Previously you stated: " Used a 8.5 kW motor."
That's almost 3 dB more power used to reach a level an order of magnitude less than your theoretical calculations indicate, so you have "a tough row to hoe" ahead to bring theory into reality.
Also, never before have heard of VLF complaints from nuclear power plants (though VLF complaints from early wind turbines were common), who is complaining about what plant?
Last edited:
Ok, I read this (actually just skimmed it) and it's not a patent and the results are spectacularly underwhelming AND did not use a ported enclosure with different sized ports as you described.
A brief summary, if I was paying attention to the appropriate details is as follows.
The input section, which is not really important here:
- a racecar engine capable of a sustained 400 hp output was used
- the engine powered dual racecar superchargers to provide the airflow
- airflow flows through a feeding tube (very likely undersized, I didn't catch that part), past a surge tank and through a modulator device, all of which are more complicated than needed
The enclosure section, which is the important part of this discussion:
- a 2.4 cubic meter concrete lined enclosure
- a MASSIVELY UNDERSIZED port to external air was used (8 inches diameter, 20 feet in length, at one point they stepped that up to 3x8 inch diameter tubes, still massively undersized)
- tested with and without an internal divider to divide the enclosure into dual chambers
- when tested as a dual chamber ports were used to join the chambers, making this a dual chamber reflex design (dcr), these internal ports were likely also massively undersized in most configurations tested
Several configurations were tested, varying mostly the internal ports diameters and cross sectional areas, and also adding up to 3 eight inch diameter output ports.
SPL was measured INSIDE the enclosure and despite being driven by 400 hp engine they were only able to achieve results approaching 160 db at 2.5 hz. This is really pretty embarrassing and I'm not sure why they even published this wholly mediocre result.
Clearly they have no idea how to properly size ducts for minimal airflow (or maybe they fully intended to have airflow problems). This result can be had using regular drivers and amps for a fraction of the cost of just the racecar engine alone.
Anyway, the point here is that they did NOT use multiple different sized ports at any one time unless I completely miss that part (and if I did please point me to the right place). It's just a dcr enclosure with the expected narrow bandwidth and mediocre spl.
My theoretical calculations did not include friction or temperature ocillations, nor did I run the more at it's max power band. Hence the inefficiency. Name another device that can make 115db groundplane at 1 meter on 6hp..... there are none.
The issue is that current noise suppression systems are failing, and the manufactures are no longer in business. Also, the current designs used introduce significant amounts of higher order distortion which makes the working environment too loud. A better design would have increased reliability and less distortion.
@ Just a guy
Here is the patent link Patent US6665413 - Infrasonic Helmholtz resonator - Google Patents
Try reading this. It has 11 different designs.
Yes, it has very narrow bandwidth.
Yes, it is inefficient.
No, I am not going to copy it directly.
Try using your imagination to understand how a variable Helmholtz resonator could be used to accomplish these goals.
If there is any constructive advice, shoot away. I am currently working on the final iteration of the design doc. My work in linearX indicates that this will work, but as far as I can tell, the wider my frequency bandwidth, the lower the efficiency of the system.
I'll post what I have here after my testing next month in the AFRL anachoic chambers.
Pics will be provided... try not to have a hissy fit. I am not scamming anyone, nor am I asking for money.
JG
Same principle, but for noise cancelling headphones the noise only needs to be cancelled at the ear, a relatively simple operation for steady state noise in an enclosed chamber (the ear muff) too small to have LF "room modes".
A noise cancelling device "perfectly out of phase" with noise sources propagating from multiple sources (industrial noise does not come from a single point source) could only be effective in one location, at other locations the noise will be non-correlated and could add an additional 3 dB noise pollution to the source, or even be in phase a wavelength behind and add 6 dB.
Your assessment of the "military design":
"It's just a dcr enclosure with the expected narrow bandwidth and mediocre spl. " seems spot on.
They probably had fun doing it though, messing with superchargers and big block engines has some appeal ;^).
Agreed.
FYI noise suppressors produce anti-notes, and 15-20 are typically spaced out around the generator. The way the original designer explained it to me was a bubble around the noise.
I don't understand it, and no, I am not qualified, but my job was just to get the transducer done. Now we are trying to get the port set taken care of, and I am just curious.
I'm curious too.
It would be interesting to see the noise spectra of the generator, and the noise spectra with the 15-20 "anti-note" noise suppressors included.
Assuming the generator is in an enclosed building, I'd think passive bass traps and proper elastomeric mounts would be a better approach to VLF attenuation than an active system for employee comfort.
Ok, I read it. It is nothing at all like the previous device you linked to and unless I'm missing something really major this is just a regular ported box with a 1/4 wave stub trap (the one closed port). I certainly hope this patent was not granted, there's nothing new here and it doesn't operate anything like what is described.
I can sim just about anything as long as the airflow device can be described with t/s parameters. So I simulated this patent device exactly as described. (With the excpetion that the ports are on the back of the box instead of the front, a distinction that might not even be considered with this software and which would make the schematic too confusing if sim'ed with front ports).
I used dual EV 180 drivers as specified. Here's the combined t/s for a pair so you can verify I didn't make any mistakes.
An externally hosted image should be here but it was not working when we last tested it.
For the enclosure I used a 27 cubic foot box with one open ended port (36 inches long) and one closed ended port (9.5 inches long), both with 6 inch diameter as specified.
An externally hosted image should be here but it was not working when we last tested it.
Here's the frequency response at 100 watts. (And this is being very kind, it's almost at double xmax at 100 watts at 6 hz).
An externally hosted image should be here but it was not working when we last tested it.
The response extends pretty high so you might think this is a wide band device, and it is, but not because of resonances. The port is only contributing in a meaningful way at the bottom of the passband, above that it's the driver producing most of the output.
Now here's the funny part (or not so funny if you understand what's going on). Here's the response if the short closed ended stub is COMPLETELY REMOVED.
An externally hosted image should be here but it was not working when we last tested it.
And there's no real change in any of the other graphs (like displacement) either. I kind of knew what was going to happen before I sim'ed it, I just wanted to see how bad it was.
They claim in room spl of 130 db in the 6 - 20 hz bandwidth of interest. This is either a blatant lie or the room gain is so monumentally incredible that it doesn't really matter what type of device is used.
Anyway, this stupid patent is just a regular ported box with a completely useless closed ended trap attached to make it seem interesting. It's hard to tell if they really don't understand what they are saying or are just making stuff up to get a patent.
Last edited:
At this point I've spent a couple of hours deciphering your links so clearly I am willing to help, at least to find a solution to an intellectual issue (if not as a design in a nuclear power plant) so don't accuse me of hissy fits again.
The problem is that you links were written by people that seem to have no idea what they are doing and their ideas are not new or special and they won't do what you want.
There are already devices that will use multiple resonances to create a 3 octave passband of acoustical gain.
What you are asking for has not been invented but you claim if you can get 5 - 20 hz bandwidth it will be easy to get 5 - 100 hz and do this at 125 db in the size of a refrigerator.
The problem you are going to bump up against is twofold. First, if you try to smush up more resonances than the existing 3 octave devices you will have phasing problems. Second, at the airflow rates you are talking about you are going to need massive vents to control velocity. Large diameter vents tuned to low frequencies are going to be large. In the device above, velocity through the open port is about 18 m/s, and that's only a couple of 6.4 mm xmax drivers pushing air.
If you have a PHD professor at your disposal, please have him describe his idea more robustly than ports that act as thermistors. If you can describe the enclosure and describe the airflow device with t/s parameters I can simulate it.
If you can't step up your game and at least present ideas that make sense I can't help any further.
The problem is that you links were written by people that seem to have no idea what they are doing and their ideas are not new or special and they won't do what you want.
There are already devices that will use multiple resonances to create a 3 octave passband of acoustical gain.
What you are asking for has not been invented but you claim if you can get 5 - 20 hz bandwidth it will be easy to get 5 - 100 hz and do this at 125 db in the size of a refrigerator.
The problem you are going to bump up against is twofold. First, if you try to smush up more resonances than the existing 3 octave devices you will have phasing problems. Second, at the airflow rates you are talking about you are going to need massive vents to control velocity. Large diameter vents tuned to low frequencies are going to be large. In the device above, velocity through the open port is about 18 m/s, and that's only a couple of 6.4 mm xmax drivers pushing air.
If you have a PHD professor at your disposal, please have him describe his idea more robustly than ports that act as thermistors. If you can describe the enclosure and describe the airflow device with t/s parameters I can simulate it.
If you can't step up your game and at least present ideas that make sense I can't help any further.
An externally hosted image should be here but it was not working when we last tested it.
Just for fun, that's the response of the same thing shown above but with the closed ended port opened up, which is more in line with what you said you wanted to try.
As you can see the tuning shifts up, it doesn't give you multiple tunings. This seems very much in agreement with the John K post I linked to earlier.
If this were as simple as adding a bunch of different ports to a box to achieve spectacular wide bandwidth response it would have been done already.
If you want to try multiple chambers either in series or parallel with either series or parallel ports that might be a step in the right direction, but as I mentioned you likely won't get anything better than a front loaded horn and you certainly won't be able to fit it in a refrigerator size. And it won't be easy to tune it right.
As I said, if you want more than 3 octaves it's best to use different subs to cover the different frequency ranges.
I'm reminded of this thread a few years ago. The "7-way" idea with one octave per driver is, as several say in the thread, mostly unworkable and bizarre, but there might be some ideas here that help the OP:
http://www.diyaudio.com/forums/multi-way/169296-thinking-building-7-way-loudspeaker.html
(and double-checking, no, "just a guy" didn't start that thread)
Just a guy, thank you so much for your input.
The "hissy fit" comment was me overreacting to people like sreten (who I know is WAY smarter than me, but also very negative) and that guy who basically called me a liar... full of hot air or whatever (top of the second page I think).
I don't have time to respond to everything in this thread tonight, but I'll do my best to update it as time goes on and I get more info.
If it's not already clear, I was contracted to build the transducer, and I am trying to figure out what the end user is going to do to make it usable over a large frequency bandwidth... I have rights to use the transducer, and if I can put it in an enclosure for my own use that gets better response than possible with a traditional setup.
As I understand it, the goal is to set up the various ports in such a way that higher frequencies will only propagate through shorter and narrower pipes. This requires that the pipe with the lowest impedance for a given mass flow rate be tailored such that the mass flow rate is the max peak-to-peak velocity *crossectional area for that given frequency.
The "hissy fit" comment was me overreacting to people like sreten (who I know is WAY smarter than me, but also very negative) and that guy who basically called me a liar... full of hot air or whatever (top of the second page I think).
I don't have time to respond to everything in this thread tonight, but I'll do my best to update it as time goes on and I get more info.
If it's not already clear, I was contracted to build the transducer, and I am trying to figure out what the end user is going to do to make it usable over a large frequency bandwidth... I have rights to use the transducer, and if I can put it in an enclosure for my own use that gets better response than possible with a traditional setup.
As I understand it, the goal is to set up the various ports in such a way that higher frequencies will only propagate through shorter and narrower pipes. This requires that the pipe with the lowest impedance for a given mass flow rate be tailored such that the mass flow rate is the max peak-to-peak velocity *crossectional area for that given frequency.
It sure looks like it was.
From what I've seen of many patents, that's not unusual.
I have a rant or few about patents (though perhaps I shouldn't complain, there are two with my name on them), but it's definitely for another subforum, and it may even be considered a totally-off-topic topic for the whole forum.
I'm not sure that this actually makes sense but I definitely think I understand what you are getting at.
Here's the problem. If there's only one chamber and you have multiple open ended ports the tuning frequency will be an average, not a bunch of different tunings. Another problem is that the short and narrow ports are going to have increased velocity compared to the long and large ports. (At least that's what this software says and at this point I have no reason to believe it isn't accurate for what's being sim'ed here, although a sanity check with Akabak wouldn't hurt), and if the velocity gets too high they will block up completely due to resistance and then it's like they aren't really there at all.
Let's try this with pictures. Here's the enclosure from before but with only one port 36 inches long, 6 inch diameter. This shows velocity at 100 watts.
An externally hosted image should be here but it was not working when we last tested it.
Now I'm going to add another open ended port 9.5 inches long and 6 inch diameter.
An externally hosted image should be here but it was not working when we last tested it.
What we can see clearly is that the two ports now have different velocity flows, I'm pretty sure the red line is the narrower port velocity which makes sense - shorter port less resistance. This is going to be a problem if you have a really narrow port. It might not be immediately evident but the fact that both ports peak velocity is at the same frequency points to the fact that they are not providing two different tuning frequencies, there is only one tuning frequency. There are blips higher up in frequency which indicate that both these pipes have their own pipe resonances which are not the same, but there is only one box tuning frequency.
Frequency response is as shown in the previous post with dual open ended ports.
If you have something specific you would like me to sim I can do it, but this one chamber with multiple different ports is not going to fly. It's not going to do what you want.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.