Mission Possible
My designs are proprietary and owned by my clients. The PWT driver power response determines what the bandwidth limits should be. Here driver selection is the determinant. The horn profiles I use do not follow the design regimes to which refer. I have looked at all the Faital driver offerings, and have already provided a recommendation. For a compression driver to work properly, a back chamber is needed to provide a 'balancing' load to backside of the driver diaphragm. The issues here are not at the spectrum top end, instead they are at the bottom. Use of a wimpy driver here is contraindicated by the design mission.
WHG
My designs are proprietary and owned by my clients. The PWT driver power response determines what the bandwidth limits should be. Here driver selection is the determinant. The horn profiles I use do not follow the design regimes to which refer. I have looked at all the Faital driver offerings, and have already provided a recommendation. For a compression driver to work properly, a back chamber is needed to provide a 'balancing' load to backside of the driver diaphragm. The issues here are not at the spectrum top end, instead they are at the bottom. Use of a wimpy driver here is contraindicated by the design mission.
WHG
This is very good to know. Your graphs looked very good and your listening reports confirm an open back, no-phase plug, option with eg M5N12 would work very well. If a sealed back + phase plug con further improve that, it would be icing on the cake. Great!
Seems hard to source.
Above this midrange I will use a Beyma TPL-150H. Initially with factory waveguides, but might take a shot at a DIY horn later on. XO in the 2-2.5kHz region.
Below: a pair of Beyma 10G40 per side in sealed enclosures, working down to 80Hz. Below that I have sealed subs. The 10G40 work well up to 1kHz by some accounts, but I'm shooting for somewhere 350-500Hz XO.
Digital crossovers: can be very steep, allow time alignment. DSP capability to linearize driver response, and in-room response.
Community M200a new - Community M200 Midrange Driver Assembly | Full Compass
sounds like it's going to be a nice system. The Faital driver is not "wimpy" as stated above and should have no problem keeping up with the Beyma AMT. Adding a phase plug will help. Sizing the back chamber will be determined by flare rate, length of horn and mouth size. If it's a full size hypex flare you will be able to get it down to the flare size in free space. Tractrix is said to sound better for a mid horn though 🙂
sounds like it's going to be a nice system. The Faital driver is not "wimpy" as stated above and should have no problem keeping up with the Beyma AMT. Adding a phase plug will help. Sizing the back chamber will be determined by flare rate, length of horn and mouth size. If it's a full size hypex flare you will be able to get it down to the flare size in free space. Tractrix is said to sound better for a mid horn though 🙂
Last edited:
I was able to go though the papers. Very informative! Thanks for pointing them out.
From the Low Frequency Horn Design paper:
- "The midrange band is defined primarily by driver and back cavity compliance rolloff on the low end". Here midrange refers to midrange band, defined between fLC and fHM (p.7)
- "If fLS approx equal or larger than fc, the driver is not well suited for operation in a horn at that cutoff frequency (fc)" (p.12)
- fLS=fSxQts. For the M5N12 fLS=180Hz x 0.45 = 81 Hz. Not sure what the cutoff frequency would be for my needs, but likely larger than twice fLS. So I should be good.
- Back cavity volume for a 300Hz cutoff frequency: Vb=Vas/(fc/fLS-1)=1.20dm3/(300/81-1)=0.44dm3
Makes sense?
The paper on the Investigation of the Air Chamber of Horn Loudspeakers was also interesting. I admit I didn't go back to my fluid dynamics books to follow what they were doing, but I think I got the general idea.
I noticed the paper was assuming a compression driver and hence the shape they have. Plus a key assumption was the size of frequencies at play are comparable to the depth of the air chamber, which is assumed to have around 0.02 inches, or 0.5mm. But in my case the driver is a cone and at 2500Hz the wavelength is about 1380mm long, so I believe what Smith derives in that paper is probably not applicable for my case.
But maybe from practical experience you know his conclusions are applicable to cone drivers in horns and to frequencies within the range I'm looking into?
I will look into Hornresp during the weekend.
I definitely appreciate what you all are contributing to my learning!!
FYI....for horn simulation and comparison...
JBL provides these T/S values for the 8" 2169H CMCD driver. (2013 data)
FS = 320 Hertz
QTS = 0.61
QMS = 6.5
QES = 0.68
VAS = 0.55 liters
EFF = 1.3% (reference efficiency with half-space load)
Pe(max) =200 watts
Xmax = 3 millimeters
Re=5.1 ohms
Le =0.9 mH
SD = 0.020 square meters
BL= 19.8
Mms = 26 grams
Gap FLux = 1 Tesla
CAN ANYONE EXPLAIN....the Le=0.9mH seems high for a differential drive motor which should provide significant symmetric inductance canceling
I use 4 of them in a horn for midrange driven by 45 triode in my big mono system 🙂 It's scary
2pcs RT4 4x4 SEALED Back Midrange | eBay
Horn Design
Yes, so far as optimizing performance the bottom end is concerned,
the acoustics are scalable.
Two issues not addressed here include:
2) Mouth Reflectance.
To mitigate this, use a farness curve and match slope and curvature of the neck at the point of departure to form a mouth bell.
3) Pattern Beaming
A Salmon neck will manifest an onset of beaming at a much lower frequency than that of a Freehafer/Geddes horn [1], but you will sacrifice bottom end loading if you use the latter. If a severe diffraction slot is formed in the neck to encourage wave spreading, this will trigger premature HOM (transverse mode) formation in the horn.
Simply put horn design is a 'greasy balloon squeeze' [2]!
With a back-box and horn in place, and where [St]<[Sd], an open-frame driver is now transformed into a compression driver.
The clearance between the phase plug and the diaphragm of course will be in excess of [xmax] and thus the front chamber volume will be larger and accompanied by a decreased high frequency response. The wave fronts of concern are circular and are traveling along the dust cap and cone, not normal to them.
The first standing wave mode that would otherwise extend across the entire cone and dust cap will be weekend by the discontinuity where the dust cap and cone are joined. The modes across the dust cap and those between there and the compression chamber boundary are probably the ones to mitigate. With this in mind, study the JBL CMCD designs and reverse engineer from there carefully.
Again the acoustics are scalable.
Reference [3] may be of help here.
N.B., An accurate set of Bessel Function Values follow, the ones used in the article lack this accuracy:
J0(1) = 2.40483
J0(2) = 5.52008
J0(3) = 8.65373
J0(4) = 11.79153
J0(5) = 14.93091
J1(1) = 3.83171
J1(2) = 7.01559
J1(3) = 10.17347
J1(4) = 13.32369
J1(5) = 16.47063
Beyond (5) is outside the audio spectrum.
I hope this installment contributes as well.
Regards,
WHG
References
[1] http://dspace.mit.edu/bitstream/handle/1721.1/46001/35368752-MIT.pdf?sequence=2
[2]http://scholarsarchive.byu.edu/cgi/viewcontent.cgi?article=3447&context=etd
[3] AES PP 1384
Yes, so far as optimizing performance the bottom end is concerned,
the acoustics are scalable.
Two issues not addressed here include:
2) Mouth Reflectance.
To mitigate this, use a farness curve and match slope and curvature of the neck at the point of departure to form a mouth bell.
3) Pattern Beaming
A Salmon neck will manifest an onset of beaming at a much lower frequency than that of a Freehafer/Geddes horn [1], but you will sacrifice bottom end loading if you use the latter. If a severe diffraction slot is formed in the neck to encourage wave spreading, this will trigger premature HOM (transverse mode) formation in the horn.
Simply put horn design is a 'greasy balloon squeeze' [2]!
With a back-box and horn in place, and where [St]<[Sd], an open-frame driver is now transformed into a compression driver.
The clearance between the phase plug and the diaphragm of course will be in excess of [xmax] and thus the front chamber volume will be larger and accompanied by a decreased high frequency response. The wave fronts of concern are circular and are traveling along the dust cap and cone, not normal to them.
The first standing wave mode that would otherwise extend across the entire cone and dust cap will be weekend by the discontinuity where the dust cap and cone are joined. The modes across the dust cap and those between there and the compression chamber boundary are probably the ones to mitigate. With this in mind, study the JBL CMCD designs and reverse engineer from there carefully.
Again the acoustics are scalable.
Reference [3] may be of help here.
N.B., An accurate set of Bessel Function Values follow, the ones used in the article lack this accuracy:
J0(1) = 2.40483
J0(2) = 5.52008
J0(3) = 8.65373
J0(4) = 11.79153
J0(5) = 14.93091
J1(1) = 3.83171
J1(2) = 7.01559
J1(3) = 10.17347
J1(4) = 13.32369
J1(5) = 16.47063
Beyond (5) is outside the audio spectrum.
I hope this installment contributes as well.
Regards,
WHG
References
[1] http://dspace.mit.edu/bitstream/handle/1721.1/46001/35368752-MIT.pdf?sequence=2
[2]http://scholarsarchive.byu.edu/cgi/viewcontent.cgi?article=3447&context=etd
[3] AES PP 1384
Last edited:
Horn Design Addenda #1
Driver Selection References:
[4] Small
https://community.klipsch.com/index.php?app=core&module=attach§ion=attach&attach_id=34546
[5] Leach
http://users.ece.gatech.edu/mleach/papers/HornPaper/HornPaper.pdf
Regards,
WHG
Driver Selection References:
[4] Small
https://community.klipsch.com/index.php?app=core&module=attach§ion=attach&attach_id=34546
[5] Leach
http://users.ece.gatech.edu/mleach/papers/HornPaper/HornPaper.pdf
Regards,
WHG
Horn Design Addendum #2
This is a recent discovery. WHG
Oclee-Brown, Jack (2012) Loudspeaker compression-driver phase-plug design. University of Southampton, Faculty of Engineering and the Environment, Doctoral Thesis , 393pp.
http://eprints.soton.ac.uk/348798/1/Jack Oclee-Brown PhD Thesis.pdf
https://eprints.soton.ac.uk/348798/1/Jack%20Oclee-Brown%20PhD%20Thesis.pdf
[Link repaired by moderation]
This is a recent discovery. WHG
Oclee-Brown, Jack (2012) Loudspeaker compression-driver phase-plug design. University of Southampton, Faculty of Engineering and the Environment, Doctoral Thesis , 393pp.
https://eprints.soton.ac.uk/348798/1/Jack%20Oclee-Brown%20PhD%20Thesis.pdf
[Link repaired by moderation]
Last edited by a moderator:
Never measured for that exactly but they have more output in a horn than the Audax PR170MO - I think those are around 95 db midband on a baffle - in a horn I'd guess maximum output would hover around 120 db and average sensitivity 500 - 2K would be 103 db
you can measure all you want.. I have used the orig,, one made in France Audax PR170MO .. those midrange speakers are great...(back in the day i used my income tax refund to get those 2 midrange speakers. lol ).. they were 200 bux in 1981..
those 2 speakers are over 25 years old and sound just as good as the day I put them in
some large oak cabinets...my 2 cabinets were driven with the small carver receiver... 90 w per channel...horn tweeter polydax 6.5 midrange and a cheap cloth according 15 inch woofer with a 2 inch voice coil rated at..100 w.. 100w l pads,,, 250w crossover....
Those 2 speakers (Audax PR170MO) have the cleanest most powerful midrange I ever heard in a home stereo.... I have not heard them horn loaded.. those cones are light and the eff when horn load has to be over 102-103 DB..
i cant imagine even in a large room... !!!
those 2 speakers are over 25 years old and sound just as good as the day I put them in
some large oak cabinets...my 2 cabinets were driven with the small carver receiver... 90 w per channel...horn tweeter polydax 6.5 midrange and a cheap cloth according 15 inch woofer with a 2 inch voice coil rated at..100 w.. 100w l pads,,, 250w crossover....
Those 2 speakers (Audax PR170MO) have the cleanest most powerful midrange I ever heard in a home stereo.... I have not heard them horn loaded.. those cones are light and the eff when horn load has to be over 102-103 DB..
i cant imagine even in a large room... !!!
Sorry, not sure why you linked to that website? Their link to www.paudio.com.ar is broken so can't see that...
Sorry it's taking me so long to come back.
Some takeaways and questions from the [4] Small paper:
- It's not clear to me if his ultimate aim is for high-fidelity systems. He talks about the efficiency and power capacity tradeoff.
- It shows how critical the driver is in a cone driver + horn system. For whatever reason I was assuming the driver was critical, so good to have that confirmed.
- There is a tradeoff between hi frequency bandwidth and efficiency. This looks like a more critical factor for PA that for a 4-way hi-fi like I'm designing.
- The driver limits the horn system performance through EBP = fs/Qes
- If hi performance is required, the driver should provide high EBP values. He notes good drivers for sealed boxes should have EBP around 50Hz, and for vented boxes EBP of around 100Hz. For horns EBP should be higher.
- EBPs of all drivers in the short list from this thread have EBP above 300Hz, so they are all good from that point of view. All equally good? Dunno.
[*]Beyma 605Nd fs: 150, Qes: 0.32 => EBP=469Hz
[*]Faital Pro M5N12-80 fs: 180, Qes: 0.54 => EBP=333Hz- Faital Pro M5N8-80 fs: 180, Qes: 0.59 => EBP=305Hz
- B&C 6 PEV 13 fs: 126, Qes: 0.39 => EBP=323Hz
Driver Spec. Missing
N.B., EBP for this driver is 416 Hz. The more 'beef' the better.
I have fond memories of my visit to your country when I did some work for Acindar in Villa Constitucion.
Speaking of beef, hard to beat the flavor of a good Argentine steak.
Reference [1]
N.B., EBP for this driver is 416 Hz. The more 'beef' the better.
I have fond memories of my visit to your country when I did some work for Acindar in Villa Constitucion.
Speaking of beef, hard to beat the flavor of a good Argentine steak.
Reference [1]
Last edited:
Math & Beef
In the choice of a 4-Way system vs. 2- or 3-way system you have already sacrificed band-width for efficiency force majeure .
Qes = ([Re]/((*[l])^2))*(([Mms]/Cms])^(1/2))
fs = 1/((2*[Pi]*(([Mms]*[Cms])^(1/2)))
EBP = [fs]/[Qes] = ?
The more beef, the better.
Higher up, gossamer is good
A good bottom end performance is the key to the kingdom. A strong driver motor and sufficient volume displacement are most important here.
Horns are the Ferraris of loudspeaker enclosures,
for best results, do not put lawnmower engines in them.
Note that a loudspeaker driver is a 'wave surfer'. Driver headroom, no matter where the music is reproduced, remains important.
In a crowd, the finer details of music get lost, particularly when the venue is a rock concert and signal compression is unbounded. The more critical listening experience remains at home, Live (unplugged) concert orchestra or pipe organ performances excluded.
Regards,
WHG
In the choice of a 4-Way system vs. 2- or 3-way system you have already sacrificed band-width for efficiency force majeure .
Qes = ([Re]/((*[l])^2))*(([Mms]/Cms])^(1/2))
fs = 1/((2*[Pi]*(([Mms]*[Cms])^(1/2)))
EBP = [fs]/[Qes] = ?
The more beef, the better.
Higher up, gossamer is good
A good bottom end performance is the key to the kingdom. A strong driver motor and sufficient volume displacement are most important here.
Horns are the Ferraris of loudspeaker enclosures,
for best results, do not put lawnmower engines in them.
Note that a loudspeaker driver is a 'wave surfer'. Driver headroom, no matter where the music is reproduced, remains important.
In a crowd, the finer details of music get lost, particularly when the venue is a rock concert and signal compression is unbounded. The more critical listening experience remains at home, Live (unplugged) concert orchestra or pipe organ performances excluded.
Regards,
WHG
Last edited: