[imassarano]

Hi all,

I'm new on this circle, and I'm starting a new speaker project and looking for some help.

I'm planning to employ 3 wide-band drivers in a series/parallel array of 88dB 1W/1m  sensitivity each. What would be the resultant sensitivity of the array?

Thanks,
Isaac

[*Scotty*]

If I have understood your question and you have only 3 drivers on the front baffle then it will probably be around 88dB. Depending on where the speaker is placed relative to room boundaries and the bass 3db down point you may do a little better than that in a real living room. This is very generalized answer, there are a lot of variables in play.
You could chose drivers that are much more efficient and hit 94db with a 4ohm load in an MTM design and then come in on the bottom octaves with a powered sub. This gives you more control over the problematic bass regions and uses less amplifier power to do the job.
Scotty

[Russell Dawkins]

I am curious about the specifics of the series parallel wiring you intend with the three drivers. Two in series and those in parallel with the third?

[imassarano]

Thanks so far.

I'll try and clarify my question:

Perhaps it would be more accurate to say full range drivers as they are specified at 45Hz - 20kHz with F_s 45Hz, and I plan to use them full range with their obvious advantages.

Since they have Xmax 5mm and 30Watts RMS power handling with only moderate sensitivity of 88dB 1W/1m,  I would like to try employing 3 of them in an array. By wiring them  series/parallel, i.e. both series and parallel, would substantially increase both the power handling and sensitivity for much greater SPLs with only a moderate decrease in system impedance making for an easy load for an amplifier.

I just wanted to know what the system sensitivity would be. I believe there is a formula for it.

[JohnR]

You can't really run 3 full range drivers in parallel and expect to get full range response. There are tools available to help visualise this. I'd be inclined to suggest that you may be better off using just two drivers in a 1.5 way configuration, or alternately a bipole.

[doug s.]

i am not sure how you could run them in series and parallel.  you could run all in series, or all in parallel.  or, you could run two in series, and run that pair in parallel w/the third.  actual impedance would depend on the driver impedance, of course.  if you do not want to run them all in series, or all parallel, i would advise a number of drivers that allows equal number of combined series/parallel connection, if this is your goal.  ie: four drivers, two series pairs run in parallel.  or six drivers, three series pairs in parallel, or two series triples in parallel, etc...

as far as efficiency, the standard rule of thumb is doubling driver count will add 3db, so two 88db drivers will give you 91db; four will give you 94db, etc.  3 would be ~92-92.5db...

doug s.

[imassarano]

Many thanks gentlemen.

It seems like I'll go for 3 in parallel and take a hit on a lower net impedance.

[doug s.]

Many thanks gentlemen.

It seems like I'll go for 3 in parallel and take a hit on a lower net impedance.

if your drivers are 8 ohms, running three in parallel will give you a 2.67 load that your amp will see.  many good amps can handle this load, and some will output considerably more wattage into such a load.  but check your amp to be sure...

doug s.

[richidoo]

Many thanks gentlemen.

It seems like I'll go for 3 in parallel and take a hit on a lower net impedance.

That will give you the highest sensitivity, if your amp can handle the lowest impedance dip. The problem is Doug is right about 2.67 ohms if that is the lowest impedance dip. But most 8 ohm drivers will have dips much lower than that. If a single driver has a dip down to 4 ohms, then your amp will see a 1 point something ohm load. (1.33?? I don't know doug's cool formula.  )  With full bandwidth drivers the lowest impedance is usually around 100-300Hz, and there is a lot of music energy there. So just be careful that the amp doesn't overheat. If your drivers are 16 ohm nominal, then you should be OK with any decent SS amp.

[imassarano]

Actually, although they are rated at Nom. Imp. 8ohm, their plot shows a gentle dip to just bellow 9ohm, so that should be fine I guess. Though it would be a little more demanding to drive in terms of current demands, the 92dB+ sensitivity should do the trick .

Am I correct in saying that there will be a tripling of the power handling?

[doug s.]

That will give you the highest sensitivity, if your amp can handle the lowest impedance dip. The problem is Doug is right about 2.67 ohms if that is the lowest impedance dip. But most 8 ohm drivers will have dips much lower than that. If a single driver has a dip down to 4 ohms, then your amp will see a 1 point something ohm load. (1.33?? I don't know doug's cool formula.  )  With full bandwidth drivers the lowest impedance is usually around 100-300Hz, and there is a lot of music energy there. So just be careful that the amp doesn't overheat. If your drivers are 16 ohm nominal, then you should be OK with any decent SS amp.

rich is right - if the actual impedance dips below 8 ohms, the paralleling of them will also dip correspondingly.  the formula is easy; yust insert the lowest actual impedance as the walue:


doug s.

[richidoo]

Thanks doug, that's not so hard.

imassarano, there's tripling of the sensitivity compared to single driver, if the amp can deliver the current. Power handling is increased enough so that the drivers have the upper hand over the amp... but the enclosure effects power handling. I would guess that as long as they are all in the same cabinet (duh) then it is also tripled...

[doug s.]

Actually, although they are rated at Nom. Imp. 8ohm, their plot shows a gentle dip to just bellow 9ohm, so that should be fine I guess. Though it would be a little more demanding to drive in terms of current demands, the 92dB+ sensitivity should do the trick .

Am I correct in saying that there will be a tripling of the power handling?

yes, theoretically, if each driver can handle 30w, then a speaker w/three of them will be able to handle 90w.  series or parallel wiring won't change this, it will only change the power that the amp is capable of making.  (that's if it's a s/s amp; non-otl tube amps will output the same power regardless of the load they see.  otl amps actually put out more power when the impedance goes up.)

[imassarano]

Brill guys

I use SS amps, and my questions were directed with that in mind.

[*Scotty*]

A loudspeakers impedance consists more than just resistance alone,there is the phase angle of the load the amplifier has to deal with at the frequency the impedance minimum occurs at.
 It is this complex load that fries output stages. The make and model of your power amplifier and its suitability to drive this kind of load has not been brought up. If you had a Krell FPB 250M there would be no question of compatibility.
 You can always connect them series/parallel for a 5.33ohm load which your amplifier should be happier driving pretty much regardless of where the impedance minimum occurs at. Thats two 8ohms in series with the third driver in parallel to get 5.33ohms.
 You might on paper get 3dB+ from a tripling of cone area and maybe 2dB from the lower impedance drawing more power from the amplifier. That is an optimistic 93dB at wave-lengths up to twice the emissive diameter of the speakers cone.
  When the wave-lengths get shorter the drivers stop mutually coupling to the air load and start to destructively interfere with one another which reduces their effective output at higher frequencies. Where a single driver might have a flat response curve into the higher frequencies,multiple drivers covering the same bandwidth clear into the high frequencies may show a downward sloping response past the point where the drivers emissive diameter equals the emitted wave-length.
 There is no free lunch with an increase at efficiency at all frequencies when drivers are connected in parallel and their overlapping response extends into the shorter wavelengths.
 Comb-filtering effects between the drivers will also roughen the previously smooth response curve a single driver might have had. It is hard to say how audible this effect will be. You will be the judge.
Scotty

[doug s.]

scotty, you are correct about the interference at higher frequencies; i believe this is what johnr alluded to earlier.  but, you are incorrect about running two drivers in series, then paralleled to a single driver.  the two drivers in series would be a 16 ohm load; run in series w/an 8 ohm load, the total load would now be ~12 ohms.  and, i would not recommend this, even if your amps have enough power at 12 ohms, as the drivers will see different amounts of power...

doug s.

[*Scotty*]

Here are a couple of links to the math required to calculate series and series parallel circuits.
http://www.electronics2000.co.uk/calc/series-parallel-resistor-calculator.php
http://learnabout-electronics.org/resistors_20.php
Basically you can't add 8ohms and 16ohms in parallel and come out with a resistance higher than 8ohms. The 16ohms in parallel will always drag down the resistance of another resistor of a lesser value to a combined value below that of the smaller resistor. Take two resistors of unequal value connected in parallel and your ohm meter and measure the resulting combination.
Scotty

[doug s.]

Here are a couple of links to the math required to calculate series and series parallel circuits.
http://www.electronics2000.co.uk/calc/series-parallel-resistor-calculator.php
http://learnabout-electronics.org/resistors_20.php
Basically you can't add 8ohms and 16ohms in parallel and come out with a resistance higher than 8ohms. The 16ohms in parallel will always drag down the resistance of another resistor of a lesser value to a combined value below that of the smaller resistor. Take two resistors of unequal value connected in parallel and your ohm meter and measure the resulting combination.
Scotty

oops, my bad.  big time brain fart here; your formulas are the same as what i posted.

but, lowering impedance by running drivers in parallel will not give you more db per watt; it will simply give your amp more headroom by allowing it to output more watts.  whether or not your speaker's power handling ability can handle the added output is another issue...

the added spl of running additional drivers is solely due to increase of the spl of each driver.  doubling the drivers gives a theoretical addition of 3db at any given wattage.  if one driver is 88db/1w, then two is 91, four is 94, eight is 97, etc...  headroom is also increased, as you can run eight 30w drivers up to 240w.  but, you are increasing the headroom of the system; each driver still max's out at 30w.


doug s.

[*Scotty*]

When two drivers have the same efficiency with a 1 watt input and one of them has an impedance of half that of the other driver, when they are supplied with the same voltage the lower impedance driver will pull twice the current from the amplifier assuming it behaves as a "perfect" voltage source.
 This means that for any given voltage input it will produce three dB more output.
 When multiple drivers are covering the same frequency range their output does not go up 3dB for every doubling of cone area indefinitely. After you have more than 2 drivers and gain 3dB you may not gain another whole 3dB with each doubling of cone area. Four drivers don't necessarily give you a 6dB increase over the output of a single driver.
 When you have as many as 6 drivers you start seeing near field effects at increasing distances from the array. In other words it stops behaving as a point source and starts to behave like a line source.
Scotty

[doug s.]

When two drivers have the same efficiency with a 1 watt input and one of them has an impedance of half that of the other driver, when they are supplied with the same voltage the lower impedance driver will pull twice the current from the amplifier assuming it behaves as a "perfect" voltage source.
This means that for any given voltage input it will produce three dB more output.

in this case all the drivers are the same.  the amp only "sees" a different load because of how the drivers are wired.  and, the lower impedance the amp sees will cause it to output more watts as the driver is is pulling more current.  it will output more db's if it's making more watts.

When multiple drivers are covering the same frequency range their output does not go up 3dB for every doubling of cone area indefinitely. After you have more than 2 drivers and gain 3dB you may not gain another whole 3dB with each doubling of cone area. Four drivers don't necessarily give you a 6dB increase over the output of a single driver.
 When you have as many as 6 drivers you start seeing near field effects at increasing distances from the array. In other words it stops behaving as a point source and starts to behave like a line source.
Scotty

agreed, which is why i said "theoretical"...

doug s.