The "damping factor" works this way:
When a woofer with heavy diaphragm returns to its neutral mechanical position after a transient, it has a lot of inertia and therefor a tendency to "overshoot" the neutral position. But a coil moving in a magnetic field, produces a current, and if the coil is short circuited (or almost) this current is high, and thereby producing a magnetic field , that has the opposite direction of the permanent field in the woofer, and this will slow down the diaphragm.
In the case of the arrays, this damping factor will always be low, except in the situation where all the drivers are parallel and the amplifier driving them has extremely low internal impedance.
Theoretical there is a little advantage in first paralleling 5 drivers as each driver will see a impedance of 8/5x3 = 13,3 Ohms, and in the series first it will see an impedance of 8x4 = 32 Ohms. As can be seen, almost no damping factor effect in any of the cases, as the driving impedance has to be low in comparison to the nominal impedance of the driver, before the damping factor has any significant effect.
I don't think the Vifa drivers are affected much by a high damping factor anyway, so , if there is a difference, it is probably more the interacting between the drivers and this is far too complex for me to say anything meaningful about.
There was also once presented a circuit to at least partially overcome this, if I remember correctly
Hi koldby, i think you have it backwards, about reducing damping factor when you series drivers.
Check out the numerator in the damping factor formula The Relationship Between Amplifier Damping Factor, Impedance & Cable - ProSoundWeb
Personally, my biggest understanding about damping factor, is that we can safely ignore it for anything other than low impedance subs with long speaker wire runs.
I do not think there is any circuit that can overcome the loss of dampingfactor when series connecting woofers.
Maybe you are referring to Poul Ladegårds (I think it was him) article on a current output amplifier?
This partly overcomes the rising impedance @ the resonant frequency
I cannot see i got it bacwards. An amplifier with very low internal impedance has a high dampingfactor, as the woofer sees into almost a short circuit.
But when you series connect two woofers each of them see into the intenal impedance of the amplifier IN SERIES with the other woofers impedance. Now it is not a short circuit at all..
What I think you are referring to is that two woofers in series has the double impedance. And you are correct that a SINGLE woofer with the double impedance would result in a higher( x2) damping factor. But this is not the case here...
I cannot see i got it bacwards. An amplifier with very low internal impedance has a high dampingfactor, as the woofer sees into almost a short circuit.
But when you series connect two woofers each of them see into the intenal impedance of the amplifier IN SERIES with the other woofers impedance. Now it is not a short circuit at all..
What I think you are referring to is that two woofers in series has the double impedance. And you are correct that a SINGLE woofer with the double impedance would result in a higher( x2) damping factor. But this is not the case here...
I think I see what you are saying....looking at it from the point of view from the driver, not the amplifier. I'll need to think about that, as I'm not sure if seeing a short circuit is that important, compared to the ongoing motor control of the signal. thx
This is not so. As I said woofers with heavy diaphragms often need an amplifier with low internal impedance (high dampingfactor) to control its behavior in the bass.
Another thing is , that long speaker cable runs can RUIN the damping factor as the cable impedance is (again ) in SERIES with the woofer. So if the cable impedance is 1 Ohm it completely ruins the damping factor in an amp with 0,008 Ohm, just as a SERIES connect additional woofer would.
I think I see what you are saying....looking at it from the point of view from the driver, not the amplifier. I'll need to think about that, as I'm not sure if seeing a short circuit is that important, compared to the ongoing motor control of the signal. thx
Excatly
If undamped, the cone keeps moving until the energy stored in the voice and the momentum of the cone dissipates.
Each voice coil is loaded predominantly by the other voice coils in series with it instead of being short circuited by the very low output impedance of the amplifier (very low if its not a tube amp). If there is a saving grace, the low mass cone and low L voice coil don't store much energy.
Suppose I rewire - 4 sets of 8 drivers in parallel series connected for net Z of 4 ohms. Then I'll have 4 groups in series with the amp's output Z. The situation will be better but nowhere near the damping factor achievable with single driver per amp. Only way to beat that seems to be multi-amping.
Each voice coil is loaded predominantly by the other voice coils in series with it instead of being short circuited by the very low output impedance of the amplifier (very low if its not a tube amp). If there is a saving grace, the low mass cone and low L voice coil don't store much energy.
Suppose I rewire - 4 sets of 8 drivers in parallel series connected for net Z of 4 ohms. Then I'll have 4 groups in series with the amp's output Z. The situation will be better but nowhere near the damping factor achievable with single driver per amp. Only way to beat that seems to be multi-amping.
Yes each group will then see the three other groups in series and as each group is 1 Ohm it will result in a 3 Ohm impedance. This could be a better way, depending on how the amp is behaving into 4 Ohm.
Multiamping is a good solution. Maybe only six amps with good behavior into low impedance is needed.
Hi koldby,
OK, a little quick research has me doubting again, whether series presents a damping problem....
Pls check out this thread, particularly post#8...Series vs. parallel sound quality
What do you think?
thx for getting my mellon in gear this morning...coffee wasn't working
OK, a little quick research has me doubting again, whether series presents a damping problem....
Pls check out this thread, particularly post#8...Series vs. parallel sound quality
What do you think?
thx for getting my mellon in gear this morning...coffee wasn't working
Hmm I am an oldschool guy , where the explanation I gave was the one everyone accepted, but I can see that there is more to it than just impedance's in series. So it seems like I stand corrected here after all , especially if the woofers and enclosure are exactly the same.Hi koldby,
OK, a little quick research has me doubting again, whether series presents a damping problem....
Pls check out this thread, particularly post#8...Series vs. parallel sound quality
What do you think?
thx for getting my mellon in gear this morning...coffee wasn't working
Yes, the thread Mark Linked is instructive and makes me feel chagrined that I didn't think of it myself. So damping and Q are directly/inversely related. If the series parallel array had different damping than the parallel series array it would show up in its QTC (the system Q). And QTC can be see in the shape of the low end roll off. If high Q, than one would see peaking just before roll off.
Check post #18 in my line array thread
Full range line array for wall or corner placement
There is some peaking there. Not sure if its due to boundary support or indicative of Q. Suspect the former as the teal trace also has similar bump and its the single driver test box
Check post #18 in my line array thread
Full range line array for wall or corner placement
There is some peaking there. Not sure if its due to boundary support or indicative of Q. Suspect the former as the teal trace also has similar bump and its the single driver test box
I looked to HornResp for an answer. Top sim compares a single SB65 driver to same driver and box with 4 ohms in series with driver. The series R definitely causes peaking and thus raises the QTC.
The lower sim compares 8 SB65s in series with 8 in parallel, with source voltage adjusted to nearly match the levels. THe curvers are identical except for the level offset. There is no apparent change in Qtc due to multiple drivers in series.
The lower sim compares 8 SB65s in series with 8 in parallel, with source voltage adjusted to nearly match the levels. THe curvers are identical except for the level offset. There is no apparent change in Qtc due to multiple drivers in series.
Attachments
Not to beat a dead horse, especcially not one I helped kill, but there is one last thought I wanted to get out. Too many interruptions at my end earlier.
The reason why there is no change in damping for the series case is that back EMF for the drivers is in phase and so adds. With N drivers you have N times as much voltage driving unitary current through N times as much impedance. Same current, same force exerted on the cones, same damping.
The reason why there is no change in damping for the series case is that back EMF for the drivers is in phase and so adds. With N drivers you have N times as much voltage driving unitary current through N times as much impedance. Same current, same force exerted on the cones, same damping.
Not to be nitpicking and still remembering I was the first to make a fundamental flaw, it is not a back emf voltage that produces the current. The movement of the coil in a magnetic field produces a current, and as this is flowing through an impedance (voice coil) there is produced a voltage...
I am not an expert here at all, but I believe , and have heard in several loudspeaker construction, that it is very beneficial to use an impedance correction network. One of the positive effects is that the amplifier has a much easier job (no big phase changes) around the resonance frequency and the rising impedance caused by the voice coil induction can also be a problem for amplifiers. Flat impedance is much easier to deal with for an amplifier. IMHO