Most back loaded horns expand in only on dimension, have many folds, and are a hybrid horn/TL/bass reflex design. I am experimenting with a pair of Fostex FE126E drivers in different enclosures and plan on building a relatively compact and efficient design for a small room listening environment.
The design would use the principle of the cochlea in the human ear, but operate in reverse with the driver replacing the "sensor" and be back loaded. It would essentially mechanically filter high frequencies and amplify low frequencies.
Why the Inner Ear is Snail-Shaped
Short article with scientific link
http://focus.aps.org/story/v17/st8
Coiled like a tuba instrument, it would expand exponentially. It could be made with paper mache with 20 gauge steel wire forming the skeleton. Wood glue would be used for strength. Damping should be similar to MDF, but it would need to be built up in layers. Has anyone tried this before? Also, how is the compression chamber volume and throat area determined for maximum efficiency?
Using the same principles, theoretically it could also be made into a 1/2 wave mass loaded transmission line with a sealed end like the B & W Nautilus. Not sure of the effectiveness of this though.
The design would use the principle of the cochlea in the human ear, but operate in reverse with the driver replacing the "sensor" and be back loaded. It would essentially mechanically filter high frequencies and amplify low frequencies.
Why the Inner Ear is Snail-Shaped
Short article with scientific link
http://focus.aps.org/story/v17/st8
Coiled like a tuba instrument, it would expand exponentially. It could be made with paper mache with 20 gauge steel wire forming the skeleton. Wood glue would be used for strength. Damping should be similar to MDF, but it would need to be built up in layers. Has anyone tried this before? Also, how is the compression chamber volume and throat area determined for maximum efficiency?
Using the same principles, theoretically it could also be made into a 1/2 wave mass loaded transmission line with a sealed end like the B & W Nautilus. Not sure of the effectiveness of this though.
With some inevitable exceptions the majority of back loaded horns can be regarded as BR cabinet variations with gigantic vents. Most typically use a hybrid of QW & horn actions because a full-sized horn is gigantic. For e.g. at 40Hz, you'd need a mouth ~85in x 85in. Even corner loading, which nominally permits an eightfold reduction in mouth area, still demands a very substantial terminus area if full horn-loading is to be maintained. And that's without bringing other issues into play.
Folds are a double edged sword. They're the enemy of absolute efficiency; however, they can be useful too. A perfectly smooth expansion is not automatically going to be an advantage.
Most designs expand only in one direction a) because it's easier for most DIYers and b) given that they're bass-horns, the added potential distortion / standing waves aren't as much of an issue down in the LF where our hearing sucks, as it would be in the midband.
Good luck & let us know how you get on. Although the words 'horn' and 'compact' sit rather uneasily together. Assuming we're talking bass-horns here anyway.
Could be rather attractive.
Seems reasonable, although you might not find the exponential contour to be your best option. I believe paper mache has been used quite a bit by different people, so yes, it's been done.
Ah. Something tells me you might want to read up about horn design before you start as these are rather basic factors. If you want maximum efficiency for example, an exponential horn is not going to be the best option. GM posed a plethora of sources in a thread here a couple of weeks back. Have a quick look through the last 3 - 5 pages of threads & you should spot it.
Well, it would be a 1/2 wave sealed line anyway. How well that would work would depend on the eventual design. Fine for midrange duties, but the 126 really does need a horn if you want to keep excursion under control.
Thanks Scott. I read some of the recent posts and read MJK's quarterwave website. I played with Hornresp software and even with instructions, could not understand it. Yes, a true horn action would be too large for even a very small room. I looked at plans on the Frugalhorn website, and it is difficult to know the advantages of each design other than looks and size. It seems few have built enough different designs for direct comparison.
Currently, the FE126E drivers are in 6.7L bass reflex enclosures tuned to 92Hz (-3dB at 90Hz). I am using this as a starting point for comparison. The Fostex recommended simple bass reflex design tuned for 61Hz seems very low, especially considering the average parameters measured by Planet10 (Fs = 92Hz). Overall, they sound very good from about 200 Hz to 15KHz. Beyond subjective listening, I do not have the equipment to do a frequency response measurement.
Would the best compromise be a 1/4 wave TL with some sort of horn action, i.e. flared mouth?
Currently, the FE126E drivers are in 6.7L bass reflex enclosures tuned to 92Hz (-3dB at 90Hz). I am using this as a starting point for comparison. The Fostex recommended simple bass reflex design tuned for 61Hz seems very low, especially considering the average parameters measured by Planet10 (Fs = 92Hz). Overall, they sound very good from about 200 Hz to 15KHz. Beyond subjective listening, I do not have the equipment to do a frequency response measurement.
Would the best compromise be a 1/4 wave TL with some sort of horn action, i.e. flared mouth?
FWIW, the best compromises for the FE126E (IMO) are either
a) a long path, high gain horn -but not taken too low. Or
b) Get it in either a baffle or sealed box, and support everything below about 500Hz with a suitable 15in HE woofer.
Yeah, the number of boxes on the FH site are a bit confusing -we'll have to have a bit of a tidy up at some point I reckon. Pick of the bunch on there for the 126, in no particular order are probably the Frugel-horn itself, Ron's A126, or the Saburo cabinets.
a) a long path, high gain horn -but not taken too low. Or
b) Get it in either a baffle or sealed box, and support everything below about 500Hz with a suitable 15in HE woofer.
Yeah, the number of boxes on the FH site are a bit confusing -we'll have to have a bit of a tidy up at some point I reckon. Pick of the bunch on there for the 126, in no particular order are probably the Frugel-horn itself, Ron's A126, or the Saburo cabinets.
As a reminder, there was that one made of PVC-- not the seraphim at
the TL site --but another stand mount PVC horn made in sections....
there were even plans on how to make it, but I tried to find the
link and couldn't come up with it. Line length winds up to be about 81 inches or so. Very sculptural. I can't remember if the FE126 was used.
the TL site --but another stand mount PVC horn made in sections....
there were even plans on how to make it, but I tried to find the
link and couldn't come up with it. Line length winds up to be about 81 inches or so. Very sculptural. I can't remember if the FE126 was used.
Interesting application of pvc pipe for straight transmission line.
I suppose you could do all the length calculations and approximate, but could you also tune it by ear? I.e., mount the driver on a baffle at one end of a mass loaded Sonotube. Then make a sliding baffle with a 2" port at the other end. Basically three mdf discs, 1 slightly smaller than pipe with firm weatherstrip around the circumference for a press fit sandwiched between 2 discs same size as the pipe. Apply a signal to the driver one octave below Fs (or lowest possible) using a frequency generator and tune for maximum resonance by sliding the rear baffle.
Keep in mind I have yet to build a TL speaker, so this is theoretical. Seems very simple to do though. A friend tuned his piano by starting with a tuning fork as a reference and adjusting the strings successively along the music scale by listening to interference created when sounds were not at 1/2 octave intervals. However, he is an engineer and violinist with a good ear.
I suppose you could do all the length calculations and approximate, but could you also tune it by ear? I.e., mount the driver on a baffle at one end of a mass loaded Sonotube. Then make a sliding baffle with a 2" port at the other end. Basically three mdf discs, 1 slightly smaller than pipe with firm weatherstrip around the circumference for a press fit sandwiched between 2 discs same size as the pipe. Apply a signal to the driver one octave below Fs (or lowest possible) using a frequency generator and tune for maximum resonance by sliding the rear baffle.
Keep in mind I have yet to build a TL speaker, so this is theoretical. Seems very simple to do though. A friend tuned his piano by starting with a tuning fork as a reference and adjusting the strings successively along the music scale by listening to interference created when sounds were not at 1/2 octave intervals. However, he is an engineer and violinist with a good ear.
Scottmoose said:
And a dark horse... Harvey. Just got an email from someone who was really thrilled with his Harvey with FE126eN ... and i don't take his comments lightly, as his main system has Feastrex...
Aangus says he is going to have his Harveys for us to hear at the upcomong diyFEST.
dave
Back when I toyed with different prototypes of commercial designs for the 126eN and other "N's" under BUF for a while, I just kept getting a simple, but important smile put on my face. It is a musical aesthetic arrived at the hard way (the only way).
Ever try series magnets to get even more juice out of her? Perhaps with the flattened response it could hello nasty.
Still hoping to meet you some day, Dave. Feel free to come out here and visit us in Yamanashi any time. With time, I am looking forward to seeing even more community driven Feastrex cabinet designs. Given the number of people working on them right now, it is a comforting type of inevitability.
-Clark
Ever try series magnets to get even more juice out of her? Perhaps with the flattened response it could hello nasty.
Still hoping to meet you some day, Dave. Feel free to come out here and visit us in Yamanashi any time. With time, I am looking forward to seeing even more community driven Feastrex cabinet designs. Given the number of people working on them right now, it is a comforting type of inevitability.
-Clark
blumenco said:
even for those of us in that grey area between DIY and commercial ventures, I think it's fair to say it's all about the smile - regardless of those who don't "get it"
could you elaborate on that? I'm either dense or forgetful of an earlier description of the subject (of course those aren't mutually exclusive conditions, are they?)
now that'd be an interesting experience: "the road to Yamanshi" a la Hope/Crosby - talk about a culture clash
all we need is some prototyping sample drivers
series magnet...
Sorry for the vaguaries.
Series magnet would be added to the back of the driver, same looking as a bucking magnet, but at different polarity (does not need glue, and serves an entirely different purpose).
The driver would grab the magnet out of your hand, and whammo, you got extra power. Now as to whether or not this is a good thing, anybody's guess.
For instance, it is pretty much the only difference between most "S" and normal edition Fostex drivers, especially the older ones before the "R" came around. they would just stick an extra magnet on the back, and charge more than twice as much (for what was admittably a significant improvement in the sound).
Many high end manufacturers will just take an off the shelf SEAS driver or something and throw extra magnets on the back of it until they have enough efficiency to match whatever tweeter and woofer they are using. Added resolution = icing on the cake. I have been in alot of factories and have actually SEEN people do this. Fairly common practice.
With a crossover, any anomolies resulting from a very powerful magnetic circuit could be tailored out by only using the "sweet spot" of the resultant driver. Full rangers don't have that luxury, but even then, a nice, powerful magnet can be a good or bad thing and it has different effects on different drivers. For instance, Feastrex drivers actually GAIN some bass slam with the higher end magnetic circuits. Counterintuitive to be sure. So it goes to show that you never know till you try.
I am pretty sure that ferrite magnets would be easy enough to remove later on down the road. If you try to use a chunk of Alnico, your job gets tougher, as the magentic storage is -significantly- higher.
-Clark
Sorry for the vaguaries.
Series magnet would be added to the back of the driver, same looking as a bucking magnet, but at different polarity (does not need glue, and serves an entirely different purpose).
The driver would grab the magnet out of your hand, and whammo, you got extra power. Now as to whether or not this is a good thing, anybody's guess.
For instance, it is pretty much the only difference between most "S" and normal edition Fostex drivers, especially the older ones before the "R" came around. they would just stick an extra magnet on the back, and charge more than twice as much (for what was admittably a significant improvement in the sound).
Many high end manufacturers will just take an off the shelf SEAS driver or something and throw extra magnets on the back of it until they have enough efficiency to match whatever tweeter and woofer they are using. Added resolution = icing on the cake. I have been in alot of factories and have actually SEEN people do this. Fairly common practice.
With a crossover, any anomolies resulting from a very powerful magnetic circuit could be tailored out by only using the "sweet spot" of the resultant driver. Full rangers don't have that luxury, but even then, a nice, powerful magnet can be a good or bad thing and it has different effects on different drivers. For instance, Feastrex drivers actually GAIN some bass slam with the higher end magnetic circuits. Counterintuitive to be sure. So it goes to show that you never know till you try.
I am pretty sure that ferrite magnets would be easy enough to remove later on down the road. If you try to use a chunk of Alnico, your job gets tougher, as the magentic storage is -significantly- higher.
-Clark
FE126e and extra magnets
Hi blumenco,
My understanding is that if the extra magnet sticks to the existing magnet, you would be REDUCING the flux in the voice coil gap. Reducing the flux in the gap may improve bass in some enclosures by reducing Qes/Qts, but this is not usually desirable for a backhorn. If you wish to increase the magnetic flux in the gap, you need to glue an extra "bucking" magnet on the back, oriented so it is repelled by the existing magnet. The bucking magnet MAY improve efficiency of the driver. I like my FE126e in a small sealed box, but in this configuration has (almost) NO bass. The magnet on the FE126e is a monster already and flux is probably close to optimum.
Regards Philip
Hi blumenco,
My understanding is that if the extra magnet sticks to the existing magnet, you would be REDUCING the flux in the voice coil gap. Reducing the flux in the gap may improve bass in some enclosures by reducing Qes/Qts, but this is not usually desirable for a backhorn. If you wish to increase the magnetic flux in the gap, you need to glue an extra "bucking" magnet on the back, oriented so it is repelled by the existing magnet. The bucking magnet MAY improve efficiency of the driver. I like my FE126e in a small sealed box, but in this configuration has (almost) NO bass. The magnet on the FE126e is a monster already and flux is probably close to optimum.
Regards Philip
Er, reducing Qe / total Q would tend to indicate an increase in relative motor power. As chances it will therefore have a relatively high mass corner roll-off, a horn of some description would be one of the more effective ways of getting some LF out of it (without resorting to Eq).
Motor-size isn't always all that accurate a guide to how much power is actually available. The 126 for e.g. doesn't have anything like the raw grunt Fostex claim with a lower Bl factor & higher Qe, in the high 0.3 regions rather than the high 0.2 stated in the official specs. The same basically goes for the Feastrex units (the practice, not the claims as they don't have any official parameters). They certainly have a high Bl factor, but their Qe, which tends to dominate a driver's resonant behaviour, is only middling.
Adding additional magnets could be a double edged sword. You may get some more efficiency, but you'd be changing the entire behaviour of the driver, lowering Qe, raising Bl & therefore its resonant behaviour, roll-off etc. Could be useful, could equally cause problems if the change isn't accounted for in the design of the speaker itself.
One quick question though -I'm not entirely sure which Fostex drivers Clark is refering to with the "S" and "R" bits & them simply adding a second magnet to the back of a regular unit -Clark?
Motor-size isn't always all that accurate a guide to how much power is actually available. The 126 for e.g. doesn't have anything like the raw grunt Fostex claim with a lower Bl factor & higher Qe, in the high 0.3 regions rather than the high 0.2 stated in the official specs. The same basically goes for the Feastrex units (the practice, not the claims as they don't have any official parameters). They certainly have a high Bl factor, but their Qe, which tends to dominate a driver's resonant behaviour, is only middling.
Adding additional magnets could be a double edged sword. You may get some more efficiency, but you'd be changing the entire behaviour of the driver, lowering Qe, raising Bl & therefore its resonant behaviour, roll-off etc. Could be useful, could equally cause problems if the change isn't accounted for in the design of the speaker itself.
One quick question though -I'm not entirely sure which Fostex drivers Clark is refering to with the "S" and "R" bits & them simply adding a second magnet to the back of a regular unit -Clark?
I think we are all crossing wires here a bit. Sorry for that.
Check out the 108ES , and 108ESII, 168ES, 208ES, and even the much older special ed. drivers.
S is for special edition, R is for radial (papermaking).
I think that bucking magnet (N-N) is for shielding, series is for added power (N-S), but truely, I could have my head up my tush (I have not researched much).
It will change everything, but might be fun, like varying voltage in a field coil. Also might be pretty boring and frustrating. I never got to try it.
The 5 inch SEAS midrange driver I am speaking of had something like 6 extra magnets stuck to the back of it without glue. assuming that this is series? they claimed that it raised the efficiency to 95 DB or thereabouts.
-Clark
Check out the 108ES , and 108ESII, 168ES, 208ES, and even the much older special ed. drivers.
S is for special edition, R is for radial (papermaking).
I think that bucking magnet (N-N) is for shielding, series is for added power (N-S), but truely, I could have my head up my tush (I have not researched much).
It will change everything, but might be fun, like varying voltage in a field coil. Also might be pretty boring and frustrating. I never got to try it.
The 5 inch SEAS midrange driver I am speaking of had something like 6 extra magnets stuck to the back of it without glue. assuming that this is series? they claimed that it raised the efficiency to 95 DB or thereabouts.
-Clark
Might well do, although it'd change most of the other parameters in the process, so that would need to be accounted for.
You wouldn't happen to have any links for some of these special ed. drivers by any chance would you? I've got those for the current regular range of course inc. the ESigmas & also for most of the older units, but the only special edn. pdf spec sheets I've ever managed to get hold of are for the 138ES-R, 166ES-R & 206ES-R.
You wouldn't happen to have any links for some of these special ed. drivers by any chance would you? I've got those for the current regular range of course inc. the ESigmas & also for most of the older units, but the only special edn. pdf spec sheets I've ever managed to get hold of are for the 138ES-R, 166ES-R & 206ES-R.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.