bear, the problem isn't as bad as it may seem because I don't have to worry much about trying to "align" everything like you do with music. I don't care at all if my "flute" and my "drum" are out of sync.
When setting up for a power run, I hit the chamber with white noise at about 120 dB and look at the spectrum produced. Then I go back to Audacity and boost the lows or attenuate the peaks until I get the spectrum I want and then kick it up to whatever is required for the run. Most of the chamber shape is to keep the noise contained. Curved surfaces are much more rigid than flat surfaces. So, I'm able to knock off the top 45 dB with no "acoustic" insulation at all. This simplifies a lot of things! All those curves also keep the noise banging around for several passes at my test object, so it saves on the energy bill.
BillyDoc
When setting up for a power run, I hit the chamber with white noise at about 120 dB and look at the spectrum produced. Then I go back to Audacity and boost the lows or attenuate the peaks until I get the spectrum I want and then kick it up to whatever is required for the run. Most of the chamber shape is to keep the noise contained. Curved surfaces are much more rigid than flat surfaces. So, I'm able to knock off the top 45 dB with no "acoustic" insulation at all. This simplifies a lot of things! All those curves also keep the noise banging around for several passes at my test object, so it saves on the energy bill.
BillyDoc
Cool, thanks for the photos. The flat strip is typical, the braid isn't - at least not in the drivers I've used. And the flat is often where the break is, even in normal abuse.
So you are going from subs to horns at 300Hz, no midrange driver? That would be hard in any system without special drivers and big horns. The old Altec "Giant Voice" stuff might stand a chance, on big horns.
Did you have to EQ the bottom range of the horns heavily? I'm just thinking that since that horn can't really do 300Hz, you'd have to boost the heck out of it down there. Thus the trouble with the broken drivers. A horn that actually loads the driver down to 300Hz would need a lot less power for a flat response. See what I mean? That's the area I would look in first.
Or maybe you need one of these:
YouTube - Vortex Cannon! - Bang Goes the Theory Preview - BBC One
YouTube - Jem v. the US Navy's latest weapon - Bang Goes the Theory - BBC One
The LRAD guys are in San Diego.
So you are going from subs to horns at 300Hz, no midrange driver? That would be hard in any system without special drivers and big horns. The old Altec "Giant Voice" stuff might stand a chance, on big horns.
Did you have to EQ the bottom range of the horns heavily? I'm just thinking that since that horn can't really do 300Hz, you'd have to boost the heck out of it down there. Thus the trouble with the broken drivers. A horn that actually loads the driver down to 300Hz would need a lot less power for a flat response. See what I mean? That's the area I would look in first.
Or maybe you need one of these:
YouTube - Vortex Cannon! - Bang Goes the Theory Preview - BBC One
YouTube - Jem v. the US Navy's latest weapon - Bang Goes the Theory - BBC One
The LRAD guys are in San Diego.
Good advice, panomaniac. I can't change the horns for physical reasons, but I've been looking hard at that BMS4592ND-MID exactly because it goes all the way down to 300 Hz. If it has the excursion distance to slap the phase plate without ripping out the lead I think that is going to solve my problem. I can't remember the numbers now, but I did have to equalize fairly heavily. Same on the driver with the braid. The difference is (I think) that the driver with the braid didn't pull the braid up taut at the end of its excursion. I wrote to BMS to ask about details, but haven't heard back yet.
Ummm... the braid takes more thermally... quite standard... the failure of the kapton + trace seems to be at the point where it meets the inner edge of the surround/diaphragm interface. Looks like it burned out. Fuse wire.
Show us the "old compression driver"? A jpeg would likely identify it or serve to confuse. Looks like a phenolic surround...
I don't think ur understanding what i said about standing waves in the enclosure.
Why don't you have a sound proofed ROOM to do this work??
The chamber will likely cause the drivers to blow up no matter what you do, assuming you have both woofers and HF drivers running at the same time...
You do know that you need a HIGH SLOPE highpass on the tweeters AND a HIGH SLOPE LOW PASS as well??
I'd find an outside space or a big big room as being much much better... that and some layers of hearing protection at all times...
A horn that does NOT load down to 300Hz. will be dropping like a rock (24db/oct) below the mouth/flare rate frequency... so all the EQ in the world won't bring it back up... you'll just fry the VC...
IF you clip those Fender amps, ur dead meat too. Better have a 'scope on it.
Are you sure you're being paid to do this work?? 🙄
_-_-bear
Show us the "old compression driver"? A jpeg would likely identify it or serve to confuse. Looks like a phenolic surround...
I don't think ur understanding what i said about standing waves in the enclosure.
Why don't you have a sound proofed ROOM to do this work??
The chamber will likely cause the drivers to blow up no matter what you do, assuming you have both woofers and HF drivers running at the same time...
You do know that you need a HIGH SLOPE highpass on the tweeters AND a HIGH SLOPE LOW PASS as well??
I'd find an outside space or a big big room as being much much better... that and some layers of hearing protection at all times...
A horn that does NOT load down to 300Hz. will be dropping like a rock (24db/oct) below the mouth/flare rate frequency... so all the EQ in the world won't bring it back up... you'll just fry the VC...
IF you clip those Fender amps, ur dead meat too. Better have a 'scope on it.
Are you sure you're being paid to do this work?? 🙄
_-_-bear
Certainly agree. You are just asking for trouble with the horn/frequency mismatch. It does not matter what driver you use on the Selenium HL14-50 horn, it's going to drop off fast under 1KHz. So from 300Hz to 1K you have a driver without proper loading that you are having to crank an extra 20-30dB into to get it near flat. Ouch!
That'd a LOT more power.I'd wager that above 1.5Khz the horns are getting 1/20th or less the full power you are throwing at them. Because they are working in the range they should.
There may be some bass power coming back up the horns, but I really agree with Bear, the low frequencies in this horn are the main problem.
panomaniac, bear, you guys would be exactly right with most horns. They wouldn't load the driver enough and it would be "flapping in the wind" under power at low frequencies. But the Selenium HL14-50 DOES load down to 300 Hz. I've attached the data sheet. It's one of the main reasons I picked it! The impedance only drops about 6 dB from 500 Hz down to 300, and the horn is rated at 500 Hz. That, and the compact size, which is a very strong consideration in this design.
It definitely looses response as it drops in frequency, though. But that probably has more to do with the driver they tested with than the horn.
My working assumption is still that the energy from the woofers is backing up the horn and driving the compression driver diaphragms with large excursions. Excursions that exceeded the ribbon connection's ability to follow because they didn't put enough slack in there.
I still don't have any response from BMS.
BillyDoc
It definitely looses response as it drops in frequency, though. But that probably has more to do with the driver they tested with than the horn.
My working assumption is still that the energy from the woofers is backing up the horn and driving the compression driver diaphragms with large excursions. Excursions that exceeded the ribbon connection's ability to follow because they didn't put enough slack in there.
I still don't have any response from BMS.
BillyDoc
Sorry Billy, have to completely disagree with you here. That ain't no 500Hz horn! It's just a 6" horn, for pete's sake. A real 500Hz horn will be 4x the size. Does not matter what driver you put on it.
Now I truly like Selenium products and their measurement department should win a prize, but that horn is no good below ~1Kz. I wouldn't use it under 2K. Look at the graph, see how it falls fast under 1K. There is a little bump at 600Hz, but that's an artifact. Note that they state 500Hz at -10dB. That's not a range you want to be in, especially at the extreme powers you are using. That is the practical, real world truth.
The bass coming back up the horns is a red herring. Tho it may have an effect, it is not the main cause of your trouble. Running those horns/drivers with a crossover point of 300Hz is. Solve that one first.
I know you have a lot invested in where you now are with this, but it isn't going to work as-is. You've seen that. It may be hard to get past, but you'll have to, to make it work. "Maybe" a more rugged driver will survive running that low on the small horn, but it's still going to be working harder than it should - much harder.
Now I truly like Selenium products and their measurement department should win a prize, but that horn is no good below ~1Kz. I wouldn't use it under 2K. Look at the graph, see how it falls fast under 1K. There is a little bump at 600Hz, but that's an artifact. Note that they state 500Hz at -10dB. That's not a range you want to be in, especially at the extreme powers you are using. That is the practical, real world truth.
The bass coming back up the horns is a red herring. Tho it may have an effect, it is not the main cause of your trouble. Running those horns/drivers with a crossover point of 300Hz is. Solve that one first.
I know you have a lot invested in where you now are with this, but it isn't going to work as-is. You've seen that. It may be hard to get past, but you'll have to, to make it work. "Maybe" a more rugged driver will survive running that low on the small horn, but it's still going to be working harder than it should - much harder.
Hey panomaniac, don't apologize! Technical disagreements are good for learning, for everyone.
Keep in mind what my goal is here, which is to flood a chamber with destructive noise in order to see if some piece of equipment can survive it. This is very different from producing music, as you no doubt know. And you're right, the signal does indeed drop below 1 KHz, but I can boost it at the source to compensate. Messy, but it works. The BMS drivers are designed for a 300 Hz crossover, so won't be working as hard at that frequency on the same horn.
BUT, to get back to a possible fix, what do you suggest? A spacer between the driver and horn to lengthen the air column? What? I think I could get a 3" spacer in there, but have to measure some stuff to be sure.
Keep in mind what my goal is here, which is to flood a chamber with destructive noise in order to see if some piece of equipment can survive it. This is very different from producing music, as you no doubt know. And you're right, the signal does indeed drop below 1 KHz, but I can boost it at the source to compensate. Messy, but it works. The BMS drivers are designed for a 300 Hz crossover, so won't be working as hard at that frequency on the same horn.
BUT, to get back to a possible fix, what do you suggest? A spacer between the driver and horn to lengthen the air column? What? I think I could get a 3" spacer in there, but have to measure some stuff to be sure.
panomaniac, to answer my own question one way, it looks like lengthening the horn might do the trick nicely . . . if I can fit a spacer in there . . . which is in no way certain.
Here's what I did. Assuming a quarter-wave horn would work pretty well, that is a horn long enough to resonate at a quarter of the wavelength of my lowest frequency, then 343 meters (speed of sound) divided by 300 (lowest frequency) divided by four gives me 286 mm for the quarter wave horn. The existing horn is 131 mm long, so a 155 mm spacer between the horn and the driver should do the job. Do you concur? Have you ever heard of an eighth-wave horn? It would be so much easier!
And panomaniac, thanks for beating me up enough to actually do the calculation. I really didn't think a spacer would make that much difference in any length I could actually do and was in denial about the need. I think I'm over that now . . .
BillyDoc
Here's what I did. Assuming a quarter-wave horn would work pretty well, that is a horn long enough to resonate at a quarter of the wavelength of my lowest frequency, then 343 meters (speed of sound) divided by 300 (lowest frequency) divided by four gives me 286 mm for the quarter wave horn. The existing horn is 131 mm long, so a 155 mm spacer between the horn and the driver should do the job. Do you concur? Have you ever heard of an eighth-wave horn? It would be so much easier!
And panomaniac, thanks for beating me up enough to actually do the calculation. I really didn't think a spacer would make that much difference in any length I could actually do and was in denial about the need. I think I'm over that now . . .
BillyDoc
I think you've just answered your own question. 😉 And therein lies the problem. It's not the length of the horn that is the big problem here (it a small problem) it's the mouth size. Does not matter whether you want music or noise, a 6" horn isn't going to give you anything below about 1K. So you have to boost like crazy. That means beaucoup more power going to the drivers, just to get the level you need. Same driver on a bigger horn would not need all that boost and would not burn up as fast - maybe not at all.
Yes, a driver designed for a lower cut-off might help, but it's designed for a lower cut-off on a bigger horn. Those darn Laws of Physics again.
Now, if you could build an adapter to put something like the BMS or MCM driver onto that horn, you'd have a more rugged driver. It's still going to need a big boost below 1Khz because the Selenuim horn is so small. But it "might" survive better than the Pyle. It's still going to be working very much at the ragged edge, tho. Not the best fix, but maybe one that could get you by.
You can't come up even an octave on the subs? Or do they roll off real low?
Hi panomaniac,
First let me comment that the subs are already pushing it at 300 Hz. These are big Kicker Solo X - 18s.
Your comments about horn mouth size are more relevant to the problem.
I've been going over an excellent tutorial of sorts by John H Sheerin that's over here: Horn Design
Mr. Sheerin makes the same points about mouth size that you do, but he adds this:
"If your horn is designed to hang 1000 feet above a cornfield, then it will effectively be radiating into full space. Leach's model assumes that the mouth of the horn will be infinite in this case. Of course you cannot build a horn with an infinitely large mouth, but a mouth that has a circumference equal to the wavelength of the lowest frequency to be reproduced has been agreed on as being close enough to infinite. If you put the horn on the ground, then it will be radiating into half space. The ground provides an acoustic boundary. In this case, a mouth with half the area of the full space mouth will be close enough to an infinite mouth. The area can again be cut in half for a horn placed on the ground and against a wall. If the mouth is placed at the junction of two walls and a floor, then the mouth area can be 1/8 the size of the full space mouth and still be considered large enough."
You couldn't know this, but the horns in my machine are in groups of two facing into the long axis of the chamber, placed on the edge of the chamber. That is, they have a wall adjacent. Moreover, as they are in groups of two in close physical proximity and also all in phase I think that the "adjacent" horn will provide a second acoustic boundary. So, instead of needing a horn circumference at least as large as the wavelength of my lowest frequency (1.14 meters), I only need one a quarter of that (286 mm). The HL14-50 has a 140 mm diameter mouth, so a circumference of 440 mm. I think that part is OK! It also explains why it didn't seem to require much boost at all in practice.
Make sense? Or did I screw something up?
BillyDoc
First let me comment that the subs are already pushing it at 300 Hz. These are big Kicker Solo X - 18s.
Your comments about horn mouth size are more relevant to the problem.
I've been going over an excellent tutorial of sorts by John H Sheerin that's over here: Horn Design
Mr. Sheerin makes the same points about mouth size that you do, but he adds this:
"If your horn is designed to hang 1000 feet above a cornfield, then it will effectively be radiating into full space. Leach's model assumes that the mouth of the horn will be infinite in this case. Of course you cannot build a horn with an infinitely large mouth, but a mouth that has a circumference equal to the wavelength of the lowest frequency to be reproduced has been agreed on as being close enough to infinite. If you put the horn on the ground, then it will be radiating into half space. The ground provides an acoustic boundary. In this case, a mouth with half the area of the full space mouth will be close enough to an infinite mouth. The area can again be cut in half for a horn placed on the ground and against a wall. If the mouth is placed at the junction of two walls and a floor, then the mouth area can be 1/8 the size of the full space mouth and still be considered large enough."
You couldn't know this, but the horns in my machine are in groups of two facing into the long axis of the chamber, placed on the edge of the chamber. That is, they have a wall adjacent. Moreover, as they are in groups of two in close physical proximity and also all in phase I think that the "adjacent" horn will provide a second acoustic boundary. So, instead of needing a horn circumference at least as large as the wavelength of my lowest frequency (1.14 meters), I only need one a quarter of that (286 mm). The HL14-50 has a 140 mm diameter mouth, so a circumference of 440 mm. I think that part is OK! It also explains why it didn't seem to require much boost at all in practice.
Make sense? Or did I screw something up?
BillyDoc
Billy Doc, I sent you a PM a few days ago - look at ur PMs??
Sorry to tell you that you simply do not understand the problems.
You can NOT just boost the power to the seleniums.
Lengthening this horn will have NO effect on the freq it rolls off at.
The placement in a sealed chamber might help but that has to be determined by testing, not guessing. The horn in question is rated to 500 Hz. and actually only works to 1kHz. It is unclear that the drivers can work at full power down to 300Hz, no matter what the horn. Few can. Maybe none in current production.
The energy from the woofer coming back into the diaphragms of the compression drivers IS NOT a "red herring" rather it is a show stopper. Won't work. Period.
You did not say why you could not use a large open space or room.
Producing broadband high SPL inside a small enclosure is problematic.
It's going to be problematic beyond this simplistic approach, IF it can be done at all.
I think you are way over your head, and will just continue to blow stuff up indefinitely.
Since you are apparently working for a company, doing industrial testing, a review of the literature on this sort of testing might be in order, or the hiring of a consultant who has experience in high SPL applications.
My take on it is to first get rid of the small enclosed chamber since it creates problems that likely can never be overcome.
After that it is still very very hard to get that sort of broadband SPL...
Send check.
_-_-bear
Sorry to tell you that you simply do not understand the problems.
You can NOT just boost the power to the seleniums.
Lengthening this horn will have NO effect on the freq it rolls off at.
The placement in a sealed chamber might help but that has to be determined by testing, not guessing. The horn in question is rated to 500 Hz. and actually only works to 1kHz. It is unclear that the drivers can work at full power down to 300Hz, no matter what the horn. Few can. Maybe none in current production.
The energy from the woofer coming back into the diaphragms of the compression drivers IS NOT a "red herring" rather it is a show stopper. Won't work. Period.
You did not say why you could not use a large open space or room.
Producing broadband high SPL inside a small enclosure is problematic.
It's going to be problematic beyond this simplistic approach, IF it can be done at all.
I think you are way over your head, and will just continue to blow stuff up indefinitely.
Since you are apparently working for a company, doing industrial testing, a review of the literature on this sort of testing might be in order, or the hiring of a consultant who has experience in high SPL applications.
My take on it is to first get rid of the small enclosed chamber since it creates problems that likely can never be overcome.
After that it is still very very hard to get that sort of broadband SPL...
Send check.
_-_-bear
Last edited:
Sure. I figured as much. Having a wall or baffle will help. We just don't know how much. I doubt the horns are working into 1/8th space, tho. The big question is - How much did you have to boost below 1KHz to get where you want to be? You haven't said. That has to be talen into account, as well as the horn not loading the diaphragm. But as the 'frams did not disintegrate, just burned out, I suspect too much power.
What you are doing is a big mis-match. An 18" sub that hardly gets above 100Hz mated to a 6" horn. Gonna be hard to make that work, whatever the system! 😛 And with your extreme needs, well.......
While I agree that it may be a problem, I don't think at this point it's the main problem. Overpowering the drivers is. (see above). But of course we don't know, we are just guessing. One way to test would be to run the woofers only - with good 'frams in the Pyle compression drivers. (you'd only need 1 good one, really) Do they break if not under power? Or under just a few watts? If not, it probably isn't the return bass causing the damage.
Pano, I dunno what you are thinking about... just consider what happens when you pump the interior of this sealed enclosure to say 2 atmospheres... and only have the horns + compression drivers present? What do you think the pressure will do to the diaphragms??
_-_-bear
_-_-bear
Perhaps if the compression driver were open back inside the same enclosure,
low frequency (below horn cutoff) coming from the other drivers, would have
less damaging effect? Wild guessing here, don't actually know...
low frequency (below horn cutoff) coming from the other drivers, would have
less damaging effect? Wild guessing here, don't actually know...
Last edited:
yeah, perhaps, but the phase relationship would suddenly get very critical - you could make it work fine at some frequencies, but not at others... again the closed, tight space is a big big problem...
_-_-bear
_-_-bear
Hey Bear.
I dunno. Certainly it could be a problem. BillyDoc IS trying to break stuff with sound, after all! But at the moment I don't think that is what killed the drivers.
As you said yourself:
Did it break from too much current or from too much excursion, or both? The rest of the diaphragm looks undamaged.
There are a couple of things to do to test it. And if it were my rig, I'd test it.
1: Run the woofers only - and at least 1 good compression driver 'fram in place. Short the 'fram or connect it to the amp (no signal) so that it has some damping. Does the SPL from the woofers destroy it when it's not being powered?
2:Run only the compression drivers at the full level. Do they break? If so, it's not the woofers causing the problem. (Check for amp clipping!)
BTW - BillyDoc. What the heck do you use to measure SPLs this high? Must be a hell of a sound meter. Or is it some sort of strain gauge?
I dunno. Certainly it could be a problem. BillyDoc IS trying to break stuff with sound, after all! But at the moment I don't think that is what killed the drivers.
As you said yourself:
(emphasis mine)
Did it break from too much current or from too much excursion, or both? The rest of the diaphragm looks undamaged.
There are a couple of things to do to test it. And if it were my rig, I'd test it.
1: Run the woofers only - and at least 1 good compression driver 'fram in place. Short the 'fram or connect it to the amp (no signal) so that it has some damping. Does the SPL from the woofers destroy it when it's not being powered?
2:Run only the compression drivers at the full level. Do they break? If so, it's not the woofers causing the problem. (Check for amp clipping!)
BTW - BillyDoc. What the heck do you use to measure SPLs this high? Must be a hell of a sound meter. Or is it some sort of strain gauge?
Morning Gentlemen,
Let me answer a few questions:
The correction in the third-octave just above 300 Hz was about 6 - 7 dB of extra power to flatten things out, tapering to nothing in 3 or 4 third octave steps.
The meter used is a Larson Davis LXT model with a high level microphone originally designed to measure artillery noise. I forget how high it goes, but definitely above 140 dB. The SL meter also has internal software for the third-octave breakdown, dump to computer, etc., and is calibrated to NIST standard. It wasn't cheap!
We haven't found any indication of a burnt out voice coil at all. All damage seems to be purely physical, and consistent with excessive diaphragm excursion. The way the Pyles are constructed the back side of the diaphragm is open to ambient pressure via cracks between several loosely fitted parts, and has more than enough room to travel to destruction. If you press on this diaphragm from the horn side you can see it come back and simultaneously twist on the horn axis. It quickly pulls the ribbon leads up tight . . . and that is what I think broke them off. It's a big diaphragm so just from a "pneumatic" point of view capable of developing a lot of force. On the other hand, it takes quite a lot of force to displace it as well, so I can't be sure. Instantaneous pressures can't be all that high in absolute terms, and these diaphragms are pretty stiff.
Bear, doing this stuff in the open just isn't an option. What would the neighbors think? Or OSHA! Plus, the enclosure does help to raise the internal pressures. As to longevity, the test setup prior to this one had been in operation since 1995 at more or less the same sound pressures and no failures at all. That one had an internal volume of only 6 cubic feet.
I'm still trying to get in contact with the BMS rep, but no call-back yet.
BillyDoc
Let me answer a few questions:
The correction in the third-octave just above 300 Hz was about 6 - 7 dB of extra power to flatten things out, tapering to nothing in 3 or 4 third octave steps.
The meter used is a Larson Davis LXT model with a high level microphone originally designed to measure artillery noise. I forget how high it goes, but definitely above 140 dB. The SL meter also has internal software for the third-octave breakdown, dump to computer, etc., and is calibrated to NIST standard. It wasn't cheap!
We haven't found any indication of a burnt out voice coil at all. All damage seems to be purely physical, and consistent with excessive diaphragm excursion. The way the Pyles are constructed the back side of the diaphragm is open to ambient pressure via cracks between several loosely fitted parts, and has more than enough room to travel to destruction. If you press on this diaphragm from the horn side you can see it come back and simultaneously twist on the horn axis. It quickly pulls the ribbon leads up tight . . . and that is what I think broke them off. It's a big diaphragm so just from a "pneumatic" point of view capable of developing a lot of force. On the other hand, it takes quite a lot of force to displace it as well, so I can't be sure. Instantaneous pressures can't be all that high in absolute terms, and these diaphragms are pretty stiff.
Bear, doing this stuff in the open just isn't an option. What would the neighbors think? Or OSHA! Plus, the enclosure does help to raise the internal pressures. As to longevity, the test setup prior to this one had been in operation since 1995 at more or less the same sound pressures and no failures at all. That one had an internal volume of only 6 cubic feet.
I'm still trying to get in contact with the BMS rep, but no call-back yet.
BillyDoc
show me pix of the 2 chambers, and the earlier compression driver - either privately or posted.
_-_-bear
_-_-bear
Oh, OK. That's new info, then. From you descriptions and photo before it looked like a burn out of the ribbon going to the VC. You don't think it is?
Still; if it's over-excursion causing the damage, is the excursion caused by the power applied to the drivers, the bass SPL modulation, or both? At this point we don't know, it's all conjecture. I still think you should test.
Thanks for the info on the SPL meter, I'm sure it wasn't cheap!
When I was at Eglin AFB, we had entire hangers for testing in. One was a giant deep freeze - on demand arctic weather in Florida. The neighbors and OSHA were not involved. 😉
- Status
- Not open for further replies.