I've been mulling around a concept that sort of came to me of a coax type 3 way system.
It employs an array of 6 x NE180W in V shape formation.
These sit behind an MTM arrangement of 2 x M74A and 1 x T34B mid and tweeter domes.
it uses the V baffle shape behind the mids and tweeter to time align the whole setup.
This may or may not have already been done by someone, but it looks like a valid concept for a monitor style speaker.
Any thoughts or inspiration?
It employs an array of 6 x NE180W in V shape formation.
These sit behind an MTM arrangement of 2 x M74A and 1 x T34B mid and tweeter domes.
it uses the V baffle shape behind the mids and tweeter to time align the whole setup.
This may or may not have already been done by someone, but it looks like a valid concept for a monitor style speaker.
Any thoughts or inspiration?
This is intended to be a fully passive system, wiring the NE180s in 3 groups of 2 drivers in series, paralleled with the 2 other groups of 16 ohm (2 x 8 in series) to make a net load of 5.3 ohms (16 ÷ 3). I'm trying to simulate the resulting sensitivity @ 2.8V which should be in the mid 90s.
There are at least a few examples of tweeters and sometimes also mids suspended over woofers to point to as proofs of viability of the concept. ME Geithan comes to mind. You won't know what the compromises are due to shading of the woofers by the bridge and reflections from it until you try it. You should be able to make a phase plug, or perhaps just foam, that sits in the V and guides the sound from the woofers past the bridge.
On the contrary, woofers will have more time delay than conventional design where woofers are on the same plane with mids and tweeter.
@AllenB Thats only in the nearfield and only if the crossover frequencies are incorrectly chosen.
This design is for mid- to far-field, so at 500 hz (0.68m WL), there is barely any directivity considering lowest frequency off axis null point in horizontal plane is at 1900 hz and vertical plane @ 950 hz .
This is why i chose to place the drivers as close as possible in the arrangement shown for best CTC, so the chosen midrange HP would be well out of the bundling zone.
The midbass drivers are all still running in full omnidirectional mode at the chosen midrange crossover point.
If someone wants to sim the approximate behavior in Edge, please feel free as it doesnt like to work well on my PC.
This design is for mid- to far-field, so at 500 hz (0.68m WL), there is barely any directivity considering lowest frequency off axis null point in horizontal plane is at 1900 hz and vertical plane @ 950 hz .
This is why i chose to place the drivers as close as possible in the arrangement shown for best CTC, so the chosen midrange HP would be well out of the bundling zone.
The midbass drivers are all still running in full omnidirectional mode at the chosen midrange crossover point.
If someone wants to sim the approximate behavior in Edge, please feel free as it doesnt like to work well on my PC.
A few questions:
- Is the V-Shape only to reduce the width of the speaker?
- Is a second mid dome needed?
- What is the use case for this design? It likely needs a sub but the speaker doesn’t look like a PA speaker.
If we do not take into account the phase rotation by the passive crossover filter. Then the acoustic center of the HF speaker and the midrange speaker in this configuration can be matched only by introducing a delay line into the HF band, since the HF speaker is closer to you than the LF speaker. But with passive filtering, the crossover filter itself will introduce a time mismatch that will not allow you to match the acoustic centers of the speakers using a HF signal delay. Simply put, if you want to match the midrange and HF bands in time, you need a DSP that can introduce a delay into the band and make a phase-linear digital crossover.
Speaker use purpose is a portable, relatively large (yet compact) style high output monitor.
I don't think the front vertical panel will negatively effect the lower mid output at those longer WLs.
Yes, the second M74A mid is chosen to make up for the narrow front baffle width and the resulting lower mid deficit, so it can be passively EQed in the crossover to be flat down to 500 hz and allow for a 2nd order HP. I don’t like using higher order filters between 300 - 1000 hz as these worsen the lower mid transient response. If the polar response is too negatively effected by the MTM, I'd consider TMM only if necessary.
I chose the 6 x 7" driver arrangement for tight driver spacing, accurate and highly detailed bass with a reasonable F3. Cone surface area is also greater for a given baffle size.
Pro drivers aren't as good in the lower register compared to their midbass. I also wanted a slightly higher crossover point to use with the M74A mid domes but didn't want the larger cone for this, as just about every PA driver like this has some sort of suspension related resonance issues in the lower mids. They also don't go that low. I already have these NE180Ws, so thats another good reason.
Sensitivity is better with a higher count of smaller capable drivers and everyone knows the reputation of the NE180W. Its an all around excellent, low distortion driver in almost every way. They handle ALOT of power (roughly 500W total long term for all them). They're also very light weight.
I don't think the front vertical panel will negatively effect the lower mid output at those longer WLs.
Yes, the second M74A mid is chosen to make up for the narrow front baffle width and the resulting lower mid deficit, so it can be passively EQed in the crossover to be flat down to 500 hz and allow for a 2nd order HP. I don’t like using higher order filters between 300 - 1000 hz as these worsen the lower mid transient response. If the polar response is too negatively effected by the MTM, I'd consider TMM only if necessary.
I chose the 6 x 7" driver arrangement for tight driver spacing, accurate and highly detailed bass with a reasonable F3. Cone surface area is also greater for a given baffle size.
Pro drivers aren't as good in the lower register compared to their midbass. I also wanted a slightly higher crossover point to use with the M74A mid domes but didn't want the larger cone for this, as just about every PA driver like this has some sort of suspension related resonance issues in the lower mids. They also don't go that low. I already have these NE180Ws, so thats another good reason.
Sensitivity is better with a higher count of smaller capable drivers and everyone knows the reputation of the NE180W. Its an all around excellent, low distortion driver in almost every way. They handle ALOT of power (roughly 500W total long term for all them). They're also very light weight.
Wrong. On both accounts.
Drawing from AllenB illustrates what I said before - there is a difference in paths from woofers and from mid/tweeter to the listening position. That is time delay. Nothing seriously wrong, but it is against your idea that
On the other hand, V-shaped design bring closer together horizontally both acoustical centers of (horizontal) woofer pairs, pushing lobing to higher frequencies (in the horizontal plane). Because crossover frequency will be at 500 Hz, there will be no lobing from woofers in a wide horizontal arc.
Align the woofer pair horizontally, not vertically, and then measure again.
Last edited:
@Sonce. A long while back I built a similar principle dual 10" 3 way with 2 Vifa M26WR LF drivers tilted vertically towards each other, then mounted a pair of Dynaudio D52s with a large ribbon in MTM arrangement. The crossover was simple and didn't require any complex all pass delays or other phase augmenting circuits.
My understanding is when being off axis in coverage, there was no multipath or early reflection triggering a jumbled wave front, as long as the reproduced WL from the tilted drivers was larger than double the cone diameter as well as larger than the driver's tilted depth offset.
When I was at ACR back in the late 80s, we designed a few kits with specific Fostex drivers. I did a notched baffle with 4 x 8" LF drivers just like I posted. When the crossover was done, the phase relationship was textbook and the directivity up to the LF LP point was improved compared to a typical vertically stacked driver flat baffle 3 way. This is what prompted me to try a multi LF 3 way with my favorite 7" midbass.
Don't get me wrong, I appreciate the input. I'm just trying to figure out why this would now be an issue after having tried other iterations of this design principle using passive filters. Designs like the TOA/Altec dispersion slot definitely had bad multipath problems, but that was at higher frequencies where the woofer was severely beaming.
My understanding is when being off axis in coverage, there was no multipath or early reflection triggering a jumbled wave front, as long as the reproduced WL from the tilted drivers was larger than double the cone diameter as well as larger than the driver's tilted depth offset.
When I was at ACR back in the late 80s, we designed a few kits with specific Fostex drivers. I did a notched baffle with 4 x 8" LF drivers just like I posted. When the crossover was done, the phase relationship was textbook and the directivity up to the LF LP point was improved compared to a typical vertically stacked driver flat baffle 3 way. This is what prompted me to try a multi LF 3 way with my favorite 7" midbass.
Don't get me wrong, I appreciate the input. I'm just trying to figure out why this would now be an issue after having tried other iterations of this design principle using passive filters. Designs like the TOA/Altec dispersion slot definitely had bad multipath problems, but that was at higher frequencies where the woofer was severely beaming.
Why would this be ok vs the horizontal arrangement?
The reason I wanted to avoid a vertical arrangement is cone/suspension sag. Probably a non issue with most smaller PA drivers but with high compliance units it could be a long term liability.
But you inherently have delay physically built into the array. This isn't about being off axis necessarily.
Earlier you said:
This is contradictory to what you've drawn. I think that's the main point people are trying to make. Whether the impact of this matters to you is a different discussion.
On axis, if we assume a 500 Hz 12 dB/octave cross point with 0.5 Q, and perfect 3" drivers with 5 inches of z offset between them, the result looks like this:
No, this is misunderstanding, sorry if I was not clear enough. I meant all woofers to be mounted on the vertical baffle together with mids and tweeter, with woofer pairs arranged horizontally, similar to your V-design.
Like this (warning: my first Microsoft Paint drawing!):
I made some calculation, with assumption of touching drivers: At 200 cm distance from the listener there is only 1.8 cm path difference between tweeter and the most distant woofer, plus the depth of the woofer acoustic center (which we don't know). I am quite confident this total path difference is smaller than in your V-design.
Flat front baffle design is much easier to build, but it is slightly taller than your V-design, also slightly wider because of larger woofer distance left and right of the tweeter. If compact dimensions are paramount, then your V-design is OK.