I'm interested in finding out how much bass a back-loaded horn (BLH) design gives in comparison with a bass-reflex one. Unfortunately for me, I couldn't find an answer with a Google search (I'm bad at physics so the technical papers are impossible for me to comprehend...).
I just need somebody to describe all the effects that a BLH have on a driver's frequency response. Please describe them in terms of peaks, slopes... in relation with possible parameters such as driver Fs, BLH tuning frequency (is there one?), etc.
Your answers are much appreciated. Thanks in advance!
I just need somebody to describe all the effects that a BLH have on a driver's frequency response. Please describe them in terms of peaks, slopes... in relation with possible parameters such as driver Fs, BLH tuning frequency (is there one?), etc.
Your answers are much appreciated. Thanks in advance!
It depends on the design and the match to the driver. If you had a huge horn mouth, proper expansion, proper driver, it can make enough bass to not need baffle step comp. and then some.
Plenty of stuff written online. Search fir MJK’s paper on design of BLH.
Learn how to use modeling software. Akabak or hornresp.
Read threads with simulations and measurements.
MJK's Fostex FF125wk BLH project
Plenty of stuff written online. Search fir MJK’s paper on design of BLH.
Learn how to use modeling software. Akabak or hornresp.
Read threads with simulations and measurements.
MJK's Fostex FF125wk BLH project
Last edited:
eugh, XRK971, altough your info is right, it's not the simpel basic info he wants. He said before in his post he does not understand the full science behind it and wants a simpel basic explenation to start understand it. Pointing him to the full complicated math behind him won't help him.
I'm not into horns so i'm certainly no speicialist, but as fas as i know the horn amplifies certain frequencies depending on the size (mainly the length), and that is used to give a driver more bass than it would have in a direct radiator.
I'm not into horns so i'm certainly no speicialist, but as fas as i know the horn amplifies certain frequencies depending on the size (mainly the length), and that is used to give a driver more bass than it would have in a direct radiator.
A back-loaded horn provides a better impedance match to air and increases the driver’s efficiency at lower frequencies below its cut-off point.
Dynic loudspeakers do not couple that well to the air they are trying to excite. A horn takes a high pressure/small area to a much larger are/lower pressure that couples better, providing greater efficiency.
dave
Dynic loudspeakers do not couple that well to the air they are trying to excite. A horn takes a high pressure/small area to a much larger are/lower pressure that couples better, providing greater efficiency.
dave
I'm poorly educated but will take a stab.
Horns are "transmission lines" and their characteristics are ruled partially by their length and quarter-wave resonance.
With an expanding cross-section line, that length may appear shorter than if a constant cross-section pipe. If the pipe has a negative taper, then the 1/4 wave frequency may appear somewhat longer than with the constant cross-section pipe.
Both bass reflex and BLH have their places and tradeoffs.
A back-loaded or front-loaded horn can produce more power for a given cone excursion than the reflex.
Sometimes a tall/skinny aspect reflex is useful and those are called MLTL (Mass Loaded T-line). Their 1/4 wave vertical dimension reinforcement can fill in a slumped area which can appear with lower Qts drivers in more cubical boxes.
Below is a rough comparison of a driver in a reflex vs a larger
BLH - I hope it isn't completely confusing.
You can see that even at 70 liters for the reflex, that's not a great response for that driver. The 2.4 meter length BLH trounces any reflex for that driver.
Horns are "transmission lines" and their characteristics are ruled partially by their length and quarter-wave resonance.
With an expanding cross-section line, that length may appear shorter than if a constant cross-section pipe. If the pipe has a negative taper, then the 1/4 wave frequency may appear somewhat longer than with the constant cross-section pipe.
Both bass reflex and BLH have their places and tradeoffs.
A back-loaded or front-loaded horn can produce more power for a given cone excursion than the reflex.
Sometimes a tall/skinny aspect reflex is useful and those are called MLTL (Mass Loaded T-line). Their 1/4 wave vertical dimension reinforcement can fill in a slumped area which can appear with lower Qts drivers in more cubical boxes.
Below is a rough comparison of a driver in a reflex vs a larger
BLH - I hope it isn't completely confusing.
You can see that even at 70 liters for the reflex, that's not a great response for that driver. The 2.4 meter length BLH trounces any reflex for that driver.
Attachments
I think it was the mass rolloff, so qes is the important value instead. The driver had a qts of .35, it was a long time ago.
Driver mass rolloff, Fmh(Hz) - Technical/Modifications - The Klipsch Audio Community
Driver mass rolloff, Fmh(Hz) - Technical/Modifications - The Klipsch Audio Community
OK, I'll try to wade in a bit here, contextually, so this is not supposed to be definitive (somebody who knows who they are 
-please forgive). I'll probably do over a few posts as / when time allows.
For a start:
Yes and no. I can't comment on the 0.35 being about right for baffle-step compensation, since the corner frequency of baffle step loss (assuming you start with the driver that has a flat response) depends on the width of the baffle. Natch. Now, I suspect what you're conflating this with is the driver's mass corner rolloff, i.e. the point at which it transitions from the rising response (acceleration) BW to the mass-controlled (flat) BW. The former is the region across which T/S parameters apply; above this their representation falls away. Mass corner Fhm = 2Fs/Qt' where Qt' = Qts + any series R in circuit from wire, connections, and / or amplifier output impedance. You can use effective Qe rather than effective Qt if you wish, which is usually a touch more 'forgiving'; again, it's Qes modified by series R or amplifier output impedance.
So from this, you can see that a Qt or Qe value of 0.xyz by itself won't tell you much, but combined with Fs, it will give you something that you might wish to work with.
By and large, mass corner is the upper limit you are likely to want the driver being supported by a bass load of any kind, and even then, it may be excessive. Assuming you're targeting a response below about 60Hz, then you probably won't want to have an mass corner above 300Hz or group delay will become excessive. Which more or less tells you everything you will need to know about the widespread assumption that a low Q drive unit is 'required' for horn loading. Not so: depends on context i.e. other driver parameters, notably Fs, type of horn, tuning, amplifier output impedance & / or series R in circuit etc.
-please forgive). I'll probably do over a few posts as / when time allows. For a start:
Yes and no. I can't comment on the 0.35 being about right for baffle-step compensation, since the corner frequency of baffle step loss (assuming you start with the driver that has a flat response) depends on the width of the baffle. Natch. Now, I suspect what you're conflating this with is the driver's mass corner rolloff, i.e. the point at which it transitions from the rising response (acceleration) BW to the mass-controlled (flat) BW. The former is the region across which T/S parameters apply; above this their representation falls away. Mass corner Fhm = 2Fs/Qt' where Qt' = Qts + any series R in circuit from wire, connections, and / or amplifier output impedance. You can use effective Qe rather than effective Qt if you wish, which is usually a touch more 'forgiving'; again, it's Qes modified by series R or amplifier output impedance.
So from this, you can see that a Qt or Qe value of 0.xyz by itself won't tell you much, but combined with Fs, it will give you something that you might wish to work with.
By and large, mass corner is the upper limit you are likely to want the driver being supported by a bass load of any kind, and even then, it may be excessive. Assuming you're targeting a response below about 60Hz, then you probably won't want to have an mass corner above 300Hz or group delay will become excessive. Which more or less tells you everything you will need to know about the widespread assumption that a low Q drive unit is 'required' for horn loading. Not so: depends on context i.e. other driver parameters, notably Fs, type of horn, tuning, amplifier output impedance & / or series R in circuit etc.
Last edited:
I'm afraid the reason you can't find out is because there isn't a single answer. In the same way there are lots of different types of 'bass reflex', there are lots of different types of horn loading, all of which give different results, and none of which can be called 'best' universally speaking.
To answer this frankly, you'd need to learn how to design horns, as in the physics behind them, which you say you're struggling with. I have on my shelf next to me, Bjorn's 1,000+ page tome on the subject, and that doesn't cover every single thing itself, not least because it somewhat depends on what you call a horn.
Keeping it simple, a horn is a form of acoustical transformer in that it efficiently couples the high mass of a drive unit to the low mass of the air. A back loaded horn is a bass horn: it only operates (and is only usable) in the low frequency region. Unlike a regular vented box, it operates over a wider range though: most do not have a single narrow peak of output, but a broader plateau over which it operates. This range is set by the length of the horn and its expansion. Due to size considerations, most bass horns have the expansion limited and unless designed for boundary loading, which provides a virtual increase of the terminus (mouth) size, are likely to show a resonance train of peaks and dips in their response until it is rolled off via a chamber, damping or similar.
here are a couple of useful references on horn loudspeakers in general
http://etd.lib.metu.edu.tr/upload/3/12607890/index.pdf
https://pearl-hifi.com/06_Lit_Archi...ec_51/4445_Acoustical_Engineering_3rd_Edn.pdf
http://etd.lib.metu.edu.tr/upload/3/12607890/index.pdf
https://pearl-hifi.com/06_Lit_Archi...ec_51/4445_Acoustical_Engineering_3rd_Edn.pdf
I have built a number of horn loaded speakers and can give a real world example of the differences between horn loaded and bass reflex speakers using the same drivers. Ten years ago, I built a set of DECWARE corner horns with Fostex 166EN (6.5inch) drivers. These are low Qts drivers suitable for horns or bass reflex cabinets. I built the first DecHorn and then built a base reflex using dimensions from Fostex. A short time later, I pulled the 166EN out of the bass reflex and built the second horn.
Using the Bode Plot (frequency response) program built into my oscilloscope, I compared the two. The Decs are very large corner horns. The base reflexes were 1/10th the size. The difference in sound output is substantial. The base reflex curve was pretty much as shown by Fostex. Somewhere around 100 hz the curve drops rather quickly with the speaker also placed in a corner. The Decs extended the flat part of the curve down another octave to 50 hz. There is an incredible amount of musical information in that additional 50hz. At 50 hz, the horns are 10 DB louder. There was almost no difference in response above 150hz. The most noticeable difference was that the DECs had a much greater dynamic range at lower frequencies. The horn loading was most noticeable in that the DEC speaker cones did not move during high output low frequency sounds where as the base reflexes moved in and out at least 1/4 inch. Speaker distortion is proportional to cone excursion and the Decs have none, the bass reflexes have some and sub woofers have lots. Everyone that has hear the Decs mention how incredibly clean the bass is.
Frequency response curves reflect the room at least as much as the speaker. The Decs are in a large workshop so the 50 cycle notes have the 25 feet needed to fully expand. In a small room, large corner horns have no advantage. The cutoff frequency of a horn is based on the mouth size (not length of sound travel) with a big output bonus for corner placement.
The Decs are corner horns, they use the walls of the corner to increase the apparent cone size which adds to the loading. Low frequency response is amazing. They are heavy (about 100 pounds each) and if the room corners are more than maybe 16 feet apart, can produce an odd sound stage. I added a center speaker (an Altec 15 inch 602 duplex in horn cabinet) which fills in the sound stage perfectly.
Hope this helps,
FE166 is not suited for a BR enclosure. It wants to be in a horn. With a highish Rout[.sub] amplifier.
All the FExx6 and FExx8 drivers fit that mold.
A more useful compraison would be a driver happy in a BR and can be horn loaded.
Tale Alpair 10p. Nicely fits in a 17 litre reflex but FHXL, Silbury extract maximum bass from the driver.
Hard to directly compare extention as both horns use the room to extend the horn so room and placement are final arbitors of ultimate bass extension.
Scott has done both he can give an estimate. I cannot find the original Silbury sim.
I do know that FHXL can reach 25 Hz, my Mar-Ken in the low 40s (miniOnkens do not try to mximize extention). F10.
Maeshowe may not go alot lower but will do so with more gain and less ripple.
FE166En (specific model is important in a reflex) miniOnken reaches the high 70s in 8.5 litres. In Victor (similar to Silbury) it reaches 50 Hz.
dave
All the FExx6 and FExx8 drivers fit that mold.
A more useful compraison would be a driver happy in a BR and can be horn loaded.
Tale Alpair 10p. Nicely fits in a 17 litre reflex but FHXL, Silbury extract maximum bass from the driver.
Hard to directly compare extention as both horns use the room to extend the horn so room and placement are final arbitors of ultimate bass extension.
Scott has done both he can give an estimate. I cannot find the original Silbury sim.
I do know that FHXL can reach 25 Hz, my Mar-Ken in the low 40s (miniOnkens do not try to mximize extention). F10.
Maeshowe may not go alot lower but will do so with more gain and less ripple.
FE166En (specific model is important in a reflex) miniOnken reaches the high 70s in 8.5 litres. In Victor (similar to Silbury) it reaches 50 Hz.
dave
At its extreme, a reactance annulled hyperbolic BLH can have ~108 dB/m eff. over ~ a decade and near that as an exponential variant as proven by the pioneers that used them for their largest early two way cinema horn systems [note the slotted open back covers for the four woofers]: http://www.audioheritage.org/images/lmco/photos/products/shearer/shearer3.jpg
Scaled back for HIFI/HT apps, the goal is just enough loading to offset any baffle step, so 3-6 dB gain out to whatever frequency it needs to be in one's app based on room size, speaker location.
BR vent gain varies depending on its tuning [Q] in relation to the driver's [Fs], box [Q], but as a general rule it's up to adding an extra half octave of F3 gain BW Vs sealed.
The rest of your answers requires some education, but all the necessary knowledge is on-line in various places as others are pointing out, though recommend starting with MJK's BLH doc as the most relevant for typical DIY speaker apps.
OK, I'll try to wade in a bit here, contextually, so this is not supposed to be definitive (somebody who knows who they are
Thank you!
Whether front or back loaded, it's a function of throat area, Scott covered the rest.