Stephen D said:
No, they do not.
That premise means that by simply increasing the number of drivers I buy, I will eventually be able to exceed 100% efficiency..which is not possible.
I seem to recall Abraham Coen stating the 3db thing, when I read him back in '74, and immediately understood the implication of >100% efficiency.
I believe he mistook the method used for testing efficiency, by not taking into account the additional current draw which occurs when the drivers are paralleled and fed the 2.83 volts specified.
I do concur that efficiency will increase for cases where the driver efficiency is compromised by the acoustic impedance mismatch between the air and the driver, but as the radiating surface increases unbounded, the efficiency can only approach 50% (assuming closed back). For more on that, check out Tom Danley at servo.
Your premise violates conservation of energy.
FWIW, my array design is for reducing the near field SPL, as the intended location is too close to some seats. 12 by 5, with some spare drivers.
For major sound levels in a small box, I use a Delta pro 12 incher, with a selenium d205ti behind a 60 by 40 horn. Plenty of mid bass and highs for a 200 person DJ app (weddings and the such, certainly not for a club) on stands, with a big cab pushing a kilo of bass. The 12" cabs are decent enough to use standalone, for the smaller apps. The cab is 15 by 13 by 24(inches), easily moved and transported.
The delta's are about 100 dB spl, rated 400 WRMS, the 205's xover'd 5 K, 3rd order, and 105 dB spl. Easily handle a 300 Wrms amp with clip lights keeping the beat..
Cheers, John
If you search the archives, you will find this:
http://www.lsv-achenbach.de/basics/db_spl_no.htm
showing that 112 dB spl is about 100% efficient..
Can't exceed 100% efficiency..
Cheers, John
http://www.lsv-achenbach.de/basics/db_spl_no.htm
showing that 112 dB spl is about 100% efficient..
Can't exceed 100% efficiency..
Cheers, John
jneutron said:
Not long ago, that was PRECISELY my argument against the 3dB gain in efficiency premise. It seems though it is a well accepted premise included even in the efficiency calculations of speaker design software programs such as WinISD pro. If you have the software (It's freeware), punch in the # of drivers used, select a moderate Q sealed alignment & then check the SPL graph with 1 watt input & see what you get as an efficiency rise.
I already know it will show 3dB increase for every doubling of drivers (at least in the passband above box roll off or Q effects).
😕
Stephen D said:
Not long ago, that was PRECISELY my argument against the 3dB gain in efficiency premise. It seems though it is a well accepted premise included even in the efficiency calculations of speaker design software programs such as WinISD pro. If you have the software (It's freeware), punch in the # of drivers used, select a moderate Q sealed alignment & then check the SPL graph with 1 watt input & see what you get as an efficiency rise.
I already know it will show 3dB increase for every doubling of drivers (at least in the passband above box roll off or Q effects).
😕
Software many times is incorrect. GIGO
Try 512 drivers at 86 dB per...the program will give you 113 dB spl 1w/1meter.
113 is over 100% efficiency...perpetual motion..
The software writer assumed incorrectly..
Cheers, John
jneutron said:
I hear ya, but the acceptance of that premise goes much wider than just an individual software. As I said it also seems a flawed premise to me, but I'm curious as to it's wide acceptance & what the real relationship is or what I may be misunderstanding because it sure doesn't seem quite right to me either.
Yah, I know...but it is indeed incorrect as a general rule, and will apply easily for a coupla 4 inchers for wavelengths much larger than the drivers, and low efficiencies.
For speakers that are acoustically impedance challenged, yes the rule works..but as you increase the array size, it can only asymptotically approaches 109 dB spl, and that for wavelengths smaller than the array size.
Wide acceptance does not mean accurate..it is a generic rule with certain bounds...obviously the bounds have not been included in the software described. They probably never saw a Parts Express closeout flyer..😀
Cheers, John
jneutron said:
The softwarewriter may have assumed that all of them was connected in paralell, giving 113dB at 2.83v
electroaudio said:
That is exactly what I think also...So, the program should not state "113 dB 1watt/1 meter, but should instead say 113 dB 2.83 v/1 meter...course, that would be 512 watts input..
The program should of course have some means of showing how the lf response is getting better as the array grows.
As to using an array of those for DJ apps...running an array of those at full power, 200 watts in his case, means getting those babies very hot with the continuous use...they will certainly overheat, as they are not designed to handle that much continuously. For DJ apps, nobody should run that close to failure.
Word of mouth on a total blowout would ruin future business..
Cheers, John
reply
You can use small speakers in line array.
Have a look on the Eminence website they have 3 special new 6.5 inch line array drivers.
2 midranges,1 with a sealed back and 1 mid bass LA6-MB
www.eminence.com
Also you could use the Alpha 6,Alpha 8,Beta 8MR,Beta 8 in line array.Eminence Beta 8 handles 225 w rms and has 2 inch vc.
Something like [8] Eminence Alpha 6's[100w rms] with [4] Eminence APT80 tweeters plus x over would make good sounding mid-top array.
And then a single 15 inch per side on the floor in a ported cab for bass.At least this array would be able to handle serious power needed in a pa application.😎
You can use small speakers in line array.
Have a look on the Eminence website they have 3 special new 6.5 inch line array drivers.
2 midranges,1 with a sealed back and 1 mid bass LA6-MB
www.eminence.com
Also you could use the Alpha 6,Alpha 8,Beta 8MR,Beta 8 in line array.Eminence Beta 8 handles 225 w rms and has 2 inch vc.
Something like [8] Eminence Alpha 6's[100w rms] with [4] Eminence APT80 tweeters plus x over would make good sounding mid-top array.
And then a single 15 inch per side on the floor in a ported cab for bass.At least this array would be able to handle serious power needed in a pa application.😎
Attachments
electroaudio said:
No, that's not the case. The SPL calculations in the program are strictly concerned with power sensitivity & not voltage sensitivity & the program help file makes that clear. The tutorial in the help file even makes specific mention of the 3 db acoustic power efficiency gain per doubling of drivers & then mentions this should be added to the 3 dB voltage sensitivity gain from lowering the ohms when electronically paralleling drivers, to give a total 6 dB increase in sensitivity when paralleling drivers. I think they just don't fully understand what the truly unbounded algorithm would be & just kept it simple, as in the majority of cases of just a few drivers 2 - 8, as a rule of thumb the 3 dB premise probably holds up to close enough scrutiny at low freq of longer wave length than the driver diameters (The program only deals with freq 500Hz & below).
Thanks to this thread getting me thinking again ( thinking is good no? 🙂 ) I think I have a more accurate picture in my head of how it actually plays out though.
😉
jneutron said:
This might be the perfect chance to hook up those 512 drivers you spoke of and give this book theory a try. See what actually happens.
Maybe pass the hat around, I'd bet you'd get some donations. I'll chip in.
Now, who's gonna build the baffle?
Sad Cal
Canucks lost tonight🙁
Wow, what a tangent we've gone off on.
I got the bass drivers for my PA system today. The boxes aren't done yet. Parts express #290-380. Go look it up. I've already linked it twice in this thread, I think. These drivers will be in 2cuft vented boxes, tuned to 40hz, so with a bit of room gain, they'll play below 40hz no problem.
I'm impressed with the drivers. I can already tell that the motor on them is fantastically powerful. The two drawbacks are the lack of linear excursion (5mm p-p is realistic; it gets nonlinear beyond that very fast) and that the motor is balanced for a 10" driver and looks pretty crummy in anything bigger.
Paulinator,
while I haven't heard the drivers in boxes yet, they sound like they'll integrate well with my current satellites. If I were to design the system for someone else (you?) I would make an active crossover and power them with gainclones. The crossover would have to be more on the complex side--second-order from the woofers to the satellites, with a treble boost and very wide, but shallow, midrange cut.
When the woofer cabinets are finished, I'll let you know how much sound I can produce, and how good it sounds.
I got the bass drivers for my PA system today. The boxes aren't done yet. Parts express #290-380. Go look it up. I've already linked it twice in this thread, I think. These drivers will be in 2cuft vented boxes, tuned to 40hz, so with a bit of room gain, they'll play below 40hz no problem.
I'm impressed with the drivers. I can already tell that the motor on them is fantastically powerful. The two drawbacks are the lack of linear excursion (5mm p-p is realistic; it gets nonlinear beyond that very fast) and that the motor is balanced for a 10" driver and looks pretty crummy in anything bigger.
Paulinator,
while I haven't heard the drivers in boxes yet, they sound like they'll integrate well with my current satellites. If I were to design the system for someone else (you?) I would make an active crossover and power them with gainclones. The crossover would have to be more on the complex side--second-order from the woofers to the satellites, with a treble boost and very wide, but shallow, midrange cut.
When the woofer cabinets are finished, I'll let you know how much sound I can produce, and how good it sounds.
By using multiple drivers there WILL DEFINITELY BE a rise in efficiency. The "doubling rule" however will only apply up to a certain point.
If there would be no rise in efficiency, this design
http://www.tactaudio.com/Products/Speakers/Speakers_LS1_TopView.htm
wouldn't achieve 102dB/1W/1m by using those low-efficiency Scan-Speak drivers.
Furthermore line-arrays wouldn't be that fashionable amongst professional users if there was no rise in efficiency.
BTW: It is perfectly possible to have more than 112 dB/W/m with an efficiency significantely lower than 100 % ! Ever heard anything about directionality ?!?
Regards
Charles
If there would be no rise in efficiency, this design
http://www.tactaudio.com/Products/Speakers/Speakers_LS1_TopView.htm
wouldn't achieve 102dB/1W/1m by using those low-efficiency Scan-Speak drivers.
Furthermore line-arrays wouldn't be that fashionable amongst professional users if there was no rise in efficiency.
BTW: It is perfectly possible to have more than 112 dB/W/m with an efficiency significantely lower than 100 % ! Ever heard anything about directionality ?!?
Regards
Charles
Totally wrong, i work with professional systems at stadiums and we have the exact amount of amplifier (The same actually) nowadays with our new linearray (JBL Vertec) as we had before with our conventional system (JBL HLA).
-And the soundpressure is lower with vertec than hla
The reason linearrays are popular amongst proffessionals is mostly because of a hype and partly because it is possible to calculate the soundpressure on every seat.
I did not want to say that effieciency is THE reason for using them. AFAIK a HLA is already a very high efficiency system and not that easily topped in this respect.
But if the efficiency of complete line arrays were as low as single speakers of the same size (as one array-element) they wouldn't be used I guess. Or am I wrong ?
BTW: The HiFi line-source mentioned above should be proof enough for the increase in efficiency for multiple drivers.
Regards
Charles
But if the efficiency of complete line arrays were as low as single speakers of the same size (as one array-element) they wouldn't be used I guess. Or am I wrong ?
BTW: The HiFi line-source mentioned above should be proof enough for the increase in efficiency for multiple drivers.
Regards
Charles
phase_accurate said:
Hi Charles.
You are entirely correct...Throughout my entire posts, I've not considered the directionality (beaming) associated with a phased array setup. For multiple speakers distributed as a sphere, pointing inward, the absolute spl will certainly go up 3db per doubling..and for a flat array, the low end extension will improve along those lines with array size where driver size limits low end coupling.
I have only been considering far field applications, as he did mention DJ use. For that, I consider only the efficiency of the drivers, the volume of the room, the absorbtion of the room, and the power used. I am talking spherical uniform propagation, more the bass end, while you speak of beaming.
Oh, BTW...my pa speaks, the eminence 12's with the selenium d205ti? I run the 12's up to 3Khz, then 1st order lp, while the tweet is 3rd at 5K...this setup will be prone to beaming before the xover point. I adjusted the rolloff to accomodate that. In large venues, I eq when off axis drops to fast.
Hey Nappy:
I also like reflex for the added boost, but watch out: If your amp has a 50 hz filter, use it. You really don't want to bottom out the VC during gigs. Freq's lower than the tune will allow driver excursion to go unchecked, as the coupling to the room will not help you too much..and powers significantly below what the driver can handle easily could damage them. Placement of the drivers at a gig could tend to mask the hits of the VC, you don't need that.
If you intend to scratch, use a 2nd or 4th order hp at about 40, you'd be surprised how much lf can get through the system to wreak havoc on a vented woof.
Good luck..post some pics..
Cheers, John
phase_accurate said:
True, but increased directivity & increased efficiency are not really the same thing. Increased directivity merely focuses the same total acoustic energy/power/efficiency into a tighter space aka, takes SPL that would be distributed elsewhere & sums it into a tighter beam.
One thing I suspect is, the degree to which efficiency is increased by just going from 1 to 2 drivers is wavelength & driver diameter dependent & is proportional to Wl/Dd.
where -
Wl = wave length
Dd = driver diameter (or cross section)
In other words it would be greatest at wave lengths much greater than the driver diameter & taper off as wavelengths approached & diminished below driver diameter. 3 dB being the theoretical maximum increase & applying only to frequencies somewhat longer than the driver diameter, with wavelengths shorter than the diameter going from little to no increase in efficiency (tapering off as wave length shortens aka. freq rises).
Taken further this would imply that as you continue to double the # of drivers in an array, the cross section of the array could be viewed as a single driver cross section (diameter) & as the array grew, at each doubling the transition freq point where there would be little to no increase in efficiency (from that freq on up) would move further & further down into the bass range. If all this is so, an interesting side effect would be a change in freq response between an array & the single driver, with the array exhibiting a sort of shelving low end boost starting up in the mid to high band (depending on the initial size of drivers). This would likely manifest itself to the listener more as a perceived HF roll off, which interestingly some who have built arrays report. Of course HF combing also plays a part in that.
Disclaimer: I've yet no documentation or empirical experimental evidence to support the above theory. It is merely the product of my sick mind. I have at this point only thought experiments to support this... yet more products of my sick mind.
FWIW, I think in general, wide horizontal dispersion (low directivity) makes for a superior PA (except in special cases where a specific small area is targeted). One major goal of a PA (weather for DJ or live band) is even coverage of both freg response & SPL. This is contrary to the goal of say a studio monitor where high directivity to a single sweet spot is generally preferred to minimize room coloration. Tall thin vertical line arrays with their wide horizontal & narrow vertical dispersion + resultant near field roll off in SPL @ only 3 dB per distance doubling (rather than 6 dB) are an appealing PA choice for distributing FR & SPL evenly to an audience.
🙂
So is the vertecs...
But today, the efficiency in pa speakers arent that critical as it was in the old days when the linearray theories first were found.
So if there would have been any increase then linearrays would have been used then, and not today when powerful amps is both lightweighted and cheap rather than nonexistant.
The use of linearrays in PA is that they give the ability to adjust the speakers within a couple of degrees from eachother and by doing so, is it possible to calculate angles that gives the same spl over the whole venue.
I just looked at your example and they had custombuilt speakers in them so they dont say anything.
And my experience with HLA and VERTEC is that the VERTEC has lower spl than HLA with the same amps.
They also have the same type of JBLs highefficient speakerelements within the boxes so if the theory was true then vertecs ratio of 16 to 1 would give a remarkable much higher spl than HLA , but they dont.
Stephen D said:
This is why I am hoping to figure a design that can get good, CONTROLLED dispersion without using horns. Maybe If each speaker has 2 arrays of mid and high frequency drivers, each pointing slightly away from each other, I can get good, large area coverage without using horns, and I was also considering having the mids in each array flanking the sides of the tweeter array so that I can control the tweeter's dispersion a bit. Of course, this would only be necessary, I think, if I did not use a slot or bullet tweeter.
From what I am reading here, I might be better off using an a small array of larger mids than a huge array of smaller drivers, because these will be strictly PA speakers. Parts Express has some good looking midrange drivers from Selenium that I might look into.
http://www.partsexpress.com/pe/showdetl.cfm?&DID=7&Partnumber=264-330
I am still tempted to use twice as many of a driver like this just for the visual effect that you never see in PA speakers. 🙂 Probably wouldn't sound bad either, they have pretty decent power handling.
http://www.partsexpress.com/pe/showdetl.cfm?&DID=7&Partnumber=290-309
What is the advantage of those Eminence PA mids compared to a good, high power set of normal mids, maybe some Audaxes or Vifa's?
- Status
- Not open for further replies.