|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
134.47.109.182
If you have 2 indentical speakers in serie, the current is halved and the impedance is doubled. - With the same voltage, the efficience ( dB/W )would decrease to the one half.P( one speaker ) = I^2xR = 1^2x1 = 1 ( 0 dB )
P ( two speaker in serie ) P = 0,5^2x2 = 0,5 ( - 3dB )Sombody tell me if this is wrong, and if so, please, please why !!
Follow Ups:
Your math is correct, the total output will decrease by 3 db or one half the power. Efficiency is an inherent property of the speaker itself. For the same amplifier voltage assuming you do not adjust the gain, you will deliver half the power to the two systems combined so you will have to increase the gain by 3db to get back to the same power level level as with one system. Efficiency is technically the acoustic watts generated divided by the electrical power dissipated. It ranges usually from a fraction of a percent to a few percent. even the most efficient loudspeakers are not efficient energy converters. Sensitivity is another way of looking at it but it is not exactly the same as efficiency. Measured on axis in an anechoic chamber, it does not take into account the total acoustic energy radiated which depends on dispersion. Sensitivity is also an inherent property of the speaker itself, not how it is connected.
Soundmind, you are absolutely correct that efficiency is an inherent property of the driver itself. And whether you connect multiple drivers in series, in parallel, or in series-parallel, their efficiencies add exactly the same way.However, if you have more than one driver covering the same frequency band, how they are connected does determine the sensitivity of the array. For example, suppose you have four identical 8-ohm drivers mounted together on the same baffle. Each driver's efficiency is 90 dB at 1 meter for a 1 watt input, and each driver's sensitivity is 90 dB at 1 meter for a 2.83 volt input.
Since their efficiencies add the same regardless of how they are connected, the array will have the same (96 dB) efficiency regardless of whether the drivers are wired in series, in parallel, or in series-parallel.
However, the array's voltage sensitivity changes dramatically according to how the drivers are are wired:
Wired in series the four drivers will present a 32-ohm load, and 2.83 volts into 32 ohms is only 1/4 watt. But that 1/4 watt is moving 4 times the cone area (the current passes through all four voice coils), so the same total amount of air is moved as for a single driver (1/4 times 4 = 1). The array's voltage sensitivity remains the same as for a single driver at 90dB for a 2.83 volt input.
Wired in parallel the four drivers would present a 2 ohm load, drawing 4 watts for a 2.83 volt input. That's 4 times the power moving 4 times the cone area, or a 16-fold increase in air movement for a 12 dB increase in voltage sensitivity (raising the array's sensitivity up to 102 dB for 2.83 volts input).
With series-parallel connection the impedance presented by the four drivers remains 8 ohms, so they'd still draw 1 watt, but with 4 times the cone area now we're moving 4 times as much air, so we'd get a 6 dB increase in output. The voltage sensitivity of our series-parallel array would be 96 dB for a 2.83 volt input.
Click on the link for a useful online calculator.
"With series-parallel connection the impedance presented by the four drivers remains 8 ohms, so they'd still draw 1 watt, but with 4 times the cone area now we're moving 4 times as much air, so we'd get a 6 dB increase in output. The voltage sensitivity of our series-parallel array would be 96 dB for a 2.83 volt input."This is not true. The amount of power dropped over each driver in comparison to a single driver would be 1/4th in the series / parallel set up. Regardless of how you wire them, there is no freepower. The resultant output would be the same. The amount of cone deflection would be less (1/4 of the original amount (?)), thus reducing ultimate distortion.
Hi Ozzy,I'm not claiming "free power" for an array of identical drivers. I'm claiming "increase in efficiency" and "increase in voltage sensitivity", the latter dependent on how the drivers are wired.
I think you're overlooking what's happening to the driver parameters in the multiple driver case.
As background, let me state that driver efficiency is predicted by three parameters: Fs, Qes, and Vas. This from Bullock, whose book I can't put my finger on right now (it's around here somewhere but I don't feel like conducting a housewide search for the sake of this post).
Now, consider that identical four woofers in the same box can be modelled as a single larger woofer. If you like, think of the array as in effect a single woofer with independent suspension systems for its four quadrants, and driven by four voice coils.
With this four-driver array, what parameters change as compared to a single driver? Fs stays the same, as does Qes. But Vas is quadrupled.
Now, if you happen to have a woofer modelling program handy, punch in the parameters of a woofer and see what the predicted efficiency is. Then, punch in a Vas four times greater, and see what happens to the predicted efficiency. Guess what - it goes up by 6 dB!
If you don't have a woofer modelling program, there's a share-ware demo version of one at this address: http://www.ggimages.com/spsim
Also, I just took some measurements of an array of identical drivers wired in series, in parallel, and in series-parallel, link below. The measurements corroborate my prediction that "with series-parallel connection... we'd get a 6 dB increase in output."
Not trying to be a smart-ass here, but my measurements contradict your statement that "regardless of how you wire them... the resultant output would be the same." I would agree that regardless of how you wire them the efficiency would be the same (and with 4 identical drivers in a cluster on the same baffle you'd have a 6 dB efficiency increase over a single driver). But the voltage sensitivity is very much determined by how you wire them.
Duke
- Actual SPL measurements of different wiring configurations, constant voltage input (Open in New Window)
For the sake of clarity and continuity, I should have pointed out that the reason the Vas quadruples with four woofers (as compared to just one) is that the cone area quadruples while the compliance stays the same. So it goes back to the increase in cone area, as I'd previously stated elsewhere in this thread.
The sensitivity of the array depends on how they are connected. The sensitivity and efficiency of each element in the array doesn't. I think we are saying the same thing.
Sorry I didn't understand your first post; sometimes I'm kinda dense. Your post just above is precise and to the point:"The sensitivity of the array depends on how they are connected. The sensitivity and efficiency of each element in the array doesn't."
Cheers,
Duke
I think you are not factoring in the doubling of cone area from adding the second woofer, which brings the output back up by 3 dB for a net output level change of zero, assuming the same voltage input (which would be half the wattage input into the now-16 ohm load).I have taken to using the term "sensitivity" when referring to the output level for a given VOLTAGE input (such as 2.83 volts, which equals 1 watt into an 8 ohm load); and using the term "efficiency" when referring to the output level for a given WATTAGE input (usually 1 watt). I think this nomenclature is appropriate, but welcome correction from anyone who actually took some EE classes in college.
So, in the case of two 8-ohm drivers wired in series to give us 16 ohms, the sensitivity stays the same as for a single driver but the efficiency goes up by 3 dB.
If we wired those same two 8-ohm drivers in parallel, the sensitivity would go up by 6 dB and the efficiency by 3 dB.
Note that in both cases the change in efficiency is the same. Note also that manufacturers don't necessarily follow these same naming conventions, and may claim that a 4-ohm speaker is "93 dB efficient" when it is really "93 dB sensitive with a 2.83 volt input, which equals 2 watts into their 4-ohm impedance".
no change in db. I built a sheet to calculate this once. Umm. www.audioDIYcentral.comI think it is on the left side. Then "conversions" tab.
I'm not sure why we're back again. I thought the explanation was pretty clear.
Well, it,s easy to tell the answere. I allways like to dicuss how things work out.
To double the cone area itself don,t have not any effect if the movement is less by less current. How can you explain that when each speaker ( in serial ) get the half of the current over the same impedance each cone will move excatly as it was alone ?
"How can you explain that when each speaker ( in serial ) get the half of the current over the same impedance each cone will move excatly as it was alone?"Each cone's movement is definitely NOT "excatly as it was alone". Rather it only moves half as far. But, there are TWO cones. I think you're missing the point about the doubling of cone area, so let me try again (pardon the liberal use of caps, but I'm trying to highlight the main points):
As you correctly surmise, for a given input voltage two identical drivers wired in series only get half the current passing through them as a single one would. But, ALL of that current passes through BOTH drivers. So EACH driver's cone will now move HALF as far as it would have in the single-driver case. And remember, now you have TWO cones - twice the cone area. TWICE the area times HALF the movement equals the SAME amount of air moved as for a single driver. Move the SAME amount of air, get the SAME sound pressure level.
So for a given voltage input, two identical drivers in series will move the same amount of air and generate the same SPL as a single driver would, even though thusly wired they're only drawing 1/2 the current and therefore 1/2 wattage.
[As an aside, in case you'd like to play around with the relationship between voltage and amperage and resistance and wattage, there's a useful calculator at the link below. Enter any two parameters, and the other two are calculated for you.]
Duke, in this reply you just stated why adding two cones in series with the same Voltage input does not increase DB output vs one cone at the same voltage. Even though it was not mentioned by you, this is how lower distortion figures are arrived at. I am well aware of how the manufacturers manipulate their numbers with the sensitivity / efficiency thing. Heck anything over 90 Dbs does not need > 100 watts to properly drive a normal sized room.
Thank you for clarifying my proplem.I was avere of the area-doubling but was miksing up effect from the amplifier and values in the speaker itself. I also based on efficincy instead of sensitivity as I compared the effect encrease/decrease.
Using 4 units in parallel/serie gives 6 dB efficiency. As I have measured this many times we don,t get this value in praksis. In most cases we don,t get any better than 2- 4 dB depending of the baffel form and size, even if the speaker is mounted close together. Often we get most effenciency under 1 kHz. Whats your experience ?
I recently used four full-range drivers in a cluster wired in series-parallel, and the increase in sensitivity was pretty close to 6 dB. The tonal balance was modified, and was a bit more downward-sloping, indicating some destructive interference at higher frequencies and/or beneficial pattern narrowing at lower frequencies. However it wasn't really a fair comparison, as I used a much wider baffle for the 4-driver array which would push the baffle step down lower in frequency.Best of luck with your project!
Duke
How did you measured the effeciency of your speaker ?
Hi up,I happened to have four identical Pioneer full-range drivers in storage, so I hauled them out to take some measurements and see if my claims in this thread stand up to measurements taken in the real world. Rather than go to all the trouble of mounting them in a box and removing them to change the wiring, I set the drivers on the floor, unbaffled, face up. I took SPL readings with an analog Radio Shack SPL meter held upside down at shoulder height above the center of the drivers. No this in not great precision, but I'm claiming sensitivity differences of 0 dB, 6 dB, or 12 dB, depending on how the drivers are wired, and this set-up should easily resolve differences of that magnitude.
My source was Track 15 of Stereophile Test CD 2, which is the pink noise track. About halfway through Track 15 the volume drops a bit, so I used the first half of the track.
I adjusted the preamp's volume control to give me a reading of 70 dB at the approximate 4.5 foot height above a single driver, and the volume control setting remained untouched throughout the tests so that the amplifier's voltage output remained constant. Of course, the wattage output varies inversely with the impedance of the array of drivers.
With one driver (8 ohms; amp voltage output roughly .28 volts; amp power output roughly .01 watts), the reading was 70 dB.
With two drivers wired in series (16 ohms; amp voltage output roughly .28 volts; amp power output roughly .005 watts): Predicted reading 70 dB, actual reading 70 dB.
With two drivers wired in parallel (4 ohms; amp voltage output roughly .28 volts; amp power output roughly .02 watts): Predicted reading 76 dB, actual reading 77 dB.
With four drivers wired in series (32 ohms; amp voltage output roughly .28 volts; amp power output roughly .0025 watts): Predicted reading 70 dB, acutal reading 69.5 dB.
With four drivers wired in parallel (2 ohms; amp voltage output roughly .28 volts; amp power output roughly .04 watts): Predicted reading 82 dB; actual reading 81.5 dB.
With four drivers wired in series-parallel (8 ohms; amp voltage output roughly .28 volts; amp power output roughly .01 watts): Predicted reading 76 dB, actual reading 75.5 dB.
These reading are within 1 dB of those predicted by my explanation of the voltage sensitivity of multiple drivers wired in series, in parallel, or in series-parallel. I believe this is close enough to corroborate my claims.
This post is made possible by the generous support of people like you and our sponsors: