Modular active 3 way - work in progress

Nice topology for an all horn system. Only an observation --if you don't mind-- I would suspend the mdbass horn one foot from the floor (making it less rectangular, more square) so the sound would be cleaner in the sensible midbass region.
Comments and suggestions always welcome.

There were compromises made for the bass horn. The height was needed to place the midrange horn at the preferred height (90cm). It also had to work (benefit) with the floor to minimize floor bounce. The one physical thing I can't change is the floor between me and the speaker. It could be made wider but I'm not convinced I need to because the mouth is large enough for pattern control. I only needed it to reduce the beamwidth and hence acoustic power. If I moved it off the floor it would reduce the LF gain (4pi radiation) and allow more floor bounce.
 
Looks really nice! How low does the front horn is loading? It is not conical - right? The side walls/guides seem to have a slight curvature.

Thanks.

The side walls are curved, and it's a Hypex T=0.7, Fc=120Hz as built (minor variation from the previous sim due to actual wood dimensions). I did try a Conic in an earlier sim version but it throws polar side spurs without a mouth termination (and it's already large), and the loading was poor.

This is the Hornresp RadImp loading prediction for the Hypex built. The RadImp peak at 230Hz actually shows up in the SPL measurement and was EQ'd flattish.
 

Attachments

  • V13_Hornresp_RadImp.jpg
    V13_Hornresp_RadImp.jpg
    55.9 KB · Views: 101
The reason for my suggestion is that I already experimented both cases. Quite an improvement in localisation --especially on cello sound and men voices-- and lack of cavernous tone (which happens because bouncing reflections in such large horn (90cm) compared to upper midbass tones (just under 500Hz, lambda is about 70cm)). You don't really need LF gain in 200-500Hz but sure need clarity. Just try elevate it from the floor, maybe make a mockup, an idea smaller (under 70cm overal), to observe the difference.
Let a proper subwoofer serve spectrum under 120Hz, but don't try to lower the midbass horn to get there, it would get muddy. If you get "greedy" by trying to solve all at once, you lose more than you gain. I am sorry if it sounds condescending, but it took me many years and lots of money to learn these things.
 
No problem with different views and experiences.

The woofer array reduces floor bounce (and increases clarity IMO) in both simulations and measured results (SPL + THD). Some speakers have too much BSC and lifting them helps balance things. I've designed with the floor as a necessary component. I agree with your subwoofer comments and I typically use 2 subs for <80Hz. I'm not looking after LF gain but I'll use it, if its free 🙂 , I'm actually trying to better match the beamwidth of the bass horn to the mid horn, so the acoustic power is similar.
 
Just to finish the data set for the V13. The model, simulation results and spinorma report from the AKABAK simulation. The transition from woofer to mid horn is smoother now (less "step") and it sounds more balanced. This is the highest DI speaker I've made.

The actual measured on axis response and THD shows the horns are fairly clean. However my current room is not sufficient to accurately measure the bass response.
.
 

Attachments

  • DSCN8944r.jpg
    DSCN8944r.jpg
    246.2 KB · Views: 190
  • V13-model2.jpg
    V13-model2.jpg
    71 KB · Views: 195
  • Mod3Way-V13-Hpolar.jpg
    Mod3Way-V13-Hpolar.jpg
    33.6 KB · Views: 157
  • Mod3Way-V13-Vpolar.jpg
    Mod3Way-V13-Vpolar.jpg
    37.4 KB · Views: 157
  • Mod3Way-V13-ANSI2034.jpg
    Mod3Way-V13-ANSI2034.jpg
    41.1 KB · Views: 162
  • Mod3Way-V13-THD.jpg
    Mod3Way-V13-THD.jpg
    191.7 KB · Views: 187
The bass response is easier to see in polar plots from the simulation. It made me wonder if I could increase the the bass horn DI a little more. These are the polar plots that show the region from 100-300Hz that could be improved to make it look like the 400-600Hz region. I could make the horn bigger (not desirable) or maybe an active cardioid design that applies pressure at the horn side edges. Might be an interesting simulation.
.
 

Attachments

  • Mod3Way-V13-Hpolar@200Hz.jpg
    Mod3Way-V13-Hpolar@200Hz.jpg
    52.2 KB · Views: 96
  • Mod3Way-V13-Hpolar@300Hz.jpg
    Mod3Way-V13-Hpolar@300Hz.jpg
    52.6 KB · Views: 92
  • Mod3Way-V13-Hpolar@400Hz.jpg
    Mod3Way-V13-Hpolar@400Hz.jpg
    52.2 KB · Views: 97
  • Mod3Way-V13-Hpolar@500Hz.jpg
    Mod3Way-V13-Hpolar@500Hz.jpg
    51.7 KB · Views: 98
  • Mod3Way-V13-Hpolar@600Hz.jpg
    Mod3Way-V13-Hpolar@600Hz.jpg
    51.5 KB · Views: 98
  • Like
Reactions: vineethkumar01
^ I keep these 3 speakers, because they behave so differently (all were EQ'd flat on axis).

pic #1 : omnidirectional : no directivity, sprays the room, imaging is OK, power response is OK, I like this with acoustic instruments. It has a single 8inch woofer and a reflector so lower mids are not as defined as I would prefer. Needs a pedestal stand and >1m spacing from any wall.
pic#2 : conventional monopole, low directivity, has the best (even) power response (balanced), with good imaging and detail. The lower mids are very good because of more woofer Sd (than pic#1) and the array working with the floor (IMO).
pic#3 : full horn, high directivity, power response is OK, has the best imaging and detail. The lower mids are excellent and the bass horn improves transients (punch) and detail compared to pic#2.
 

Attachments

  • DSCN2076r.jpg
    DSCN2076r.jpg
    75.7 KB · Views: 174
  • DSCN8121r.jpg
    DSCN8121r.jpg
    171.2 KB · Views: 180
  • DSCN8944r.jpg
    DSCN8944r.jpg
    246.2 KB · Views: 178
Just to see how wide I could get the midrange band for V13 (D2200 + ZXPC). These tests were configured using a DSP-408.

pic#1 : Hpolar from AKABAK with midrange XO [500, 6K], with better EQ this time
pic#2 : Filter templates (RePhase) and measured output, with EQ
pic#3 : THD to check the driver
 

Attachments

  • Mod3Way_V13-XO-500-6K-Hpolar3.jpg
    Mod3Way_V13-XO-500-6K-Hpolar3.jpg
    31.7 KB · Views: 125
  • Mod3Way V13 Midrange  LR4 500-6K.jpg
    Mod3Way V13 Midrange LR4 500-6K.jpg
    166.6 KB · Views: 121
  • Mod3Way V13 Midrange THD LR4 500-6K.jpg
    Mod3Way V13 Midrange THD LR4 500-6K.jpg
    194.5 KB · Views: 111
A few more graphs. I think 500-6K is about the max usable range and all I require is 100dB max, which it easily does. I'm not sure about the actual diaphragm excursion under these conditions.

Pic#1 : LR4[500-6K] various SPL levels
Pic#2 : LR4[200-10K] to see where it falls off a cliff
Pic#3-5 : LR4[500-6K] THD at various SPL levels
 

Attachments

  • Mod3Way V13 Midrange Levels.jpg
    Mod3Way V13 Midrange Levels.jpg
    209.9 KB · Views: 105
  • Mod3Way V13 Midrange Bandwidth.jpg
    Mod3Way V13 Midrange Bandwidth.jpg
    155 KB · Views: 106
  • Mod3WayV13-THD 83dB.jpg
    Mod3WayV13-THD 83dB.jpg
    203 KB · Views: 92
  • Mod3WayV13-THD 88dB.jpg
    Mod3WayV13-THD 88dB.jpg
    208 KB · Views: 91
  • Mod3WayV13-THD 95dB.jpg
    Mod3WayV13-THD 95dB.jpg
    212.1 KB · Views: 94
  • Like
Reactions: Viper_user
Just realized the freq range graph above (pic#2) is wrong (LP@6K was still on). This is the update and the D2200 driver can go to 8Khz as per its spec.
 

Attachments

  • Mod3Way D2200 ZXPC Range EQ.jpg
    Mod3Way D2200 ZXPC Range EQ.jpg
    153.3 KB · Views: 100
  • Mod3Way D2200 ZXPC THD LR4 200 10K.jpg
    Mod3Way D2200 ZXPC THD LR4 200 10K.jpg
    202.9 KB · Views: 94
  • Like
Reactions: Viper_user
The bass horn has good directivity >300Hz but I interested in extending it lower into the 100-300Hz range. All these methods will involve active drivers to shape the beam width, using standard filters and time delays. The main woofer chambers are always sealed. Pic#4 shows a beam shaping improvement using a thin (3cm) horn mouth edge port. It's just not enough improvement, so we'll place shaping radiators in other locations, and try again.

pic#1,2 - baseline V14 (same as V13) model and Hpolar
pic#3,4,5,6 - V14.1 horn edge port with Hpolar's
 

Attachments

  • model_V14_baseline.jpg
    model_V14_baseline.jpg
    33.3 KB · Views: 101
  • BaselineV14Hpolar.jpg
    BaselineV14Hpolar.jpg
    24.3 KB · Views: 90
  • V14-2-model.jpg
    V14-2-model.jpg
    46.8 KB · Views: 103
  • Hpolar.jpg
    Hpolar.jpg
    25.4 KB · Views: 88
  • V14-2-Hpolar-200Hz.jpg
    V14-2-Hpolar-200Hz.jpg
    48.9 KB · Views: 107
  • V14-2-Hpolar-300Hz.jpg
    V14-2-Hpolar-300Hz.jpg
    48.2 KB · Views: 109
With side ports (still active driven). A little better, but still not enough improvement.
.
 

Attachments

  • V14-3-SidePort-Model.jpg
    V14-3-SidePort-Model.jpg
    37.2 KB · Views: 104
  • V14-3-SidePortsOnly-Hpolar.jpg
    V14-3-SidePortsOnly-Hpolar.jpg
    26.6 KB · Views: 107
  • V14-3-SidePortsOnly-Hpolar-200Hz.jpg
    V14-3-SidePortsOnly-Hpolar-200Hz.jpg
    48.6 KB · Views: 103
  • V14-3-SidePortsOnly-Hpolar-300Hz.jpg
    V14-3-SidePortsOnly-Hpolar-300Hz.jpg
    48.1 KB · Views: 98
  • Like
Reactions: Viper_user