I'm toying with idea of building a 3-way system in the form of a near cube (just for kicks). Say, 250mm H x 200mm W x 200mm D. With X-Over frequencies @ 250hz & 5kz.
The bass driver is a 6.5" down-firing sub-woofer. (Set at 30 degree angle). The midrange unit is a 6.5" forward facing bass-mid driver. And the top is provided by a dome tweeter.
The idea behind the low x-over frequency and the oversized midrange unit is that the mid-unit also serves as a passive radiator for the bass unit.
I'm not looking for a flat frequency response but can somebody explain why it wouldn't work?
The bass driver is a 6.5" down-firing sub-woofer. (Set at 30 degree angle). The midrange unit is a 6.5" forward facing bass-mid driver. And the top is provided by a dome tweeter.
The idea behind the low x-over frequency and the oversized midrange unit is that the mid-unit also serves as a passive radiator for the bass unit.
I'm not looking for a flat frequency response but can somebody explain why it wouldn't work?
Is there a chamber for the midrange driver? If not the woofer sound waves inside the enclosure creat pressures on the midrange driver cause distortion for the midrange driver sound waves.
Last edited:
Hi Surtsey!
A midrange unit can not serve as a passive radiator for a sub-woofer. Rather than augment the bass, the midrange is more likely to be pummelled to destruction by the sub's back pressure!
A midrange unit can not serve as a passive radiator for a sub-woofer. Rather than augment the bass, the midrange is more likely to be pummelled to destruction by the sub's back pressure!
No - that's the whole point. It needs to serve as a passive radiator for the bass driver.
You cannot simply use a midrange driver as a passive radiator. Firstly you'd have to cut it off from the amp at low frequencies (which actually could work, e.g. with a series cap). Bigger problems arise though. The passive radiator cone has to be tuned to the desired resonance frequency and in order to do so, you mostly have to add mass to the cone. Which would impair it's function as a midrange. Thirdly, the huge cone displacement of a passive radiator would again ruin the qualities of the midrange due to nonlinearities in the system.
I wouldn't know though why one would opt for a design like this. A closed 2,5-way solution, possibly with some active EQ-ing would give far better results.
I wouldn't know though why one would opt for a design like this. A closed 2,5-way solution, possibly with some active EQ-ing would give far better results.
None of this makes sense to me. (1) The midrange driver is already cut off from low frequencies via the cross-over. (2) The bass driver and the mid-range driver are both specified as 6.5" - giving them a similar mass. (3) A typical cross-over point for a 2-way system would be 6Khz, requiring the bass driver to provide mids as well. (4) The potential benefit of the design is size.
Please accept the advice of the more knowledgeable members, even if that not makes sense to you (at the moment).
Resonant frequency of the loudspeaker is determined from both cone mass and a spider+surround compliance. So, even if woofer and midrange cones have exactly the same mass (and they have not!), their resonant frequencies will be wildly different. Resonant frequency of a midrange is always higher than the woofer one. Compare the resonant frequencies of the woofer and the midrange you are intending to use.
To work properly, passive radiator resonant frequency must be lower than the woofer one.
Also:
and:
Please accept above advice from Galu and markbakk.
No, a typical crossover point for a 2-way is 2.5 or 3 kHz. Almost all bass drivers will manage that, except the real subwoofers.
Even if your midrange driver had the mechancial properties to be used as a passive radiator the whole construction wouldn't make the slightest sense.
The choice of a three-way system with its increased complexity is usually made because it allows the mids being radiated by a driver that exhibits very low excursion and therefore lower distortion etc.
If you now come and modulate that midrange (that you have electrically relieved from bass) mechanically you make the whole point of using a three way obsolete.
Regards
Charles
The choice of a three-way system with its increased complexity is usually made because it allows the mids being radiated by a driver that exhibits very low excursion and therefore lower distortion etc.
If you now come and modulate that midrange (that you have electrically relieved from bass) mechanically you make the whole point of using a three way obsolete.
Regards
Charles
I can test the theory without having to build the speakers. When I'm not on lock-down I work in a recycling shop. There's a single AV-120 somewhere. Acoustic Solutions AV 120 Main / Stereo Speakers | eBay
There's 3.3 [FONT=Verdana, Helvetica]uF capacitor on the tweeter (6Khz). If I rewire the speaker using a 3-way cross-over I'll get some idea what to expect, no?[/FONT]
[FONT=Verdana, Helvetica]
I'm off work for 8 weeks. I've got nothing better to do![/FONT]
[FONT=Verdana, Helvetica]
[/FONT]
With the right TS parameters, the mid can be used synergistically with the woofer. There is a patent on this. I forget the number but it’s old. You have to choose suspension stiffness (Vas) such that the mid (more likely a large full range) is similar. For example, a W8-1722 is really a flappy full range and makes a nice mid. Then use a typical 8in woofer with Vas to fit the box volume. This could all be simulated in Akabak and would help to optimize driver choice. However, you will have Doppler distortion on the mid due to cone motion from bass and playing mids at same time. It might not be so bad though as that happens on a full range anyway when a big excursion happens. With the tweeter facing up, it’s more for ambiance so that mid will need to cover up to 12kHz.
A cheap way to try this might be a PA130-8 for the mid and DC130 for woofer with a cheap 1in silk dome for the tweeter. It could be done in a tall box like a conventional standmount to test the concept. A cube looks neat to try and uses less material.
A cheap way to try this might be a PA130-8 for the mid and DC130 for woofer with a cheap 1in silk dome for the tweeter. It could be done in a tall box like a conventional standmount to test the concept. A cube looks neat to try and uses less material.
Everyone seems to have forgotten that, for full use of a PR you typically need 2x the volume displacement of the woofer.
This is typically achieved by using a pair of PRs for each woofer.
Chris
This is typically achieved by using a pair of PRs for each woofer.
Chris
You are referring to a Nestorovic Woofer System, but it works in a somewhat different way. In fact, all loudspeakers which used Nestorovic Woofer system (some Speakerlab models and of course Nestorovic Labs models) were 3-way and 4-drivers, with a genuine midrange driver and with (superficially looking) two woofers, with a different size. The larger woofer is partially woofer and partially a passive radiator, thanks to the crossover which is a part of the whole patented system.
Last edited:
There was a line of speakers made in the 1980's which basically cube-shaped 3 ways. The front baffle would have a 5 inch mid and a 1 inch dome, and then on the side would be an 8 inch or 10 inch woofer. I can not remember the name of this manufacturer.
The easiest way for you to achieve this kind of cube-like enclosure architecture with a tweeter and a pair of 6 inch drivers is to use the 2.5-way design.
As others have said, if you proceed with a 3 way design where your midrange is exposed to the back radiation of the woofer in hopes that it will become a passive radiator, you have a low probability of success.
The easiest way for you to achieve this kind of cube-like enclosure architecture with a tweeter and a pair of 6 inch drivers is to use the 2.5-way design.
As others have said, if you proceed with a 3 way design where your midrange is exposed to the back radiation of the woofer in hopes that it will become a passive radiator, you have a low probability of success.
Acoustic Solutions AV-120 is cheaply designed and build loudspeaker and a 3.3 uF cap can hardly be called a typical 2-way crososover. Quality 2-way loudspeakers have 2.5 - 3 kHz crossover point.
Theory is already proved and tested. With two identical midbass drivers (Asoustical Solutions AV-120), one with additional capacitor as a high-pass filter (turning it to a "midrange"), you will get one huge, deep hole at the midbass resonant frequency (say 50 Hz) and a big peak higher up (say 80 Hz).
That illustration is Fig. 6 from the US patent 3,984,635 by Nestorovic and Modaferri, describing the woofer system only.
LS1 is a regular woofer (8" or 10"), and LS2 is a slightly bigger (10" or 12", respectively) woofer. Fig. 3 and Fig. 4 show frequency response and impedance respectively for frequencies below 200 Hz only.
I have never seen a 2-way loudspeaker with Nestorovic Woofer System (although I suspect that one older Celestion model used it), but yes, you are right, it can be made as a quasi-2-way system, only if it follows these steps:
1. Pick one good 6" midbass.
2. Pick one good 8" woofer with substantially lower resonant frequency (both from higher mass and much higher compliance) and connect one big capacitor C (and a optional parallel resistor R) serially to it, as in above illustration.
3. Design a low-pass filter for 6" midbass + 8" woofer/passive radiator, preferably at 2.5 - 3 kHz.
4. Pick one good tweeter and design a high-pass filter which complements above low-pass filter, to get flat overall frequency response.
So, it can work, but it is completely opposite to the system proposed by OP.
Thanks to xrk971 for provoking a constructive debate. 🙂
LS1 is a regular woofer (8" or 10"), and LS2 is a slightly bigger (10" or 12", respectively) woofer. Fig. 3 and Fig. 4 show frequency response and impedance respectively for frequencies below 200 Hz only.
I have never seen a 2-way loudspeaker with Nestorovic Woofer System (although I suspect that one older Celestion model used it), but yes, you are right, it can be made as a quasi-2-way system, only if it follows these steps:
1. Pick one good 6" midbass.
2. Pick one good 8" woofer with substantially lower resonant frequency (both from higher mass and much higher compliance) and connect one big capacitor C (and a optional parallel resistor R) serially to it, as in above illustration.
3. Design a low-pass filter for 6" midbass + 8" woofer/passive radiator, preferably at 2.5 - 3 kHz.
4. Pick one good tweeter and design a high-pass filter which complements above low-pass filter, to get flat overall frequency response.
So, it can work, but it is completely opposite to the system proposed by OP.
Thanks to xrk971 for provoking a constructive debate. 🙂
Last edited:
While we don't often close threads on technical merit, this one leads me to believe that OP has not done even a little Googling on why this in counterintuitive. For that reason, the Mod team is closing this thread. Do not start a thread of this nature without at least doing a little bit of research.
- Status
- Not open for further replies.