AJinFLA said:
I don't use eminence alpha tens, and i high pass 18db per octave at 70 Hz and don't 'boost' the bass, the xmax and the sd seem reasonable to me (they don't use big rubber surrounds) with the warrior- probably .25" like you state not the .5" max you left out although I have never witnessed the coil bottoming out - they barely move at deafening levels when high passed.
At 70 hz without the high pass filter they are 3-4 db UP from 200 hz or around 94 db at a meter each with 2.83 volts into 16 0hm (actually much higher we are at resonance!) . The Peerless at the same frequency and drive voltage is calculated at 83 db or 11 db less sensitive! So you need around 4 of the Peerless in parallel to match the output of one Warrior on a board or 20 to match the five. Now try and drive 20 Peerless in parallel!
QTS does matter
They're cheap - if your curious about the importers claims buy one and measure it. They don't work without high pass (like I have repeatd over and over again) and they don't require eq like I have stated (over and over)
panomaniac said:
http://www.diyaudio.com/forums/showthread.php?postid=1378591#post1378591
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I only use one horn sub now with the JBL serious 18" 2242, the old system with three horns and top of the line EV, CV or Jbl 2241 wasn't as good as (solid and tight) as a single 2242 horn in the center.
Besides the old 3 horn system wasn't doing much for my home's foundation. Have a friend that's into pipe organs, he would come over and we watch the bottom end crack the mortor between concrete blocks
Here is a back channel 'ambiance' sub used to enlarge the room with delay. It used four JBL 2235's 15's in push pull.
Besides the old 3 horn system wasn't doing much for my home's foundation. Have a friend that's into pipe organs, he would come over and we watch the bottom end crack the mortor between concrete blocks
Here is a back channel 'ambiance' sub used to enlarge the room with delay. It used four JBL 2235's 15's in push pull.
An externally hosted image should be here but it was not working when we last tested it.
Magnetar said:Have a friend that's into pipe organs, he would come over and we watch the bottom end crack the mortor between concrete blocks
You guys are having way too much fun!
Funny, I was just discussing this sort of thing with a "bass fiend/ pipe organ fan" buddy the other day. I had to admit that I've never been in a church or concert hall where the organ pedal notes actually shook the building. Maybe close in a concert hall. The pedal tones sound huge, deep, powerful - but in no way like the Hip-Hop or car sound shaking bass. Hmmmmmm.....
Magnetar said:
Couldn't you just get a speed boat or a fast car or something?
"I only use one horn sub now with the JBL serious 18" 2242, the old system with three horns and top of the line EV, CV or Jbl 2241 wasn't as good as (solid and tight) as a single 2242 horn in the center.
Besides the old 3 horn system wasn't doing much for my home's foundation. Have a friend that's into pipe organs, he would come over and we watch the bottom end crack the mortor between concrete blocks."
I came over to Mike's house for one of his Midwest famous "Beers and Ears" listening sessions. At that time Mike had his horn system set up. I recall that after more than a few beers and a lot of music that the RMS sound level was cranked up to somewhere in the 100 to 105db level. After a short time of that, I resorted to hanging out on the patio outside of the listening room. I recall that from the outside patio, it sounded like a live band playing in his basement.
The micro and macro dynamics were outstanding. The one thing that my system (Lowther mounted in a Hedlund Horn) could somewhat exceed (to my ears) of his system was the wholosity of sound, compared to his four way horn speakers. I recall reading in one of his posts later on that he took steps to phase-align his speakers, so he probably closed the gap in that arena as well.
There are certain people who, when they post, I read whatever they say. Of course Lynn is one of those people. For magnetics, I would put Dave Slagle in that category. I would put Magnetar in that category when he speaks about speaker design.
Retsel
Besides the old 3 horn system wasn't doing much for my home's foundation. Have a friend that's into pipe organs, he would come over and we watch the bottom end crack the mortor between concrete blocks."
I came over to Mike's house for one of his Midwest famous "Beers and Ears" listening sessions. At that time Mike had his horn system set up. I recall that after more than a few beers and a lot of music that the RMS sound level was cranked up to somewhere in the 100 to 105db level. After a short time of that, I resorted to hanging out on the patio outside of the listening room. I recall that from the outside patio, it sounded like a live band playing in his basement.
The micro and macro dynamics were outstanding. The one thing that my system (Lowther mounted in a Hedlund Horn) could somewhat exceed (to my ears) of his system was the wholosity of sound, compared to his four way horn speakers. I recall reading in one of his posts later on that he took steps to phase-align his speakers, so he probably closed the gap in that arena as well.
There are certain people who, when they post, I read whatever they say. Of course Lynn is one of those people. For magnetics, I would put Dave Slagle in that category. I would put Magnetar in that category when he speaks about speaker design.
Retsel
Magnetar said:
Well - maybe not all my words - the measurements were taken 2" from the dustcap! In other words Linkwitz's 90 db is actually around 65...... So he needs 16 woofers per channel to be low distortionI think 24 would be better
Well, this measurement was done at 91dB @ 1m
http://www.linkwitzlab.com/TAC-review.htm
(scroll to fig 3)
40Hz, 2nd harmonic 43dB down, 3rd -60dB. I still think XLS's are very viable OB option. Yes, lots of them would work better, but even a sigle pair is not that bad.
Magnetar said:
This is one of my all-time favorite posts!!! Thank you Magnetar!
In the same mood, here's a little animation that'll bring back memories for the folks that grew up with the big plastic Zenith AM/FM radio sitting in the kitchen ...
Y'know, I've been thinking about this whole baffle-step compensation thing. Yes, in principle, it's the right thing to do. But the usual simulation shows a single transition from half-space radiation (what the tweeter sees) to full-space (what the woofer sees for wavelengths significantly larger than the cabinet).
But this simple transition from half to full-space radiation is really only true if the loudspeaker and the listener are suspended in space, at least 20 feet off the ground. If you gently lower the loudspeaker and listener back to Earth, the reflection off the ground becomes significant, and the radiation transitions back to half-space again at the lowest frequencies.
So in an outdoor environment, at least 50 feet away from any other structures, the radiation from the speaker is half-space at the highest frequencies (in the working range of the tweeter), transitioning to full-space as the wavelengths exceed the size of the cabinet, and at the lowest frequencies, when the ground reflection and direct-arrival waves approach in-phase, the speaker transitions back to half-space radiation again.
The two transition frequencies are furthest apart if the speaker is physically very small and sitting on an acoustically transparent stand; even a simple measure like extending the front panel down to ground level will move the transition frequencies closer together and decrease the depth of dip region. (I've found that a baffle extension is good for 1 to 2 dB decrease of the "dip" region, depending on the size of the speaker.)
But listening to a speaker outdoors 50 feet away from any other structures is still pretty unrealistic. Add a rear wall, and the radiation transitions to a quarter-space at the lowest frequencies, and add a side wall, and the transition is to an eighth-space - effectively, a simple conical corner-horn horn loading.
We're still listening outdoors to one speaker - maybe at a corner where two exterior walls join. Let's move to real room indoors. Now we have a full set of reflections from the the floor (and carpeting only absorbs high frequencies), side walls, and rear wall. We also have room modes, which is a separate phenomena than simple changes in driver loading. Room modes store energy, and don't come up to full strength in the first milliseconds. The length of the decay time strongly depends on room furnishings and losses from the room into other rooms and the outdoors.
Which then gets us into the whole area of LF losses through diaphragmatic absorption of the walls and windows. As the frequency is decreased, these structures reflect, then resonate at multiple frequencies, and at the lowest frequencies, become acoustically transparent. This is why bass from boom cars seems to go through everything - it takes very heavy and rigid structures to resist the lowest frequencies. Almost everything is transparent to a 55-foot wavelength (20 Hz).
See how far we've gone from the baffle-step model with its 6 dB transition? This model is only dB-for-dB accurate for a speaker (and listener) suspended in mid-air. Even the presence of the floor reflection adds another transition frequency only a couple of octaves lower than the baffle-step transition. Add in rear and side walls, stored energy from room modes, diaphragmatic wall and window losses, and I seriously wonder how much validity the model has left.
Maybe the prevalence of the baffle-step model is why speakers have gotten more boom-and-tweet over the last 15 years. The convenience of computer modelling has to some extent replaced careful real-world measurements. If the models were accurate to a fraction of a dB, that would be one thing, but LF measurements are notoriously room, and even house, dependent - and the power emitted in the near-field is not going to be the same as the SPL at the listening position.
But this simple transition from half to full-space radiation is really only true if the loudspeaker and the listener are suspended in space, at least 20 feet off the ground. If you gently lower the loudspeaker and listener back to Earth, the reflection off the ground becomes significant, and the radiation transitions back to half-space again at the lowest frequencies.
So in an outdoor environment, at least 50 feet away from any other structures, the radiation from the speaker is half-space at the highest frequencies (in the working range of the tweeter), transitioning to full-space as the wavelengths exceed the size of the cabinet, and at the lowest frequencies, when the ground reflection and direct-arrival waves approach in-phase, the speaker transitions back to half-space radiation again.
The two transition frequencies are furthest apart if the speaker is physically very small and sitting on an acoustically transparent stand; even a simple measure like extending the front panel down to ground level will move the transition frequencies closer together and decrease the depth of dip region. (I've found that a baffle extension is good for 1 to 2 dB decrease of the "dip" region, depending on the size of the speaker.)
But listening to a speaker outdoors 50 feet away from any other structures is still pretty unrealistic. Add a rear wall, and the radiation transitions to a quarter-space at the lowest frequencies, and add a side wall, and the transition is to an eighth-space - effectively, a simple conical corner-horn horn loading.
We're still listening outdoors to one speaker - maybe at a corner where two exterior walls join. Let's move to real room indoors. Now we have a full set of reflections from the the floor (and carpeting only absorbs high frequencies), side walls, and rear wall. We also have room modes, which is a separate phenomena than simple changes in driver loading. Room modes store energy, and don't come up to full strength in the first milliseconds. The length of the decay time strongly depends on room furnishings and losses from the room into other rooms and the outdoors.
Which then gets us into the whole area of LF losses through diaphragmatic absorption of the walls and windows. As the frequency is decreased, these structures reflect, then resonate at multiple frequencies, and at the lowest frequencies, become acoustically transparent. This is why bass from boom cars seems to go through everything - it takes very heavy and rigid structures to resist the lowest frequencies. Almost everything is transparent to a 55-foot wavelength (20 Hz).
See how far we've gone from the baffle-step model with its 6 dB transition? This model is only dB-for-dB accurate for a speaker (and listener) suspended in mid-air. Even the presence of the floor reflection adds another transition frequency only a couple of octaves lower than the baffle-step transition. Add in rear and side walls, stored energy from room modes, diaphragmatic wall and window losses, and I seriously wonder how much validity the model has left.
Maybe the prevalence of the baffle-step model is why speakers have gotten more boom-and-tweet over the last 15 years. The convenience of computer modelling has to some extent replaced careful real-world measurements. If the models were accurate to a fraction of a dB, that would be one thing, but LF measurements are notoriously room, and even house, dependent - and the power emitted in the near-field is not going to be the same as the SPL at the listening position.
Yes, I know. I linked that very similar Eminence to show typical (but honest spec'd) SD of a 10" (guitar) driver. The xmax, etc. similarities were purely coincidental.
Sorry about any confusion, your original post stated 50hz. I actually didn't mention XO frequency, but both could be looked at in Linkwitz's Max Spl spreadsheet , given the possible displacement of the driver and baffle size. I think it was wise to now raise it to 70hz. If you note, driver sensitivity and Qts are irrelevant parameters.
Which? The quoted specs or the lower figure? You could simply measure the cone diameter. It's much more likely to be 8" than 9". Xmax would be a bit more tricky, but you could use a free program like ARTA to measure the distortion of the driver. I suspect a cheap guitar driver will have quite a bit of distortion, even before nearing Xmax. With a simple motor (no shorting ring) my guess would be that distortion skyrockets anywhere near Xmax. Keep in mind that higher distortion will sound subjectively louder than a lower distortion system. The Xmech of the driver is probably much higher than the linear operation travel, so you would have a hard time bottoming them, especially with no eq.
Not quite sure what you mean here. You're saying the 0.75 qts drivers show a 3-4db measured rise above the 200hz level in an open baffle? Interesting. Baffle sims like xbaffle predict a loss even in room with floor/wall influence. Can you post these measurements? The voltage sensitivity of the Peerless XLS12 is around 90db/2.83V. A pair of XLS would require over 500 watts to get above 117db at 200hz. Not unreasonable.
How did you arrive at your numbers?
What? Please explain.
The shape of the acoustic roll off of your system can be determined many ways. They were a lot cheaper when they were being sold on ebay a while back. IIRC, you could pick up a pair of the executioner 15's for around $100! I think the 18's around $200 and even that whopping 21 incher for around $250/pr if you were patient. I was tempted.
However I prefer to shape the acoustic roll off (and associated group delay) by other methods, as I find the high distortion motors and noisy suspension of the cheap drivers a bit too much of a fidelity compromise, although they could make a fun, cheap "loud" system.BTW, those bass horn subwoofers in the pic above are huge! Look about 5-6' tall. What frequency is the horn loading effective down to for these subwoofers?
cheers,
AJ
There is the baffle step 2pi-4pi effect. There is also the floor bounce 4pi-2pi effect one or two octaves below. There are other boundary (walls, ceiling, etc) effects. With OB bass, low frequency measurement is even more troublesome. I have been wondering why we are bothered with near field measurement at all. Could we just use gated far field measurement throughout the entire audio bandwidth and have the window set to about 5ms? Within approximately 5ms sound is perceived as part of the original, direct sound and above that distinct sound. If we use 5ms gated measurement for XO EQ design, we may be able to completely ignore BSC, floor bounce, etc, and keep things simple and accurate. Normal practice is to make the speaker sound 100% flat in free space, then calculate and add in EQ for BSC, floor bounce, etc, then, when taking into a room, use room EQ to patch it up. This may be necessary for commercial products. But for DIYers who make speakers for their own use in their own rooms, why be bothered?
This is what I have been thinking lately. Please correct me if I am wrong.
Regards,
Bill
This is what I have been thinking lately. Please correct me if I am wrong.
Regards,
Bill
Here is a Warrior 10 XBaffle sim on a 4x4 baffle based on the Shred site parameters. Please correct me if I have entered incorrect data.
An externally hosted image should be here but it was not working when we last tested it.
Most of the DIY speaker designers I know are well aware of the effects of proximity effects of the floor and walls on baffle step. But let’s not confuse the 2Pi to 4Pi transition with other factors which are dependent on both the speaker position AND the listener position. The latter would include such things as floor bounce and cancellations/augmentation at specific frequencies which arise due to specific path length differences to the listening position.
The 2Pi to 4Pi transition is best examined by looking at power response. If we consider an omnidirectional point source radiating uniform intensity in all directions we can look at the power response as it is placed close to a wall for floor. In free space (4Pi) the power response would be constant vs. frequency. Let that be the 0 dB level. If placed on (or very close to) an infinite baffle the source will radiate into 2Pi and we know that the intensity will be 6dB higher, but the radiated power will be only 3dB greater. This is because we have 4 time the intensity radiated into ½ the space and 4 x ½ = 2= 3dB. So we can see the 4Pi to 2Pi transition by looking at the behavior of the radiated power as the source gets close to a wall or floor. In acoustics “close” typically means that the wave length is greater than between 3 to10 times the distance to a secondary surface.
The figure below shows two calculations of power response. To the left is a case where the source is located 7” away form a ground plane as might be the case for a speaker with a 10” woofer centered close to the floor. Thus we would expect the source to undergo a transition form 4Pi to 2Pi radiation for frequencies with wave lengths greater than 21” to 70”, or for frequency lower than some where between 644 and 193 Hz. The plot to the left clearly shown that the 2Pi to 4Pi transition occurs within this range. The source can be considered to be operating in 2Pi space up to 150 to 200 Hz. At the right the same result is presented for the case where the source is 3 ft from a ground plane. We would expect the 2Pi to 4Pi transition to occur between 37 Hz and 125 Hz which is indeed the case.
So what does this mean with regards to dipole woofers and equalization? That depends on your objectives and point of view, but IMO it suggests that the woofer system should be designed to have the driver(s) close to the floor and the woofer to midrange crossover point (in a 3-way) chosen such that the woofer system operates in the 2Pi region over it’s entire frequency range. While this will provide 6dB greater voltage sensitivity (and 3dB greater efficiency) it will not alter the need to apply equalization for the dipole roll off. This is the approach used in speakers such as the NaO and Orion. Positioning the speaker more than 3 feet from a side or back wall will also minimize the further transition from 2Pi to Pi space. However, as observed in the figure to the right, there may be some deep bass boost, below 50 Hz or so, curtailing the need for full dipole equalization at the lowest frequencies. In the NaO design this is addressed by a bass “tone control” which allows the listener to adjust the slope of the dipole eq to match his room/taste. The Orion has a 50 Hz, 1st order HP filter which can be switched in or out, and for which, I presume, is for addressing the same effect (or not?).
The 2Pi to 4Pi transition is best examined by looking at power response. If we consider an omnidirectional point source radiating uniform intensity in all directions we can look at the power response as it is placed close to a wall for floor. In free space (4Pi) the power response would be constant vs. frequency. Let that be the 0 dB level. If placed on (or very close to) an infinite baffle the source will radiate into 2Pi and we know that the intensity will be 6dB higher, but the radiated power will be only 3dB greater. This is because we have 4 time the intensity radiated into ½ the space and 4 x ½ = 2= 3dB. So we can see the 4Pi to 2Pi transition by looking at the behavior of the radiated power as the source gets close to a wall or floor. In acoustics “close” typically means that the wave length is greater than between 3 to10 times the distance to a secondary surface.
The figure below shows two calculations of power response. To the left is a case where the source is located 7” away form a ground plane as might be the case for a speaker with a 10” woofer centered close to the floor. Thus we would expect the source to undergo a transition form 4Pi to 2Pi radiation for frequencies with wave lengths greater than 21” to 70”, or for frequency lower than some where between 644 and 193 Hz. The plot to the left clearly shown that the 2Pi to 4Pi transition occurs within this range. The source can be considered to be operating in 2Pi space up to 150 to 200 Hz. At the right the same result is presented for the case where the source is 3 ft from a ground plane. We would expect the 2Pi to 4Pi transition to occur between 37 Hz and 125 Hz which is indeed the case.
So what does this mean with regards to dipole woofers and equalization? That depends on your objectives and point of view, but IMO it suggests that the woofer system should be designed to have the driver(s) close to the floor and the woofer to midrange crossover point (in a 3-way) chosen such that the woofer system operates in the 2Pi region over it’s entire frequency range. While this will provide 6dB greater voltage sensitivity (and 3dB greater efficiency) it will not alter the need to apply equalization for the dipole roll off. This is the approach used in speakers such as the NaO and Orion. Positioning the speaker more than 3 feet from a side or back wall will also minimize the further transition from 2Pi to Pi space. However, as observed in the figure to the right, there may be some deep bass boost, below 50 Hz or so, curtailing the need for full dipole equalization at the lowest frequencies. In the NaO design this is addressed by a bass “tone control” which allows the listener to adjust the slope of the dipole eq to match his room/taste. The Orion has a 50 Hz, 1st order HP filter which can be switched in or out, and for which, I presume, is for addressing the same effect (or not?).
An externally hosted image should be here but it was not working when we last tested it.
total power doubled going from 4Pi to 2Pi?
John,
Maybe I am missing some thing here, but I don't quite follow your statement:
"Let that be the 0 dB level. If placed on (or very close to) an infinite baffle the source will radiate into 2Pi and we know that the intensity will be 6dB higher, but the radiated power will be only 3dB greater. This is because we have 4 time the intensity radiated into ½ the space and 4 x ½ = 2= 3dB."
Could you please explain to me why the intensity of the radiation from the omnidirectional source will be four times higher, and the total power will double, when the radiation changes from 4Pi to 2Pi (assuming that the source is close to the infinite baffle, not right "on" it -- there is a difference)? In particular, the doubling of the radiation power seems to violate the law of conservation of energy, unless there is something elso going on. If the coversion rate of electrical power to acoustic power by the transducer remains the same, where does the extra 100% power come from? Is there some kind of assumption that the conversion rate is doubled due to the change from 4Pi to 2Pi? If so, what is the mechanism? Better coupling with air?
Cheers,
Kurt
John,
Maybe I am missing some thing here, but I don't quite follow your statement:
"Let that be the 0 dB level. If placed on (or very close to) an infinite baffle the source will radiate into 2Pi and we know that the intensity will be 6dB higher, but the radiated power will be only 3dB greater. This is because we have 4 time the intensity radiated into ½ the space and 4 x ½ = 2= 3dB."
Could you please explain to me why the intensity of the radiation from the omnidirectional source will be four times higher, and the total power will double, when the radiation changes from 4Pi to 2Pi (assuming that the source is close to the infinite baffle, not right "on" it -- there is a difference)? In particular, the doubling of the radiation power seems to violate the law of conservation of energy, unless there is something elso going on. If the coversion rate of electrical power to acoustic power by the transducer remains the same, where does the extra 100% power come from? Is there some kind of assumption that the conversion rate is doubled due to the change from 4Pi to 2Pi? If so, what is the mechanism? Better coupling with air?
Cheers,
Kurt