If you can believe three versions can be made in theory with all else the same, then this is a data point.
Those specs don't come from the engineers, they come from marketing. Or engineers working as marketing prostitutes (I was once guilty 😱). There is a tendency to put out simplified numbers to avoid questions from the ignorant, kinda like how Honda implemented massive high frequency blend on their radios-it sounded bleh, but avoided any complaining phone calls.
It's been a while since I was building woofers, but my recollection is that no you can't really make different impedances have exactly the same parameters. First of all the coil has to be different, so the mass and the BL is different. Then offhand it seems you could modify the gap to lower BL and increase the Q to match the higher impedance version however again my recollection is it's just not that simple. I'll ask some friends (one of whom works at JBL in fact).
One friend (not the JBL one) says that "It's true to some degree, and they have a patent on it. Basically it uses bifilar, series/parallel configs to make the parameters match. But seems some turns are mostly along for the ride so it's not as efficient." I'd add that I highly doubt the parameters match EXACTLY and that anyway parameters are not so exact in the first place-the numbers you get depends on the measurement technique, the drive level, and vary sample-to-sample.
I think that my earlier point about the angular flow of electrons, by whatever coil is a correlate of the motor factor, is valid. So how that is arrived at is irrelevant; many thin wires or few thick ones, for a given geometry and gap flux density that is.
Where it is relevant is a secondary effect where a higher current produces a greater induced current in the motor structure and the likelihood of greater non-linearities as a result.
This is actually an interesting point because short coil/long gap drivers which are more linear than long coil/short gap drivers in terms of their displacement-dependent inductive non-linearities will likely perform worse in respect of these current-dependent inductive non-linearities. Essentially the geometry of the induced current flow (determined by the skin effect) permeates deeper into the motor in the former class of drivers - and changes with displacement but is relatively constant in the latter. But heading off-topic once again...
Again, for example a lower impedance version of a driver usually have lower Qes-Qts, if you replace a 8 Ohm version from a Qtc 0.7 aligned box, the 4 Ohm version of the driver produces lower Qtc in that box, which can be subjectively better, because the higher group delay of the higher Qtc is a distortion too, just in the time domain.
Anyway, we don't know how many samples or what type of box the OP has heard with the same driver just different impedance versions. All this can be just subjective preference too, which can be based on from small volume differences to placebo.
Last edited:
But once again, this is a different topic because you are changing a different variable. You would assume in a valid comparison issues pertaining to these linear issues would have been taken care of. The audibility of resonances is completely different to the audibility of current-dependent driver non-linearities.
There are drivers with 2 voice coils. Monacor make some (eg sph 135 tc).
Are there any differences if they are run with one coil-8ohm, both in parallel-4ohm, or both in series-16ohm?
Are there any differences if they are run with one coil-8ohm, both in parallel-4ohm, or both in series-16ohm?
Speaking about the voice coil impedance only is a tricky game, a little to hard for me, i can't see anyting at this level of simplification.
Running only one 8-ohm coil will make Qts higher and the maximum input power is half - not good. Use both coils in parallel for total of 4 ohms - maximum SPL and minimum Qts.
Perhaps I made my point poorly.
The circumferential flow of electrons, ie current in the VC, is unaffected by whether or not it occurs in many fine conductors or fewer larger ones; there is no implicit reason on this factor alone for a 4 or 8 to be different in performance, but other incurred factors may do so.
Of course attempts to increase the occupying volume of the VC have been made, with the use of square section wire, and even hexagonal I believe.
I didn't know that there were disadvantages to the short coil long gap approach.
Last edited:
That is precisely the point I have been trying to make, that the coil resistance is an irrelevance. It is the power transfer that is the critical issue, and the showing of Rc in the normally presented expression for driver sensitivity is misleading.
I am not aware of the explanation given in my previous post as having been published anywhere before - I would have referenced it for you if so 😉 Nevertheless it is measurable and has been the subject of considerable efforts to ameliorate what manifests itself as a particularly harsh and intrusive audible midrange distortion, either by adopting highly conductive or highly insulating pole pieces - two different approaches that are literally poles apart 🙂
The link between damping factor and current delivery is often confused. Additionally the fact that lower impedance wants more current is often misinterpreted into the mix.
Of course you can build an amp for 4 ohms with the same philosophy and methods as for 4kohms and get a comparable result.
In theory, anyway. In a similar vein, I have also argued that you can cover the same amount of dynamic range either on a +5V single supply, +/-15 V or a 300 V B+ tube circuit.
In practice, 100 W in 4 kOhms is 632 Vrms, or almost 1800 Vpp. More than a bit dangerous, for one.
Not even high-power vacuum tubes are that high in impedance - by the time their supplies are in the kVs, we tend to be talking several hundred watts or even kilowatts of output power, e.g. ham radio linear PAs.
Likewise, available analog dynamic range tends to shrink more and more below 5 V supplies (try finding some ultra low noise low-voltage opamps - good luck), and the number of parts suitable for supplies well in the kOhms is quite low.
Practical parts availability and economics tend to guide our design decisions in this regard. This is such a critical factor that we have been struggling to operate landers on the surface of Venus (where it's a balmy 450°C) for longer than a very limited amount of time - we just don't have very much in terms of electronics or batteries that would be suitable for these temperatures. Basically you have to start almost from scratch and duplicate years, probably decades worth of R&D.
And so it probably is with speakers. In theory, you're only changing the primary winding on a transformer. In practice, it isn't just that.
You can get some headphones with drivers ranging from 32 to 600 ohms. Their performance tends to be similar but not identical - the higher-impedance ones possibly have lighter voice coils, as the 32 ohm versions seem to be a bit shorter on treble extension.