Yes, if you have a ported chamber on both sides of the driver you have a 6th order bandpass. And you can have multiple ported chambers on each side of the driver if you like.
But like I said a few times, even a single chamber on only one side of the driver is going to require massive ports to keep velocity in check, and your chamber/port is going to be more than the size of a refrigerator to tune low enough and keep velocity down. Notice all the previous simulated examples the box was 27 cubic feet with a 6 inch diameter port 36 inches in length. That's close to refrigerator size already and your port is going to have to be orders of magnitude larger (wider diameter and longer) to control wind noise. By the time you get the port long and large enough to control velocity the design will almost certainly have naturally morphed to a transmission line instead of a ported box. Considering the massive amount of air you want to push and the low tuning you want to achieve, any feasible design is going to be very large. If you could give some details about the jet velocity control device you say you are using that might help a bit but it's still going to be big.
And we haven't even investigated the possibility that a chambered design might not work at all depending on the airflow device design. If it's a fan type design which has open area not covered by blades there's a path of least resistance back out through the fan itself so a box might not pressurize properly. It's one thing to push air into a low resistance, it's another completely to be able to hold back and build up a positive pressure. There's a lot to consider here but without knowing any details at all of the device properties there's no way to know how it might perform in any given configuration.
Last edited:
Actually they don't mention that the SPL was in the bandwidth of interest, "dB was in the 130 range" sounds like they only measured distortion level with a dB meter with a calibrated microphone. The EVX-180B can handle "near full power" without tearing up, but as you noted at 100 watts they would exceed Xmax by 6 dB. In reality, the suspension is tight enough that they won't exceed Xmax by much no matter how much power they are given, a change from the original foam surround EVX-180, which was nixed rather quickly after mechanical failures. At 800 watts or so, distortion would so far exceed the fundamental level as to render it just part of the noise.
"In order to evaluate the aforementioned system, it was placed in a room having an opening of approximately 3 square feet. With the speaker operating at a frequency of 8 Hz or 14 Hz, and with the Helmholtz resonator tuned to those frequencies at a few watts of power, a 118 dB level was measured in the room with a calibrated microphone. With the room closed and the speaker operating near but not at maximum power, the dB was in the 130 range."
I wonder how they got a 27 cubic foot cube shaped cabinet (nine square feet per side) through "an opening of approximately 3 square feet"...
Last edited:
It does specifically say "with the speaker operating at 8 or 14 hz" which does make it sound like they measured at 8 or 14 hz, but as you point out this is not necessarily so. The whole document seems to be an exercise in deception.
It should not have been tested in a room at all, obviously, but this is what I think happened.
I think they tested it in a variety of different rooms with a variety of different sub locations and mic placements. I think they published the highest spl result they were able to obtain, which happened to be in a specific room (probably a concrete room) with a window open a bit (equal to 3 sq ft "opening") and put the mic right in the middle of the biggest room mode they could find. This explanation seems to be on par with the deceptive nature of the rest of the writeup and the design itself.
Hi,
No, it doesn't work well at all. Parallel tuning never does.
(Higher frequencies short the lower frequencies.)
rgds, sreten.
Last edited:
First, your speaker is not the only speaker named "Marimba" and apparently google does not think your speaker is very important so I had a very hard time finding it. I did not realize it was a diy build on this forum. A link would have been greatly appreciated and since I took the time to find it, here it is - http://www.diyaudio.com/forums/multi-way/270544-line-array-varying-length-tl-pipes.html
Second, if you had bothered to read this thread it would be overwhelmingly apparent that your speaker is not even remotely similar to the project being discussed in this thread. A brief description of your speaker is as follows-
- 9 identical drivers, each end loaded into it's own constant csa tl (cardboard tube), each tube having a different length
- the lengths of the tubes are 114, 104.5, 95, 85.5, 76, 71.25, 66.5, 61.75 and 57 cm
- these speakers are arranged in line array fashion and presumably run off a single channel with no crossover to separate passbands - they all get the same signal
The idea behind this speaker was "ST decided to try and "fill the hole", which is the Achillie's Heel of QWT designs: the nulls at higher mutiples of the resonant frequency of each tube"
Third, since I had to go to the trouble of finding and reading about your speaker that has nothing at all in common with this discussion I'll tell you a few good reasons why it's a bad idea and you should have listened to the people that were telling you that.
1. The initial premise is flawed. Properly designed quarter wave devices do not have to have nulls in the response. Making a tapered line (not a tube), offsetting the driver carefully somewhere along the length of the line (not end loaded), and adding a small to moderate amount of stuffing gives you the opportunity to get a very flat response and if you are good with a simulator you can get almost dead flat response with no nulls at all.
2. The lowest tuned driver in the system needs to do a lot of work. It should be larger and more powerful than the rest (or there should be multiple drivers handling the lowest notes). In your speaker the lowest tuned driver is one of nine identical drivers so the lowest notes are only supported by 1/9 of the bass drivers. Furthermore, each of the other 8 drivers are tuned sequentially higher and higher in frequency and are not protected in any way below their tuning frequencies. That means that the lower notes that the single lowest tuned driver can play will completely unload all the other drivers since they have no protection below their tl tuning. This limits the power the system can handle because the unloaded drivers are going to flop around like fish out of water. AND the ONLY frequencies that are properly loaded in this system are those above the tuning of the shortest tube which is only 57 cm long. I guess the good news is that notes higher than the tuning of the shortest pipe can be played just fine by all 9 drivers - the problem though is that it's completely backwards - the low bass is the area that needs to be the strongest, not the highest notes.
3. Having 9 differently tuned transmission lines all tuned within about an octave of each other playing the same music and overlapping each other is a frequency response and phase disaster.
So in conclusion, you might have successfully achieved your goal of filling in the nulls created by a bad tl design (which could have been avoided by simply designing a good tl), but you did it by choosing bad compromises that horribly cripple the max spl potential of the system. Only 1 driver is able to play the lowest notes while all the rest are completely unloaded and since all 9 drivers have completely different frequency and phase response they are all fighting each other, adding at some frequencies and cancelling at others.
It's not stated how high in frequency you are playing these but you probably have some pretty severe lobing problems as well.
Last edited:
I presume you (and maybe the OP, but it's hard to tell) are talking about using several different lengths and/or diameter ports in a bass reflex system. The ports are indeed low-pass, and the driver would be unloaded below the highest tuned port.
A quick thought comes to mind of replacing the higher tuned port(s) with passive radiators. As they effectively make a band-pass with the air in the cabinet, this would fix the unloading problem, but whether this would be an "appropriate" fix for the rest of the OP's needs, I'll leave that for the OP to decide.
I probably shouldn't send this post ...
Ugh. Could someone actually read the first post before commenting?
I was very clear. No cabinet. No "port"
As such the only time we are dealing with compliance of acoustic suspension is below the tuning of the system. At tuning, the mass in the system in increased by about 4%, and no air flows out the rear "port"
That makes this a transmission-line system, as a result.
JG
I was very clear. No cabinet. No "port"
As such the only time we are dealing with compliance of acoustic suspension is below the tuning of the system. At tuning, the mass in the system in increased by about 4%, and no air flows out the rear "port"
That makes this a transmission-line system, as a result.
JG
I read ALL the posts as well as the links to patents you provided and I'm still not really clear an exactly what you are proposing. And these two last posts seem to contradict each other - first you say no box and no port, then you say transmission line system, then you say you are using a 2.25' cubed box. All these things contradict each other. There's no pictures or even a decent description of what you want and no details whatsoever on the airflow device. No explanation of what you mean by "deviate from tuning frequency or try to add additional systems". To be honest I have no clue what you are doing. I'm not sure what you are doing with a 2.25' cube box either, you can't tune that to your desired frequency and adequately control airflow, it just is not remotely possible.
Anyway, if it's steady state noise you are killing there's no need to worry about group delay.
Last edited:
There isn't a passive radiator in the world that could handle the amount of airflow OP says he is capable of producing. And you don't have much space for multiple passive radiators or a horn flare since OP said he wanted to the system to be about the size of a small refrigerator.
Thank you! Now I have a 2nd excretory orifice 
First, your speaker is not the only speaker named "Marimba" and apparently google does not think your speaker is very important so I had a very hard time finding it. I did not realize it was a diy build on this forum. A link would have been greatly appreciated and since I took the time to find it, here it is - http://www.diyaudio.com/forums/multi-way/270544-line-array-varying-length-tl-pipes.html
Thank you for your dillegence i would have provided a link but can only do so much on my 3" Android phone.
Second, if you had bothered to read this thread it would be overwhelmingly apparent that your speaker is not even remotely similar to the project being discussed in this thread.
I did, in fact read the thread. I assumed the OP wanted enhanced output from a speaker design. I concede the point that the Marimba is not a similar design. However, It'd seem rather stubborn to restrict oneself to only one approach to solving a problem. If you're fighting a fire and you have a bucket of water, are you going to be pissed off at me if I show up with an "ABC" fire extinguisher?
Third, since I had to go to the trouble of finding and reading about your speaker that has nothing at all in common with this discussion I'll tell you a few good reasons why it's a bad idea and you should have listened to the people that were telling you that.
1. The initial premise is flawed. Properly designed quarter wave devices do not have to have nulls in the response.
How dare you imply my Marimba was properly designed
Making a tapered line (not a tube), offsetting the driver carefully somewhere along the length of the line (not end loaded),
...hard to do when I limited myself to 4" cardboard tubes
2. [deleted your quote for brevity] Your criticisms may be true. Alternatives that "might" remove these objections: What if every driver has its own amplifier? What if we band-limit (high pass) each speaker's signal so that it doesn't have to work below whatever its cut-off is? If one could custom design each separate enclosure, each driver's response could be further optimized. No requirement that the same type driver be used, either.
3. Agree the "Marimba" is a phase disaster and probably lobing problems (not sure about this one) but the active EQ seems to help a lot. May still be a whacked speaker
First, your speaker is not the only speaker named "Marimba" and apparently google does not think your speaker is very important so I had a very hard time finding it. I did not realize it was a diy build on this forum. A link would have been greatly appreciated and since I took the time to find it, here it is - http://www.diyaudio.com/forums/multi-way/270544-line-array-varying-length-tl-pipes.html
Thank you for your dillegence
i would have provided a link but can only do so much on my 3" Android phone.Second, if you had bothered to read this thread it would be overwhelmingly apparent that your speaker is not even remotely similar to the project being discussed in this thread.
I did, in fact read the thread. I assumed the OP wanted enhanced output from a speaker design. I concede the point that the Marimba is not a similar design. However, It'd seem rather stubborn to restrict oneself to only one approach to solving a problem. If you're fighting a fire and you have a bucket of water, are you going to be pissed off at me if I show up with an "ABC" fire extinguisher?
Third, since I had to go to the trouble of finding and reading about your speaker that has nothing at all in common with this discussion I'll tell you a few good reasons why it's a bad idea and you should have listened to the people that were telling you that.
1. The initial premise is flawed. Properly designed quarter wave devices do not have to have nulls in the response.
How dare you imply my Marimba was properly designed
Making a tapered line (not a tube), offsetting the driver carefully somewhere along the length of the line (not end loaded),
...hard to do when I limited myself to 4" cardboard tubes
2. [deleted your quote for brevity] Your criticisms may be true. Alternatives that "might" remove these objections: What if every driver has its own amplifier? What if we band-limit (high pass) each speaker's signal so that it doesn't have to work below whatever its cut-off is? If one could custom design each separate enclosure, each driver's response could be further optimized. No requirement that the same type driver be used, either.
3. Agree the "Marimba" is a phase disaster and probably lobing problems (not sure about this one) but the active EQ seems to help a lot. May still be a whacked speaker
The difference here is substantial and pretty crucial - you are using 9 drivers and the OP has only one air flow device - not sure what it is. So I'm not sure what you are suggesting - putting 9 transmission lines on the back of this one air flow device? I already showed in previous post sims that indicate this will not work. The tuning will be an average, not several different tunings.
...hard to do when I limited myself to 4" cardboard tubes
If you primary concern is getting rid of nulls caused by harmonics I would strongly suggest ditching the cardboard tubes and making a tapered offset tl with a bit of stuffing. What can I say? Your proposed solution is seriously flawed.
Yeah, this could work if you could work through the frequency response and phasing issues since there is still going to be overlap even with active crossovers separating each passband. But are you really suggesting a 9 way active system with 9 tunings spanning only about a single octave? Even if you did this the lowest tuned driver would still have to handle all the heavy lifting by itself.
If you are absolutely stuck on using cardboard tubes I would probably recommend using the same length tubes on all 9 drivers and have them all tuned the same. Use them as the woofer section of your speaker and cross over to the mid at a low frequency to try to avoid the null. That would solve all the problems - they all have the same frequency response and phase (if we ignore the boundary loading), cross over low to avoid the null and lobing problems, and all 9 speakers are tuned the same so they all share the heavy lifting playing the lowest notes. There's a reason why it's usually done this way.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.