Dual Voice Coil Speaker For Crate CA112D

[Mike Sulzer]

No mystery with the impedance, it's either 4 or 16 ohms, it's done every day with car subwoofers.
Speaker will be somewhat dull but usable.

As of using it as an attenuator, both coils are exact same number of turns and wound on the same former, so sound output should be 0 .

Why dull? All I can think of is that at the higher frequencies where the impedance becomes inductive the coupling between the two coils reduces the current.

[J M Fahey]

Oh, it's not an electrical but a *mechanical* problem.

Twice the mass means half the acceleration, just that.

On one side, you have double the wire pushing but on the other side besides having double the mass, you also have weaker magnetic flux, and loss is proportional to the square of gap thickness, so it quickly overrides the original gain.

In practice, you gain some (40% in theory, some 20% in practice) at lowest frequencies, while losing at higher ones, that's why this is only popular in subwoofers.

FWIW I actually made some 4 layer coil speakers (same thing, only it's a single continuous wire) and the somewhat increased Bass was not compensated by loss of punch.

FWIW2: after my homewound coil experiments, I asked my then coil winder for a batch of 4 layer coils, I wanted to make an 8" speaker for a small Bass practice amp, and he answered: "what for? , you want to fill the gap with copper?, I can wind for you edgewound single layer copper ribbon coils, same or better results and simpler (for him) winding"



I still have a ton of them, unused.

[Mike Sulzer]

My naive idea is that if you have a single coil speaker that puts out x watts with y grams of copper in the voice coil, the equivalent dual voice coil speaker uses y/2 grams in each coil.

Oh, it's not an electrical but a *mechanical* problem.

Twice the mass means half the acceleration, just that.

On one side, you have double the wire pushing but on the other side besides having double the mass, you also have weaker magnetic flux, and loss is proportional to the square of gap thickness, so it quickly overrides the original gain.

In practice, you gain some (40% in theory, some 20% in practice) at lowest frequencies, while losing at higher ones, that's why this is only popular in subwoofers.

FWIW I actually made some 4 layer coil speakers (same thing, only it's a single continuous wire) and the somewhat increased Bass was not compensated by loss of punch.

FWIW2: after my homewound coil experiments, I asked my then coil winder for a batch of 4 layer coils, I wanted to make an 8" speaker for a small Bass practice amp, and he answered: "what for? , you want to fill the gap with copper?, I can wind for you edgewound single layer copper ribbon coils, same or better results and simpler (for him) winding"



I still have a ton of them, unused.

[J M Fahey]

Sorry but it does not work that way

Watts are irrelevant here, and there's no "watts per gram" relation whatsoever.

Let's put some numbers into it:

I have no data on that speaker but a popular Eminence voice coil size is 1.5"or 38mm (nominal)

A common Eminence plate thickness is 8mm , really 7.92mm which comes from some wonky "Imperial but liberated in 1776" measure, based on the King's body parts dimensions so all were *slightly* altered just to show AngloAmericans really "meant it" so from now on I'll stick to Metric.

A normal voice coil will have 2 layers of enamelled round copper wire.

I choose 0.18mm diameter (why?because I design and wind custom voice coils all the time and already know it, but let's do the Math in detail) ... a.k.a. AWG33

Being an efficient Guitar speaker I choose winding length same as gap height , so also 8mm.

Average turn length: Pi*38mm=0.12m

Nominal 0.18mm wire will actually use 0.19mm because of enamel thickness.

N* of turns: (8/0.19)*2 (layers)=84

Wire length: 84*0.12=10.1m

Wire tables give me "2450 meters per 500g" , probably because nearby 1lb spool is a popular commercial size , so weight is:
(10.1/2450)*500g=2g

Is that too much or too little?

A 12" cone is a piston which moves around 2 grams of air, so to move a 2 grams "passenger" (useful load) we must also move another 2 gram "car frame" , a useless load which requires power to be moved back and forth but can't be heard.

Now consider we add another voice coil tightly wound around the original one, so wasted weight is now 4 grams ... mass is larger but motor is weaker (wider gap means lower magnetic flux) : sluggish response: dull sound.

Copper weight is such a problem at high frequencies thatb tweeters often use one of these solutions ... which are also a compromise:

* aluminum wire coil.
Aluminum is way less dense: 2.8g/cm^3 vs 8.9g/cm^3 , a 3:1 advantage, but also has 1.6X higher resistivity, so you must use thicker wire which negates some of the advantage and to boot again requires a wider lossier gap.

* skip one layer ....many tweeter coilss are single layer which to boot means thinner wire to lower "deadweight".

* shorter than the gap (underhung) voice coils.
Waste of magnetic energy but again a weight saving.

So in a nutshell speaker design is NOT Black Magic, but there are so many unavoidable compromises involved, none good, that thare are many (poor) ways to solve the same basic problems, all different sounding.

That's the reason why speakers are horrible 1% efficient or less , with very few exceptions (JBL/EV/etc.) which are very expensive and even so barely reach 2% or 2.5% (not a typo).

[Mike Sulzer]

Thanks Juan Manuel, that was very interesting, especially the part about the tweeters and the aluminum wire. I had not thought about that. But in the particular case of an auto sub woofer, consider for example a common configuration, a so called DVC 4 with two four ohm voice coils. Series gives you 8 ohms, parallel gives 2 ohms. If each of the 4 ohm windings is one of the two usual layers, then the eight ohm single voice coil equivalent would just be a single two layer coil where each layer is identical to one of the two windings in the dual coil speaker. The dual and single coil versions would have nearly identical characteristics other than the additional impedance option with the dual coil.