Adding an offset mid-bass driver to a 160hz tractrix horn, Wow

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have been enjoying a 3 way all horns system , tri amped with minidsp. The top end is covered by a JBL 2446 compression driver in a 160 hz paper mache tractrix down to 300hz, xover to a Emminence delta 15 in a 60hz bass horn that goes to 60hz and xover to a BIB subwoofer with a tangband W8-740P. The sound is amazing, specially with the JBL and tractrix. However I always thought that it was kind of low to ask the JBL to go and the xover freq 300hz is kind of high for the tractrix to integrate well with the bass horn . The tractrix is sitting on top of the bass horn. Then I start reading about synergy horn and an idea came into my mind to add a mid bass driver as an offset to the tractrix to extend the response down to 160hz. Since I have 2 minidsps and 2 miniDigis that would allow me to have 4 way all horn system. Hornresp is a great tool, but it does not simulate offset driver for tractrix. The closest thing that it simulates is offset driver for exponential. I want the mid bass driver to cover 160-500hz. The hornresp simulation with offset/ exponential horn for this Partsexpress buyout 5.25" NS-525 Aurasound fullrange was disapointing, the response was peaky in the pass band. I figured I could use miniDSP to smooth out the peaks. Once the desire to experiment overcomes the fear of ruining a perfectly good set of horns, I start cutting a 1.5 hole in the paper at about 5" from the throat of the tractrix and mount the mid-bass in its own enclosure. Wow , the improvement in the mid-bass region is definitely noticeable. Male vocals have weight and authority and I did not have to smooth out any peaks in the band pass region 160-500hz. A successful mod ! The horn in white is before the mod. The horn in yellow is after the mod and after I applied a coat of shellac. The wood looking horn is definitely unexpected.
 

Attachments

  • 2014-02-16 14.37.02.jpg
    2014-02-16 14.37.02.jpg
    280 KB · Views: 559
  • 2014-02-16 14.36.34.jpg
    2014-02-16 14.36.34.jpg
    544.2 KB · Views: 584
  • 2014-09-10 07.21.37.jpg
    2014-09-10 07.21.37.jpg
    599.5 KB · Views: 579
  • 2014-09-09 23.07.30.jpg
    2014-09-09 23.07.30.jpg
    517.5 KB · Views: 553
  • 2014-09-05 18.41.34.jpg
    2014-09-05 18.41.34.jpg
    566.1 KB · Views: 547
  • 2014-02-16 14.36.06.jpg
    2014-02-16 14.36.06.jpg
    460.5 KB · Views: 238
The hornresp simulation with offset/ exponential horn for this Partsexpress buyout 5.25" NS-525 Aurasound fullrange was disapointing, the response was peaky in the pass band. Once the desire to experiment overcomes the fear of ruining a perfectly good set of horns, I start cutting a 1.5 hole in the paper at about 5" from the throat of the tractrix and mount the mid-bass in its own enclosure.
The NS-525 has many times the displacement of the JBL diaphragm, so a dramatic improvement in clean SPL response would be expected.
Any measurements?
 
Looks like you've been visiting inlowsound web site...!?

Well done on your tests

I've built the 90hz Inlow bass horn, and presently have 340hz azura horn with jbl 2445.

Yes inlowsound.com has been a source of inspiration, information and designs and material ideas for the mold and paper horn. John has been very helpful with answering questions and giving encouragement. I had a few implementation details that differ from his. For example I put a layer of plastic wrap between the mold and horn to guarantee horn release as I did not take time to
make the mold ultra smooth like John recommended
 
The NS-525 has many times the displacement of the JBL diaphragm, so a dramatic improvement in clean SPL response would be expected.
Any measurements?

I have been playing with REW measuring software on Linux but some calibration must be off because the measurement don't show much changes when I use minidsp to equalize the response in certain area and I know that there's is a change because I can hear it and the RS meter also reported a change at that frequency. More tinkering with measurement tool is needed
 
The NS-525 has many times the displacement of the JBL diaphragm, so a dramatic improvement in clean SPL response would be expected.
Any measurements?

What I have found is SPL does not tell the whole story. Prior to the mood, the Spl in the 160-500 hz was acceptable, granted I had to boost the 2446 at 300-350hz by 5 db or so to get flat response in that region. But I figured the 2446 can handle it since I am only going to push at most 1watt thru it. The horn is 102+ dB sensitive. The mid bas s freed up the jbl from having to work harder down low and allowing to cross lower to the bass horn all help in making the system that much
Better
 
Founder of XSA-Labs
Joined 2012
Paid Member
FWIW, my tractrix made with foam core had a lot of coloration and harmonic distortion generating vibrations. Once I added a second layer with flexible latex caulking in between it dropped the HD significantly. Now at circa -50dB levels. The max SPL level went up too. I would be surprised if paper mâché horns do not also suffer from similar effect. The vibrations on the horn walls were very noticeable by touching it, and once the constrained layer damping (CLD) was applied, the vibrations that could be felt by hand went down considerably.

Here is my before and after CLD measurements.

Before:
434942d1408971865-prv-5mr450-ndy-fast-applications-tractrix-no-cld-old-eq.png


After:
434944d1408971865-prv-5mr450-ndy-fast-applications-tractrix-cld-new-eq.png


http://www.diyaudio.com/forums/full-range/259293-prv-5mr450-ndy-fast-applications-13.html#post4036324
 
Great work bqc.. Where the horns hard to build? Any sort of noticable colouration in the sound with that horn profile?

Horns were not hard to build, just take time and patience, inlowsound.com is great site for information and John is very helpful. The mold, turn table took me about couple days to make. Another day to smooth the mold surface and seal. Laying paper is the most time consuming takes about 5 to 6 layers of paper. If I work all day, I can get one horn done, but you should take a break after 3 layers, let it dry and continue with the rest of the layers. After spreading a thin coat of vaseline on the mold, I also cut strips of cling wraps and lay them down before starting with the paper so the horn will release easily. I dont have tools to measure distortion or coloration just my ears. To me they dont have as much coloration as the bass horn where there are large panels of wood to vibrate. John @ inlowsound.com said that paper has great damping property and I think he is right. The horn just sound amazing especially driven with the 2A3 SET. I think I may have changed its sound characteristic a bit with the coat of shellac :) . I have owned in the past Pi Speakers: 2Pi, 4Pi, 7Pi. Sachiko, Avebury and this by far is the most dynamic system. The horns fit my corner placement requirements. They are so directional that proximity to wall boundaries made no differences to the razor sharp stereo imaging and sound stage is huge. The down side is the sweet spot is narrow but I usually listen to music alone so that does not matter
 
What I have found is SPL does not tell the whole story. Prior to the mood, the Spl in the 160-500 hz was acceptable, granted I had to boost the 2446 at 300-350hz by 5 db or so to get flat response in that region. But I figured the 2446 can handle it since I am only going to push at most 1watt thru it. The horn is 102+ dB sensitive. The mid bas s freed up the jbl from having to work harder down low and allowing to cross lower to the bass horn all help in making the system that much
Better

Another perfect example of why you can't just EQ things to appear flat. Its not the same as a proper design in the first place. By changing your design you made it more applicable to the frequency range you were trying to cover. I like your flexibility in trying to modify your design to better meet your needs. People that take a design and try to EQ it to fit a different application are missing the boat. Proper design trumps signal processing every time.
 
expedit-1.jpg

nd91%2520vs%2520nd91%2520-%2520frequency%25201.jpg


One of my most successful projects was this clone of the Sound Physics Labs B-Deap. It's the exact same layout, but a fraction of the size. It's +/- 3dB, from 150hz to 450hz. It's insanely simple to build; you can literally make it in less than an hour.

The efficiency on it is so high, I'm actually thinking about a new way to fold it to REDUCE the efficiency! (It's so efficient it drowns out my mains.)
 
expedit-1.jpg

nd91%2520vs%2520nd91%2520-%2520frequency%25201.jpg


One of my most successful projects was this clone of the Sound Physics Labs B-Deap. It's the exact same layout, but a fraction of the size. It's +/- 3dB, from 150hz to 450hz. It's insanely simple to build; you can literally make it in less than an hour.

The efficiency on it is so high, I'm actually thinking about a new way to fold it to REDUCE the efficiency! (It's so efficient it drowns out my mains.)

I take that you did not use electronic xover in your setup. With minidsp, a click of a mouse to lower the gain and you're done. :)
 
The NS-525 has many times the displacement of the JBL diaphragm, so a dramatic improvement in clean SPL response would be expected.
Any measurements?

Finally got the REW software and RS SPL meter working together reasonably well. The weak link here is probably the RS meter. Here is the SPL graph with 1/6 octave smoothing. The BIB sub has a peak at 25hz. I have more gain in the amp that I use to drive the 60hz bass horn to lift the response from 60-100 but the bass horn has a resonance a 120hz so I have to leave the level as is. The bass sounds pretty much even and balanced with the rest of the system. The bass horn has a way of making string instruments such as cello sounds so good, very weighty and authoritative. Percusion is tight and can impact your chess if turned up. Nevertheless controlling resonances in the bass horn has been challenging due to so many large panels.
 

Attachments

  • plot2.jpeg
    plot2.jpeg
    161.7 KB · Views: 177
nice work.

Someday I'd like to try a straight bass horn that would be flat to 100hz without eq.

In my current room, I can tolerate the 6ft2 mouth (8ft2 in free air, 4ft2 coming out of the middle of a wall, so I estimate at 6 sitting on a concrete slab floor).

I cannot tolerate the less than 4' depth needed (old rule that mouth length should be 1/4 wavelength of 1/2 octave lower than the mouth Fb).

Flat to 150hz horn would be so much easier, but I don't like to cross there. 100hz is about my sweet spot to not mess with male voices and drums.


Norman
 
Founder of XSA-Labs
Joined 2012
Paid Member
nice work.

Someday I'd like to try a straight bass horn that would be flat to 100hz without eq.

In my current room, I can tolerate the 6ft2 mouth (8ft2 in free air, 4ft2 coming out of the middle of a wall, so I estimate at 6 sitting on a concrete slab floor).

I cannot tolerate the less than 4' depth needed (old rule that mouth length should be 1/4 wavelength of 1/2 octave lower than the mouth Fb).

Flat to 150hz horn would be so much easier, but I don't like to cross there. 100hz is about my sweet spot to not mess with male voices and drums.


Norman

I was able to cross to the bass horn at about 100Hz with the Trynergy (tractrix synergy) as shown here:
437044d1410006435-40-full-range-synergy-image.jpg
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.