Hi Everyone,
I'm new here, but found this blog because I have been researching a nasty problem and found a lot of good information here already. But I could sure still use some help!
I'm trying to design a testing apparatus that needs to develop sound levels pretty close to what you would get if you were standing next to a Shuttle launch. Specifically, over 150 dB, but at frequencies from about 2 Hz up to 500 Hz. This will be "pink" noise, so great fidelity is not a requirement. I've found some speakers, Kicker SoloX 18s, that can take 10,000 watts . . . briefly. I only need to turn these guys on for about 10 seconds, max, and there will be four of them. Then they get a chance to cool down a bit.
My first thought was a full bridge (H-bridge) with some high current MOSFETS driven by a PWM scheme of some sort, and I did locate a few chips to do some of the heavy lifting on the front end of that. BUT you guys have a LOT more experience with high current amps than I do, so how would you go about this problem? Which chips would you use for the PWM? Can you recommend some MOSFETs for the H-bridge? Gate drivers? Am I crazy for trying this?
Thanks for your help!
BillyDoc
I'm new here, but found this blog because I have been researching a nasty problem and found a lot of good information here already. But I could sure still use some help!
I'm trying to design a testing apparatus that needs to develop sound levels pretty close to what you would get if you were standing next to a Shuttle launch. Specifically, over 150 dB, but at frequencies from about 2 Hz up to 500 Hz. This will be "pink" noise, so great fidelity is not a requirement. I've found some speakers, Kicker SoloX 18s, that can take 10,000 watts . . . briefly. I only need to turn these guys on for about 10 seconds, max, and there will be four of them. Then they get a chance to cool down a bit.
My first thought was a full bridge (H-bridge) with some high current MOSFETS driven by a PWM scheme of some sort, and I did locate a few chips to do some of the heavy lifting on the front end of that. BUT you guys have a LOT more experience with high current amps than I do, so how would you go about this problem? Which chips would you use for the PWM? Can you recommend some MOSFETs for the H-bridge? Gate drivers? Am I crazy for trying this?
Thanks for your help!
BillyDoc
There a a lot of subs that can take 10kw, Ground zero, kickers, Digital Designs, Audiopulse and maaaany more...
all that are used dB drag races, car audio...
If you think you will design 10kw amp yourself, well you have a looong way before that, it could be cheaper and way faster to buy one or two that will do that..
So how are they powered? 12vDC, 120/230vAC?
I mean you have some options on your hands
all that are used dB drag races, car audio...
If you think you will design 10kw amp yourself, well you have a looong way before that, it could be cheaper and way faster to buy one or two that will do that..
So how are they powered? 12vDC, 120/230vAC?
I mean you have some options on your hands
Billy!
I'd suggest using IGBTs (eg. IRGP50B60, maybe paralelled), or if you stay at MOSFET, then you have to use BCA topology.
Optimal supply voltage is somewhere around 400...550 V, so you need 8 ohm impedance. IR2113 is a good driver, however there is a problem with very narrow pulses, so you have to limit duty cycle!
I'd suggest using IGBTs (eg. IRGP50B60, maybe paralelled), or if you stay at MOSFET, then you have to use BCA topology.
Optimal supply voltage is somewhere around 400...550 V, so you need 8 ohm impedance. IR2113 is a good driver, however there is a problem with very narrow pulses, so you have to limit duty cycle!
14000 watts
http://www.labgruppen.com/products/fp_plus_series/c/fp_14000
2 Hz and 150 dB? Have you really thought about this? How does your speaker look like?
Curious, but what is your application?
http://www.labgruppen.com/products/fp_plus_series/c/fp_14000
2 Hz and 150 dB? Have you really thought about this? How does your speaker look like?
Curious, but what is your application?
BillyDoc,
are you talking about a dB drag race?
if yes, you need something to make only a loud noise. IMMO it's something different to audio devices. If you know words like "150 db" and "10000 watts", and you don't know words like THD, Xmax and clipping, you don't need an amplifier. You need something to make noise. And you're a lucky guy, because building a raw noise maker is simplier than building an good audio amplifier.
if not, you need it for audio. Maybe someone sells some good 10kW amplifiers in this world, buy one of them. It's simplier, cheap, less time-wasting and less dangerous. IMMO.
are you talking about a dB drag race?
if yes, you need something to make only a loud noise. IMMO it's something different to audio devices. If you know words like "150 db" and "10000 watts", and you don't know words like THD, Xmax and clipping, you don't need an amplifier. You need something to make noise. And you're a lucky guy, because building a raw noise maker is simplier than building an good audio amplifier.
if not, you need it for audio. Maybe someone sells some good 10kW amplifiers in this world, buy one of them. It's simplier, cheap, less time-wasting and less dangerous. IMMO.
If this isn't just a car SPL competition but some sort of outdoor noise testing a traditional amplifier/loudspeaker may not be the best appraoch. This may not be all that crazy if you think outside the box a bit. If you can live with the noise being a little "whiter" than pink, an amplifier isn't really needed. Ever look at the waveform for the "MLS impulse" used for typical gated loudspeaker response measurements? It's a series of pseudo-random spaced pulses of "unity" amplitude. If you limit the "bandwidth" to a kHz or so, pulse widths can get wider, well within what big-*** IGBTs can do. You end up needing a big +/- power supply (10kVA isolation trafo, rectifiers and caps), and IGBTs to slam it rail to rail driving the loudspeakers directly. If the woofers are horn loaded with a very long (50 foot) path length you could get efficiency down in the teens or lower. The horn wouldn't even need to be permanently assembled - build it so the panels pin together any time you set this rig up. To get to 500 Hz, it may, however, require a 2-way. The upper frequency horn would be easy to build with a high-EBP 10 or 12". The horns themselves would filter harmonics and aliasing frequencies significantly.
wg_ski
I was thinking at it.
big power supply, NE555 for square waveform output, signal trasformers for driving IGBT or high-voltage MosFET (on-off).
The same ancient NE555 may be used to modulate a low-frequency PWM signal, so you could have a very low-fidelity bass amp.
The "core" of that pseudo-amp is simple, but power supply and output filter cold be very difficult to build.
I was thinking at it.
big power supply, NE555 for square waveform output, signal trasformers for driving IGBT or high-voltage MosFET (on-off).
The same ancient NE555 may be used to modulate a low-frequency PWM signal, so you could have a very low-fidelity bass amp.
The "core" of that pseudo-amp is simple, but power supply and output filter cold be very difficult to build.
Wow! That was a lot of responses since last night! Let me try to answer some questions first, as I realize that this is not your usual living-room hi fi project.
Actually, I'm designing a test chamber that will be used to qualify components for an unnamed space industry contractor. These are components that must be rugged enough to withstand a rocket launch. Literally!
The chamber will have an internal volume of about 60 cubic feet and be made of heavy steel. There will be four Kicker 18" SoloX speakers at one end, plus eight midrange drivers. I need the power amp mostly for the Kickers, and it looks like four of the Labgruppen FP14000 pointed to by peranders will fill the bill nicely! Thanks peranders! I had no idea such big amps were commercially available. I couldn't find them when searching the web for some reason.
I have to use speakers so that I can shape the frequency spectrum in the test chamber, which I will be recording with two microphones, digitizing, and FFT analyzing via computer for subsequent analysis.
SPL drag races, eh? I wish I had some copies of photomicrographs of coclear hair cells after exposure to high noise (those are the sensitive elements in the microphones on each side of your head) to share with you. Normal hair cells are in neat, tidy rows that look something like wwwwwwwwwwwww, and after high noise exposure they look like a forest that has just been through a hurricane. Or maybe a stack of "pick-up" sticks. And they DON'T grow back! Just a thought.
Anyway, if the Labgruppen amps are'nt too pricy I'll be going with them. I've got to admit, though, that after reading quite a bit on this site yesterday I've kinda got the bug to try building one of these babies. I REALLY like the elegance of the class D approach! We'll see!
It's been a pleasure meeting you guys, and thanks for all the responses!
BillyDoc
Actually, I'm designing a test chamber that will be used to qualify components for an unnamed space industry contractor. These are components that must be rugged enough to withstand a rocket launch. Literally!
The chamber will have an internal volume of about 60 cubic feet and be made of heavy steel. There will be four Kicker 18" SoloX speakers at one end, plus eight midrange drivers. I need the power amp mostly for the Kickers, and it looks like four of the Labgruppen FP14000 pointed to by peranders will fill the bill nicely! Thanks peranders! I had no idea such big amps were commercially available. I couldn't find them when searching the web for some reason.
I have to use speakers so that I can shape the frequency spectrum in the test chamber, which I will be recording with two microphones, digitizing, and FFT analyzing via computer for subsequent analysis.
SPL drag races, eh? I wish I had some copies of photomicrographs of coclear hair cells after exposure to high noise (those are the sensitive elements in the microphones on each side of your head) to share with you. Normal hair cells are in neat, tidy rows that look something like wwwwwwwwwwwww, and after high noise exposure they look like a forest that has just been through a hurricane. Or maybe a stack of "pick-up" sticks. And they DON'T grow back! Just a thought.
Anyway, if the Labgruppen amps are'nt too pricy I'll be going with them. I've got to admit, though, that after reading quite a bit on this site yesterday I've kinda got the bug to try building one of these babies. I REALLY like the elegance of the class D approach! We'll see!
It's been a pleasure meeting you guys, and thanks for all the responses!
BillyDoc
A 10 second burst certainly wouldn't hurt a Labguppen amp. But its power supply might complain about that long a burst - usually burst testing is done for around 50ms on 120V lines. If it's strapped for 240V operation it may be ok with it (usually input current is limited per safety agency specs).
I have just sent devices out for vibration testing by a lab.
You get a nice Certificate you can pass onto the customer.
But I could see making a quick and dirty tester for small things by just gluing a board to a woofer. Or for larger things, take apart a few woofers and attached the voice coils to a platform at each corner. With a few layers you can get 3 axis shaking. No amp needed as wg_ski points out just pulses.
No sense turning the vibration into sound - there is no sound in space!
Big Bang is a myth - there was no air so nothing to go Bang in.
Yeah I know it's for the launch, but still there will be orbit adjusting rockets and vibration from those in space.
You get a nice Certificate you can pass onto the customer.
But I could see making a quick and dirty tester for small things by just gluing a board to a woofer. Or for larger things, take apart a few woofers and attached the voice coils to a platform at each corner. With a few layers you can get 3 axis shaking. No amp needed as wg_ski points out just pulses.
No sense turning the vibration into sound - there is no sound in space!
Big Bang is a myth - there was no air so nothing to go Bang in.
Yeah I know it's for the launch, but still there will be orbit adjusting rockets and vibration from those in space.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.