect linearity, barely showing up in the impedance curve. The system topped out at 130 dB/2m @ 600W with a few percent THD from 50 - 400 hz.
When designing a high output slot loaded enclosure, its critical to load the driver cones evenly across their surface area. Neglecting to do this will result in uneven forces across the cone inducing rocking and twisting motion. It will significantly raise distortion and even damage driver VCs, even on drivers with dual spider suspensions.
Some of the enclosures I'm referring to here are technically a BP design with the primary chamber fed by the triangular shaped manifold I just described (internally slot loaded BP using separate external port). Designing a lower SPL slot loaded enclosure can be more tricky because of the size form factor. Using the V shaped geometry described above can sufficiently suppress the cavity resonance. The use of multiple smaller drivers in combination with the modified slot/manifold allows a significantly higher cutoff to the mids. I wasn't specifically saying it can only be designed this way, but its at least an option run smaller mids using a higher cutoff.
In a previous 4th order bandpass design I've used elliptical coupling openings to a large common manifold using 4 x 12" neo drivers discharging into elliptical ducts out of a common V shape slot (similar to an MEH but without the WG loading). The shape and geometry of the port opening pushed the secondary resonance mode past 500 hz and rendered the primary mode dampened enough to not affect linearity, barely showing up in the impedance curve. The system topped out at 130 dB/2m @ 600W with a few percent THD from 50 - 400 hz.
When designing a high output slot loaded enclosure, its critical to load the driver cones evenly across their surface area. Neglecting to do this will result in uneven forces across the cone inducing rocking and twisting motion. It will significantly raise distortion and even damage driver VCs, even on drivers with dual spider suspensions.
Some of the enclosures I'm referring to here are technically a BP design with the primary chamber fed by the triangular shaped manifold I just described (internally slot loaded BP using separate external port). Designing a lower SPL slot loaded enclosure can be more tricky because of the size form factor. Using the V shaped geometry described above can sufficiently suppress the cavity resonance. The use of multiple smaller drivers in combination with the modified slot/manifold allows a significantly higher cutoff to the mids. I wasn't specifically saying it can only be designed this way, but its at least an option run smaller mids using a higher cutoff.
I was actually just thinking about this a little while ago. While I know that asymmetrical internal structures for speakers are preferred, left or right of center would make it uneven. Then it would be an uphill battle of balancing the enclosure. Not worth the hassle. It's most likely going to come down to a larger sub, PRs or both. Bending over backwards just to make one small aspect work does not a good design make.
Funny enough I'm using a low cutoff point for the mids. MTMs suffer from a directivity issue, but this can be alleviated by using a lower crossover point for the tweeter. The T34B should be a good stand in for the RAAL I originally had in mind.
Slot loading can lower Fs a bit. It makes more of a difference on lower mms drivers, but those are more susceptible to cone rocking. There's a different slot shape which avoids this and also suppresses standing waves. I have to dig it out of my paper notes. The last time I built an enclosure with this type of vent, it was crossed over very high with great success.
Ok so a quick and dirty for LF drivers. Each curve is tuned for 20L with a 200Hz lowpass. The CSS 10" kind of took me by surprise, but the ported curve looks a little weird. Two things to note:
1) Chose the Ultimax II 10" over the 12" just for the smoother transition (see the spec sheets for the 10" and the 12")
2) Purifi 10" isn't in the running, but I thought I would put it there since there's a lot of 10" and 12" options and it would be a closer "apples to apples" comparison
@weltersys Can a woofer still use a PR on the left and right if the driver is off center on the baffle?
Cool. Would I need to anything differently?
Also, what size do I get them? In relation to the driver or one size up?
If I recall, you’re looking for a nearfield monitoring solution?…….passive radiators have awful impulse response so I’d take them out of consideration quickly…..it’s not hard to see why when you examine a motor less membrane flopping around with excess energy from within a box.
Since you mentioned super low distortion, your slot loaded solutions from 100hz act as an acoustic filter so you’re on the right track with your design goals. Problem is I’ve never been convinced that a person could actually hear odd order harmonic distortion at such low frequencies and long wavelengths after being modified by the room and boundaries. There’s quite a few practicalities that get thrown out with the baby and the bath water in the house of audiophoolery……be careful.
Since you mentioned super low distortion, your slot loaded solutions from 100hz act as an acoustic filter so you’re on the right track with your design goals. Problem is I’ve never been convinced that a person could actually hear odd order harmonic distortion at such low frequencies and long wavelengths after being modified by the room and boundaries. There’s quite a few practicalities that get thrown out with the baby and the bath water in the house of audiophoolery……be careful.
Thanks for the info. Sealed it is then! Not like it's the end of the world. Whole point was a three way so I'm just fine with that. If that's the case I may as well go with the CSS 12" to compensate for the lack of port.
https://www.klippel.de/listeningtest/?v=3 someone linked this to me a long time ago and I'm glad they did. You're definitely right in your assertions. Although it still surprises me how manufacturers don't focus more on this aspect (in regards to flat response not loudness... the auto industry took care of that one). I would imagine a subwoofer an integral part of a system just like any other driver.
I live and die by the spec sheet 🛠️ It's one thing to acknowledge something sounds good. It's another thing to say something is better without data. A fact without data is an opinion.
One thing I always do is to first verify the factory TSPs and if they stray (which they do 90% of the time), figure out under which conditions they were measured.
There are certain giveaways which expose a poorly made, cheap driver. Large scaling in FR curves with heavy smoothing is almost always an indication you're dealing with a POS driver. Well engineered examoles will have TSPs correlating to suitability for a practical range of enclosure sizes, yielding reasonably low cutoff frequencies.
Many cheaper drivers will have higher Qts with high Vas, making them unsuitable for most real world applications. Poorly vented, noisy motors are another indication you're dealing with a turd. What good does lots of VC overhang do if the driver generates too much noise and distortion of its own?
In the case of TSP discrepancies, its important to understand how they change with temperature and by how much. If the specs are off by a considerable amount, it should be investigated if there are other engineering problems plaguing the suspect driver, perhaps making it less suitable for its intended use.
In the case of the motor not being able to shed its heat quick enough, it won't sound the same at different SPLs for given periods of time. Even the materials chosen for the VC can make a difference - Alu wire has a higher temp coefficient than copper wire and it also has a lower fatigue temp, making it more failure prone. The lower midrange generates alot of heat, being the range where system impedance typically dips the lowest with minimal VC excursion, further compounding the problem, not moving enough cooling air across the VC windings. These are reasons why (for pro use) I always choose a driver with copper VC windings and good motor heatsinking abilities for use in the lower midrange. Most of the time, lower grade adhesives will fail before the windings do, but repeated heat cycling of the windings will anneal the wire, leading to early failure.
Reliability issues aside, thernal management is more important than most people think. The amount of drift in the enclosure tuning and alignment alone can be audible if not considered. Power compression can be a big issues if left unchecked. In extreme cases, turning up the volume to compensate for the SPL loss literally can result in no output gain whatsoever.
Most practical use scenarios won't run into these conditions, but they can be an issue if the driver isn't chosen appropriately for its intended use. I see quite a few cone midrange drivers with small VC diameters being employed in higher output / demand situations. Under these conditions, a speaker can significantly change its tonal balance and character as time and heat accumulate. That's why the choice of a suitable midrange / midbass driver is so important. Its likely the most critical driver in the entire system.
There are certain giveaways which expose a poorly made, cheap driver. Large scaling in FR curves with heavy smoothing is almost always an indication you're dealing with a POS driver. Well engineered examoles will have TSPs correlating to suitability for a practical range of enclosure sizes, yielding reasonably low cutoff frequencies.
Many cheaper drivers will have higher Qts with high Vas, making them unsuitable for most real world applications. Poorly vented, noisy motors are another indication you're dealing with a turd. What good does lots of VC overhang do if the driver generates too much noise and distortion of its own?
In the case of TSP discrepancies, its important to understand how they change with temperature and by how much. If the specs are off by a considerable amount, it should be investigated if there are other engineering problems plaguing the suspect driver, perhaps making it less suitable for its intended use.
In the case of the motor not being able to shed its heat quick enough, it won't sound the same at different SPLs for given periods of time. Even the materials chosen for the VC can make a difference - Alu wire has a higher temp coefficient than copper wire and it also has a lower fatigue temp, making it more failure prone. The lower midrange generates alot of heat, being the range where system impedance typically dips the lowest with minimal VC excursion, further compounding the problem, not moving enough cooling air across the VC windings. These are reasons why (for pro use) I always choose a driver with copper VC windings and good motor heatsinking abilities for use in the lower midrange. Most of the time, lower grade adhesives will fail before the windings do, but repeated heat cycling of the windings will anneal the wire, leading to early failure.
Reliability issues aside, thernal management is more important than most people think. The amount of drift in the enclosure tuning and alignment alone can be audible if not considered. Power compression can be a big issues if left unchecked. In extreme cases, turning up the volume to compensate for the SPL loss literally can result in no output gain whatsoever.
Most practical use scenarios won't run into these conditions, but they can be an issue if the driver isn't chosen appropriately for its intended use. I see quite a few cone midrange drivers with small VC diameters being employed in higher output / demand situations. Under these conditions, a speaker can significantly change its tonal balance and character as time and heat accumulate. That's why the choice of a suitable midrange / midbass driver is so important. Its likely the most critical driver in the entire system.
Last edited:
Don't know what you are asking.
The general recommendation is the passive radiator displacement should be a minimum of double that of the driver's Xmax.
As to the PR impulse response, you could expect it to be similar to a vented system with the same Fb (frequency of box tuning).
So I did a lot of looking around and found this woofer today https://sbacoustics.com/product/10in-sw26dbac76-4/ it's got good specs and, most importantly, it's thin enough that it solves the Vas problem.
Kept it simple on this one. Can probably be made a bunch of ways. Maybe a little more bracing, but I think we have a winner.
Kept it simple on this one. Can probably be made a bunch of ways. Maybe a little more bracing, but I think we have a winner.
As with practically any speaker, the most important aspects to concentrate on is the enclosure design and execution. Some drivers are more picky in this regard but in general, you'll need to keep the enclosure as "quiet" as possible, through a careful combination of stiffening and dampening. You need both of these features to complement each other. Neither one trait will work without help of rhe other. Bass improves with stiffening and bracing while midrange benefits primarily from dampening. Saying all this may sound obvious and redundant, but most speakers suffer from insufficient dampening, relying mostly on excessively stiffened and braced panels. They generally lack dampening where its most necessary, which in return would really clean up the lower mids.
I'm not even sure an enclosure that small could be braced. That and bracing may even take up too much volume. One thing I thought about is maybe 3d printed materials, something strong enough to keep the enclosure from flexing in any way. Still need to do more research on that front. As for dampening, I'm more than likely going to use resonix CLD and their multilayer foam and use decidamp as a paste to adhere everything to the walls of the enclosure. Polyfill (or something like it) will fill the remaining air space. Also, as per the loudspeaker cookbook, well nuts for every screw.
Getting it ready for prime time. I added better dimensions in order to make a more realistic model (I also accidentally added a couple of inches in depth, but nothing to cry over). The curve on the baffle was the only way I can think to do this. The baffle is probably going to be a separate piece entirely, but after a recent post I made ( https://www.diyaudio.com/community/threads/mechanical-decoupling.417853/#post-7797060 ) I think that might be the better plan anyway.
Next thing I need to figure out is the logistics of the decoupled MTM enclosure (I'm fairly certain I already have an idea but wouldn't mind any constructive criticism or ideas). It will have to be mounted from the back. One idea I got from the previously mentioned post is to use retaining hardware on the back of the drivers, so maybe some retaining hardware on the back of the enclosure isn't a terrible idea.
I think I settled on materials though. HDF would probably optimal here. It's dense and it's pretty rigid, great for the LF portion especially (since these speakers aren't going to be that big, I figure I can put up with the weight hit in terms of design). The MTM enclosure I probably have no choice but to get 3d printed. That's the only way I can think of that would get me the maximum space necessary for the drivers and their encapsulation. It would, in a way, also allow for a simpler design of the enclosure itself.
So here's how I think I'm going to do the MTM enclosure. Because of the wave guide on the front, it's going to have to be mounted from the inside. The front piece and the back piece will be bonded together and sealed via those pegs. Once together, it can be mounted with standard hardware from the back. I'm fairly certain this will need to be 3d printed. Doing this with wood would be more than laborious. With that said, I also had to move the speakers closer together in order to accommodate the guides. Real estate on that baffle is going to be tight.