So maybe its me with some antiquated formulas but I keep coming up with my F3 cutoff before my resonant frequency!
Whats up here?
Is my chosen driver unsuitable for vented enclosure? The EBP is spot on at 93.33.....
Here are the specs...could someone double check my math...perhaps my chosen formulas are flawed.
Fb 35.830... Hz
F3 41.188...Hz
_______________________________________________________Rick....
Fs 28 Hz
Vas 15.6 cu Ft.
Qts .29
_
SPL 99Db
Qms 8.28
Qes 0.3
41-2.4K Hz
Whats up here?
Is my chosen driver unsuitable for vented enclosure? The EBP is spot on at 93.33.....
Here are the specs...could someone double check my math...perhaps my chosen formulas are flawed.
Fb 35.830... Hz
F3 41.188...Hz
_______________________________________________________Rick....
Fs 28 Hz
Vas 15.6 cu Ft.
Qts .29
_
SPL 99Db
Qms 8.28
Qes 0.3
41-2.4K Hz
Attachments
Last edited:
Specifics
Thanks Loren..
The make is the Eminence Sigma Pro-18A2...an 18" driver from the "pro" application side of things.
The intent is to create a highly efficient system(100+ Db) to use a low powered single-ended tube amp.
I'm tending to shy away from the intensive woodworking side of horned enclosures & trying to stick with large driver, ported, Horn Mids & HF drivers.
My limited skills with wood and relative lack of appropriate tools make for straightforward port versions being the way to go.
This driver at 99Db seems to be the highest I have found with an appropriate low Fs for this project.
________________________________________________Rick.........
Thanks Loren..
The make is the Eminence Sigma Pro-18A2...an 18" driver from the "pro" application side of things.
The intent is to create a highly efficient system(100+ Db) to use a low powered single-ended tube amp.
I'm tending to shy away from the intensive woodworking side of horned enclosures & trying to stick with large driver, ported, Horn Mids & HF drivers.
My limited skills with wood and relative lack of appropriate tools make for straightforward port versions being the way to go.
This driver at 99Db seems to be the highest I have found with an appropriate low Fs for this project.
________________________________________________Rick.........
Hi Richard,
It is not at all unusual for the F3 to be higher then the FB in a bass reflex enclosure. I just threw the published params for my 10" woofers into akabak and brought up the alignment list (see attachment).
This is a vifa M26WR-09-08 which is recommended for reflex enclosures.
Note that all the alignments shown have F3 higher than FB.
Tony.
It is not at all unusual for the F3 to be higher then the FB in a bass reflex enclosure. I just threw the published params for my 10" woofers into akabak and brought up the alignment list (see attachment).
This is a vifa M26WR-09-08 which is recommended for reflex enclosures.
Note that all the alignments shown have F3 higher than FB.
Tony.
Attachments
Well, here is the comparison of your selected driver compared to a JBL 2245. I did this because when I saw the response curve for the Eminence I was very disheartened. I would not choose that driver and I will explain why shortly.
An externally hosted image should be here but it was not working when we last tested it.
The problem with the Eminence is that the cone breakup at 1kHz and beyond is very nasty. While the SPL is high, it is measured at 1 kHz, a frequency that you would never want to produce with this driver anyway. As the frequency goes down it drops at least 3 dB, so the actual region where the driver makes sound has an SPL closer to 95 - 96 dB.
I ran the cabinet volume way up to 18 cubic feet to get as much bottom end as possible.
Secondly, all that breakup will require a lot of crossover work to suppress. This will further rob valuable amp power with inductors.
I compared the Eminence with the pro JBL 2245H. You can view the raw driver data HERE.
I modled the JBL in a 16 cubic foot cabinet. The F3 is about the same as the Fb. If you drop the volume down to 10 cubic feet F3 goes up to 34 Hz.
Obviously, there is more to this story than simply the woofer, so a lot depends on what the rest of the drivers are that you intend to use and the system architecture.
If it were me I would not bother trying to squeeze a super high SPL out of a loudspeaker at the expense of performance of that loudspeaker. My system is only 93 dB/W/m, but I have 60 WPC.
My personal opinion with SET amps is that they are more a novelty amp and not something I would use for serious listening. A class A or AB push pull would yield more power and better result, which gives you a wider array of choices for drivers.
A second tact to increase efficiency is to use two woofers in parallel. This gives you a 3 dB boost in efficiency, but you are required to double the cabinet volume. If you have a lot living space (and money) you can get away with this, but you cut the impedance in half and the amp still needs to handle the lower impedance (i.e., 4 Ohms).
Lastly, you can further improve the performance of the system by bi-amping and using an active crossover instead of the passive crossover. The bass (500 Hz and lower) is where 50% of your amp's power is required. So, building a second stereo SET would give you two things. First, you would gain 3 dB in system power. Second, you would eliminate the low end crossover for the woofer, which will make the woofer operate more efficiently without the huge inductors.
Using an 18" woofer implies that you are at least considering a 3-Way system. You can get away with a passive crossover for the mid and tweeter and then use an active solution for the woofer to mid. The results would be superior to a passive 3-way crossover and nearly as good as a fully active system.
I would be interested in seeing what your plans are for the rest of the system and what drivers you are considering. My conclusion would be to either reconsider the Eminence driver, possibly substituting the JBL or reconsider the way you amplify your system.
Last edited:
You might consider doing an IB thing, if you own your home... You'll get all the way to 28Hz. then.
But 40Hz. is quite reasonable!
Most systems although they say they are, are not really flat to 40Hz.
Some are, and some go lower of course.
Your box calc does not take into account "room lift" or room placement either.
Depending on what that does and where you put the boxes that will have a big effect. In the corners, you will get significant boost as you go lower.
Placed in a large triangular pyramid like box and shoved into the corners with the driver low to the floor will give you rather remarkable results, depending still on the actual room and floor materials... etc.
Using a SE amp, your Qt figure will likely tend to be a bit higher than the actual figure in practice...
Although as the previous poster suggested a biamp'd system has numerous merits. If ur going tube only, I'd suggest a P-P amp for the bottom end myself.
Also, the lower you go with the "mid" horn the better... check the L'Clerc'h horn thread(s) for one way to go there...
SE amps are not novelty amps - they are not all equally good, but matched properly to the right speakers they are capable of magic. Improperly designed, built and matched, they are not very good.
Definitely SE amps are not "best" in most cases for bass drivers.
Paralleled woofers are something to consider. You get 6dB if you count the halving of the impedance... but in a biamped system, who cares? Just set the levels properly!
_-_-bear
But 40Hz. is quite reasonable!
Most systems although they say they are, are not really flat to 40Hz.
Some are, and some go lower of course.
Your box calc does not take into account "room lift" or room placement either.
Depending on what that does and where you put the boxes that will have a big effect. In the corners, you will get significant boost as you go lower.
Placed in a large triangular pyramid like box and shoved into the corners with the driver low to the floor will give you rather remarkable results, depending still on the actual room and floor materials... etc.
Using a SE amp, your Qt figure will likely tend to be a bit higher than the actual figure in practice...
Although as the previous poster suggested a biamp'd system has numerous merits. If ur going tube only, I'd suggest a P-P amp for the bottom end myself.
Also, the lower you go with the "mid" horn the better... check the L'Clerc'h horn thread(s) for one way to go there...
SE amps are not novelty amps - they are not all equally good, but matched properly to the right speakers they are capable of magic. Improperly designed, built and matched, they are not very good.
Definitely SE amps are not "best" in most cases for bass drivers.
Paralleled woofers are something to consider. You get 6dB if you count the halving of the impedance... but in a biamped system, who cares? Just set the levels properly!
_-_-bear
No way could you use the Eminence driver in an open baffle. F3 approaches 90 Hz and 28 Hz is 15 dB down.
However, what other IB driver do you know of that has a high SPL?
In any ported system designed for maximum flatness the lowest cutoff frequency is achieved with a woofer Qts of .383 or very slightly less if a passive crossover is used. Q other than this cutoff will always be above box frequency.
I have no idea why one would use an 18 at 1kHz and the importance of the response up there. That cross should be like 200Hz into a suitable mid.
As for that JBL mentioned it squawks so bad we spray the entire cone surface with the cheap undercoat from Walmart automotive department and set them in the Sun to dry.
Please remember, small ports don't work so if the port is not at least 1/3 the area of the woofer it is pretty much a waste of time. Port will go into power compression long before the driver runs out of ump.
Large drivers do not at all mean low bass automatically. I have a 3 inch here that goes down to 40Hz but it does not play loud at all as it runs out of cone excursion. For low bass choose in a small cabinet choose a driver with a low resonance and smaller cone with Qts= .383. All things being equal like Fs, resonance, and Vas the smaller driver will have a proportionally smaller cabinet. The reality of fabrication of a large cabinet like 16 cubic feet is quite difficult and takes a lot of bracing. A quality cabinet that size will weigh like 250 pounds- 110 kilos. That makes smaller better in a lot of ways.
I have no idea why one would use an 18 at 1kHz and the importance of the response up there. That cross should be like 200Hz into a suitable mid.
As for that JBL mentioned it squawks so bad we spray the entire cone surface with the cheap undercoat from Walmart automotive department and set them in the Sun to dry.
Please remember, small ports don't work so if the port is not at least 1/3 the area of the woofer it is pretty much a waste of time. Port will go into power compression long before the driver runs out of ump.
Large drivers do not at all mean low bass automatically. I have a 3 inch here that goes down to 40Hz but it does not play loud at all as it runs out of cone excursion. For low bass choose in a small cabinet choose a driver with a low resonance and smaller cone with Qts= .383. All things being equal like Fs, resonance, and Vas the smaller driver will have a proportionally smaller cabinet. The reality of fabrication of a large cabinet like 16 cubic feet is quite difficult and takes a lot of bracing. A quality cabinet that size will weigh like 250 pounds- 110 kilos. That makes smaller better in a lot of ways.
The idea is that you still have to deal with the cone breakup! Even though you apply a crossover slope of 6 or even 12 dB per octave, the breakup is so obnoxious that it will still intrude into the midrange frequencies.
At 200 Hz and 12 dB/octave you are only 24 dB down at 800 Hz and 36 dB down at 1600 Hz. Looking at the graph you have a wide 6 dB spike, which will require additional notch filtering to get flat. The more components you throw at this driver (i.e., steeper slopes and notch filters) the more power is dissipated as heat and not music. This is very bad for a flea powered amp.
If you cross at 200 Hz you need a midrange with very low Fs, at least 50 Hz if not lower, and it still needs to have a high SPL. The alternative is a 4-way, which is a monster to try to design.
Crossing low implies large inductors, which are both expensive and power robbing. It is better to cross as high as possible. 400 to 500 Hz is realistic, but then that aggravates the problem with the cone breakup.
The bottom line is the Eminence 18 is not a good driver for the design goals.
I do agree with you about port area and it would be prudent to design your system for higher power applications in the future (even though the SET will never develop enough port velocity to be audible), but that is a side issue compared to all the other problems he is trying to juggle.
Personally, I would disagree with you about driver size and bass capacity (particularly for his application). In this instance I would recommend at least 12" to 15" for deep bass production, assuming a very efficient driver. A 3" driver simply can not effectively produce a good polar response, let alone move enough air to be heard with a SET.
Leren42 is mostly correct.
There is that handy formula I do not have handy (oops) which is something like box size times cutoff frequency cubed times something else (?air density?) gives maximum achievable efficiency . If it comes out of the pile will post it but it is on other threads as have seen it. The key point is driver size is not in the formula. Small boxes with low cutoff frequency must have low efficiency. Larger boxes with the same cutoff will have higher efficiency if the correct driver can be found. Given the formulation one can pick a cutoff and box size and then calculate the efficiency or any other combination of known parameters.
If anyone has that formula handy would that person be so kind as to post it?
-SUM
There is that handy formula I do not have handy (oops) which is something like box size times cutoff frequency cubed times something else (?air density?) gives maximum achievable efficiency . If it comes out of the pile will post it but it is on other threads as have seen it. The key point is driver size is not in the formula. Small boxes with low cutoff frequency must have low efficiency. Larger boxes with the same cutoff will have higher efficiency if the correct driver can be found. Given the formulation one can pick a cutoff and box size and then calculate the efficiency or any other combination of known parameters.
If anyone has that formula handy would that person be so kind as to post it?
-SUM
Leren42 is mostly correct.
There is that handy formula I do not have handy (oops) which is something like box size times cutoff frequency cubed times something else (?air density?) gives maximum achievable efficiency . If it comes out of the pile will post it but it is on other threads as have seen it. The key point is driver size is not in the formula. Small boxes with low cutoff frequency must have low efficiency. Larger boxes with the same cutoff will have higher efficiency if the correct driver can be found. Given the formulation one can pick a cutoff and box size and then calculate the efficiency or any other combination of known parameters.
If anyone has that formula handy would that person be so kind as to post it?
-SUM
How can you determine box size with out driver size? That almost implies that one sized shoe fits all and we know that is not true.
I think the minimum amount of data you need to determine box volume is the driver Fs, Qts, Vas, Xmax, and Sd. Sd is directly related to driver size.
I should add, any changes to box volume or port tuning beyond the ideal is called a misalignment, which will impact the knee of the frequency response curve and the slope of that fall off. In the old days they gave those different misalignments names, which I no longer remember.
Last edited:
To design a specific box driver information is of course required. To know the maximum efficiency which can be achieved in a box of a given size with a given cutoff frequency no real info about the driver is needed.
The image is from one of the R. Small papers. It is not exactly what I was looking for but reveals the concept that F3 and maximum achievable efficiency are unrelated to driver cone size and dependent on box volume instead. That is my key point. Cone area size, Xmax, and cutoff frequency (F3) determine how loud the speaker can play at that cutoff frequency but have nothing to do with the efficiency available in a given box size with a given F3.
Of course there are practical limits as loren42 wrote. No 3 inch driver will play 120dB at 40Hz however this particular 3 inch mentioned will play 85dB at 40Hz which was its' design goal. A very very small speaker with bass for use more like headphones for a tiny room than in any kind of big system.
The image is from one of the R. Small papers. It is not exactly what I was looking for but reveals the concept that F3 and maximum achievable efficiency are unrelated to driver cone size and dependent on box volume instead. That is my key point. Cone area size, Xmax, and cutoff frequency (F3) determine how loud the speaker can play at that cutoff frequency but have nothing to do with the efficiency available in a given box size with a given F3.
Of course there are practical limits as loren42 wrote. No 3 inch driver will play 120dB at 40Hz however this particular 3 inch mentioned will play 85dB at 40Hz which was its' design goal. A very very small speaker with bass for use more like headphones for a tiny room than in any kind of big system.
Attachments
The driver Q is a bit low, so the low end response may droop and this will give you an F3 that seems "high" compared to tuning (for vented boxes). For closed boxes F3=Fb for Q=0.707 and for Q< 0.707 F3>Fb. Vented boxes will have similar thing going on, depending on the tuning used... If you increase the tuning frequency, the droop will go away, but you will then have a rather high tuning.
Typically with a Q around 0.3 in a vented box that is not "too small" and not "tuned high" you can expect to need some sort of "boost" around the tuning frequency. I actually prefer low Q drivers like this for 6th order vented and passive radiator alignments: the vented or PR box and driver, plus a second order HP filter, usually with 1<Qfilter<3. There is lots of info on this and it is nothing new. Go back to papers by Small and I think it is one of his alignments. It has the advantage that cone excursion can be made to drop off to zero below tuning because of the high-pass filter. This eliminates problems with out of band signals causing large cone excursions and leading to distortion by generating harmonics within-band.
If you want more flexibility in designing your system, ditch the "formulas" for design programs that are much more flexible and let you check for vent resonances, add an active circuit in to the response, etc. Check out:
Unibox
Woofer Box and Circuit designer
etc.
Also, if you want to increase Qts, you can do it a couple of ways:
-Charlie
Typically with a Q around 0.3 in a vented box that is not "too small" and not "tuned high" you can expect to need some sort of "boost" around the tuning frequency. I actually prefer low Q drivers like this for 6th order vented and passive radiator alignments: the vented or PR box and driver, plus a second order HP filter, usually with 1<Qfilter<3. There is lots of info on this and it is nothing new. Go back to papers by Small and I think it is one of his alignments. It has the advantage that cone excursion can be made to drop off to zero below tuning because of the high-pass filter. This eliminates problems with out of band signals causing large cone excursions and leading to distortion by generating harmonics within-band.
If you want more flexibility in designing your system, ditch the "formulas" for design programs that are much more flexible and let you check for vent resonances, add an active circuit in to the response, etc. Check out:
Unibox
Woofer Box and Circuit designer
etc.
Also, if you want to increase Qts, you can do it a couple of ways:
1. Use a dual voice coil driver. Connect a resistor across one set of coils and drive the other. This is call resistively damped operation:
http://stereointegrity.com/Files/RDOOperation.pdf
http://stereointegrity.com/Files/RDOOperation.pdf
2. For a single voice coil driver, add a resistor in series to increase increase Qts. This also dissipates power, so you need a power resistor.
3. You can add negative feedback around the amplifier using a small value (e.g. 0.1 ohm) sense resistor. This results in an increase of the amplifier's output impedance, which causes an increase in Qts (but you would probably want only that driver connected to the amp). This does not cause power dissipation, except in the sense resistor (which is small and thus negligible).
-Charlie
Thanks for the good link.
Constant K in Small's paper is related to Vas.
Vas = rho*c^2*Cms*Sd
Where rho is the density of air and c is the speed of sound. So there appears to be a direct relation of Sd and Cms (mechanical compliance) for calculating Vas. We can assume rho*c^2 can be though as a constant if all calculations are always based on STP.
I'm just having a hard time wrapping my brain around the idea that you can speak in terms of a relationship between box efficiency with respect to absolute box volume and box resonance frequency without discussing the driver which is used.
I did not mean to hijack this discussion. The topic is interesting, but I hope the original poster does not mind the diversion nor find it confusing with his original request.
Maximum box efficiency
For a sealed box the maximum efficiency is:
0.000002 * F3*F3*F3 * Vb = output in acoustic watts where 120dB = 1 acoustic watt.
Where Vb is in cubic meters, F3 is in Hertz and the constant is for a sealed box and spherical radiation. Half sphere (normal situation) the constant is double this value = .000004
For any vent, port, or passive radiator the constant is double the value shown as the energy from the back of the driver is used so .000002 becomes .000004 and for half space becomes .000008
These are the limits of maximum efficiency versus cutoff for sealed and vented enclosures in spherical and half space arrangements. Please notice drives area is not part of the formula. No matter the tuning these values cannot be exceeded in vent or sealed arrangements. For bandpass and horns of course this does not apply.
Driver specifications only affect maximum SPL and so on and do not affect box efficiency. Maybe I will find the paper that shows this by Small. Know it was a much later paper. This is from my notes of design.
One can always make boxes that are less than maximum efficiency for the box.
I think what happened is the thread starter said making the box bigger did not help much and I said driver size does not matter and then someone else said how can the driver size not matter for efficiency or something like that and then we got this side track on box efficiency.
This guy is probably looking for a moderately high efficiency driver that will come in the form of a 3-4 inch voice coil 12 or 15 inch driver that will fit into the box of a size he wants and go high enough to suit the need. Look for a blasting 12. If I see one will post the link. Maybe part-express has one or someone else.
Note: 1 cubic meter equals 35.31 cubic feet.
For a sealed box the maximum efficiency is:
0.000002 * F3*F3*F3 * Vb = output in acoustic watts where 120dB = 1 acoustic watt.
Where Vb is in cubic meters, F3 is in Hertz and the constant is for a sealed box and spherical radiation. Half sphere (normal situation) the constant is double this value = .000004
For any vent, port, or passive radiator the constant is double the value shown as the energy from the back of the driver is used so .000002 becomes .000004 and for half space becomes .000008
These are the limits of maximum efficiency versus cutoff for sealed and vented enclosures in spherical and half space arrangements. Please notice drives area is not part of the formula. No matter the tuning these values cannot be exceeded in vent or sealed arrangements. For bandpass and horns of course this does not apply.
Driver specifications only affect maximum SPL and so on and do not affect box efficiency. Maybe I will find the paper that shows this by Small. Know it was a much later paper. This is from my notes of design.
One can always make boxes that are less than maximum efficiency for the box.
I think what happened is the thread starter said making the box bigger did not help much and I said driver size does not matter and then someone else said how can the driver size not matter for efficiency or something like that and then we got this side track on box efficiency.
This guy is probably looking for a moderately high efficiency driver that will come in the form of a 3-4 inch voice coil 12 or 15 inch driver that will fit into the box of a size he wants and go high enough to suit the need. Look for a blasting 12. If I see one will post the link. Maybe part-express has one or someone else.
Note: 1 cubic meter equals 35.31 cubic feet.
Last edited:
That is a very general statement. Small ports can work if the designer knows the operating limits. Something like WinISD will tell us very easily what the velocity and max power requirements are, plus we are not talking full range here so the XO point could be 80Hz and a sub included reducing the port velocity even more. There is always a compromise somewhere and if a 15" or 18" woofer is used as part of a main speaker we probably don't need to see 6" ports. Those size ports are left for Sub boxes.
Why Fs = 50Hz, I thought we only need to be an octave higher so 200Hz XO requires a mid range with an Fs = 100Hz or better?
btw, I do agree that an XO at 500Hz is more pratical just curious to the XO vs Fs comment.
Last edited:
You want at least two octaves and three preferably. A second order filter slope only puts the signal at -12 dB down one octave away. If there is a resonance at that point (or nearby) and a corresponding rise in level, it will be added into the merge of the woofer's response curve and the midrange's response curve.
You could make a higher order filter as a means to get around the problem (or add a notch filter), but the components for a 200 Hz filter are going to be big and expensive. The more parts you add down there will cause the cost to skyrocket and decrease the damping factor between woofer and amp.
Usually as in almost always driver phase response goes wild quite a ways before the frequency response does. As example about any of the curved cone woofers no matter the brand in 12", 15", and 18" flip phase about 200Hz. This makes practical crossovers a real issue. Drivers that are in phase with constant directivity which automatically means rolling off of the high end and play 5 octaves are pretty much considered full range drivers. This translates into 3 octaves of overlap on a 5 octave driver... at least a 4 way if not a 5 way system. loren42 builds to some pretty exclusive standards so I commend him for that but do not quite know where to find these wideband drivers
Still looking for a straight cone 12 for your application. That will go the highest and will likely have the efficiency your looking for in a box that may be managed. If found will link to it. As a note it will be a 2 layer voice coil straight cone with a Q in the .35 area and Fs of like 30-35Hz. I actually will not tune ported systems between 20Hz and 40Hz because of the delay at the port. The port output is one full cycle behind the active driver. Tuning at 30Hz means 33ms of delay and that is getting very close to echo. 40Hz is 25ms delay and not so bad.
Keep working at it- there is a solution in here somewhere!
Still looking for a straight cone 12 for your application. That will go the highest and will likely have the efficiency your looking for in a box that may be managed. If found will link to it. As a note it will be a 2 layer voice coil straight cone with a Q in the .35 area and Fs of like 30-35Hz. I actually will not tune ported systems between 20Hz and 40Hz because of the delay at the port. The port output is one full cycle behind the active driver. Tuning at 30Hz means 33ms of delay and that is getting very close to echo. 40Hz is 25ms delay and not so bad.
Keep working at it- there is a solution in here somewhere!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.