dual voice coils & impedance

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Say we have a 1:1 transformer with two identical primaries and one secondary. The secondary has a load of 8 ohms.

If we drive one primary, the load presented to the drive signal is 8 ohms. If we then parallel the second primary winding with the first, the impedance to the drive signal is still 8 ohms because the turns ratio is unchanged; only the primary winding is effectively twice as thick.

Now let's say we have a speaker with a dual voice coil. If we drive one coil the impedance is nominally 8 ohms. If we parallel the second coil, many people would say the impedance is now 4 ohms. I say it would still be 8 ohms, only the winding is effectively thicker, like the transformer in the above example.

What do you think???
 
Circlotron said:
Say we have a 1:1 transformer with two identical primaries and one secondary. The secondary has a load of 8 ohms.

Not necessarily, it depends on the transformer as to the impedance.

Circlotron said:
If we drive one primary, the load presented to the drive signal is 8 ohms. If we then parallel the second primary winding with the first, the impedance to the drive signal is still 8 ohms because the turns ratio is unchanged; only the primary winding is effectively twice as thick.

Now let's say we have a speaker with a dual voice coil. If we drive one coil the impedance is nominally 8 ohms. If we parallel the second coil, many people would say the impedance is now 4 ohms. I say it would still be 8 ohms, only the winding is effectively thicker, like the transformer in the above example.

What do you think???

No, and proven by the millions of people out there with dual coil parallel drivers and their ohm meters.
 
Circlotron said:
Say we have a 1:1 transformer with two identical primaries and one secondary. The secondary has a load of 8 ohms.

If we drive one primary, the load presented to the drive signal is 8 ohms. If we then parallel the second primary winding with the first, the impedance to the drive signal is still 8 ohms because the turns ratio is unchanged; only the primary winding is effectively twice as thick.

Now let's say we have a speaker with a dual voice coil. If we drive one coil the impedance is nominally 8 ohms. If we parallel the second coil, many people would say the impedance is now 4 ohms. I say it would still be 8 ohms, only the winding is effectively thicker, like the transformer in the above example.

What do you think???

No.

The main reason for this is that a loudspeaker's nominal impedance is dominated by Re, ie the resistance of the coil.

The impedance on the input of the transformer that you describe is the impedance of the load "seen trough" the transformer, and this is different from the loudspeaker case. The reasoning also neglects any resistance in the transformer coils.

There is a slight similarity though, in that the part of the loudspeaker impedance that comes from the mechanical side (that forms a resonace around fs) does not scale to half the impedance as Re does. Likewise, Le does not either.

But Re scales from 8 (or slightly lower) to 4 ohms (or slightly lower) if the coils are connected in parallel.
 
If our transformer had more than it's fair share of leakage reactance (lousy coupling), let's imagine 8 ohms on the secondary appears as 20 ohms on the primary. Add the second primary and the coupling improves so the impedance reflected to the primary would be lower, closer to 8 ohms the turns ratio figure would dictate.

I can imagine a second voice coil would improve the total "grip" on the magnet (lower Qes) therefore lowering the impedance somewhat. But Re would seem to be the dominant factor.
 
Circlotron said:
If our transformer had more than it's fair share of leakage reactance (lousy coupling), let's imagine 8 ohms on the secondary appears as 20 ohms on the primary. Add the second primary and the coupling improves so the impedance reflected to the primary would be lower, closer to 8 ohms the turns ratio figure would dictate.

I can imagine a second voice coil would improve the total "grip" on the magnet (lower Qes) therefore lowering the impedance somewhat. But Re would seem to be the dominant factor.

Hi,

Your analogy seems slightly wrong because the "floating primary" is in fact a floating secondary coil before it's plugged in.

However, I see what you're saying and the inductive reactance should be fairly stable as long as the windings are closely coupled and wound in the same direction.

I've never seen much point in dual voice coil speakers though. The only advantage that I can see is convenience when you need an unusual impedance. For example, one thing I'd look at is designing a low current amplifier for a high resistance speaker, enabling the use of smaller and faster MOSFETs with lower distortion. But in general it seems like a waste of precious magnetic gap real-estate, especially if only half of the coil is being used.
 
Cal Weldon said:


Car audio. Two channels of amp and room for one woofer.

Sure, but even then it's just a stop-gap solution. Attempting convert stereo into mono by brute force mechanical means is just silly. Likewise, the trend to run DVC speakers in parallel with amplifiers rated for 2ohms and using ultra thick wiring also sounds silly.

AFAIK the original reason for using low impedance speakers in cars was because of the limited supply voltage (typically 12V), before the days of PSU converters. Nowadays, where practically ALL subwoofer amps have built-in boost converters of one sort or another, I see no good reason why the trend should be towards higher currents rather than higher voltages when people want more power.
 
Hi CeramicMan,

Cal did not say mono to stereo at a DVC voice coil.

Car head units put out a mono sub bass which is fed to two channels of a suitable amplifier.

Sounds like you have not made use of some of the excellent DVC drivers there are.

Stop gap - no !
There are dual 2 and 1 ohm drivers too.
This keeps amplifier output driver operation better within SOAR than raising internal PSU rail voltges to achieve the same power, thus a better use of technology with recent designs much more reliable.

Typical sup/plate amps for non-vehicular use are different beasties.


Cheers ......... Graham.
 
Apart from the convenience factor of being able to switch the speaker between various different systems, what possible benefit is there from splitting a mono signal into two amplifiers and then recombining it? It seems pointless.

Like I said before, the magnetic gap in a speaker is expensive real-estate. Why waste it on unnecessary insulation?
 
let's suppose you can humour me.
I have two amplifiers that each produce 100W into 4ohms (actually a redundant stereo amplifier).
I have a sub-bass speaker that has a driver with dual 4+4ohm voice coils.

I can put the dual voice coils in series and get 50W into them from my existing amps.

or

I can put the dual voice coils in parallel and blow my amp up attempting to drive 200W into 2ohms.

or

I can attach one amp to one voice coil and a second amp the the other voice coil,

The resulting power in the third arrangement is 200W into the speaker. I have gained 3db and I was able to use existing equipment.
It's the flexibility that dual voice coils offers that is attractive to users.
When both voice coils are used and fed with identical signals then there is no waste of voice coil gap. Dual layer or even 4layer voice coils work rather well in comparison to single layer voice coils.
 
AndrewT said:
When both voice coils are used and fed with identical signals then there is no waste of voice coil gap. Dual layer or even 4layer voice coils work rather well in comparison to single layer voice coils.

If two thin ribbon wires are stacked together (where the edge width is negligible), they effectively need twice as much insulation compared to a thicker single wire.

And even if the shape of the wiring is optimized so that it's hexagonal or square, a series configuration will merely resemble a single coil with a centre tap. And a parallel configuration will still have wires connected in parallel for no good reason.

If you can get two 4 ohm amps, you can just as equally get one 2ohm amp and save components on the input stages.
 
Hi CeramicMan.

You suggest >> If you can get two 4 ohm amps, you can just as equally get one 2ohm amp and save components on the input stages.<<

Yes,
BUT -
If the circuits are the same except for output transistors the 2x 4 ohm amplifiers will still exert better cone control than a 1x 2 ohm chassis.

And please do not argue against this without first checking out in a real-world situation with a 2ohm amplifier driving a 2x 4ohm DVC wired in parallel.
I have !
The difference in driver cone control between a 1x 2ohm, parallel 4ohm DVC, and a 2x separately driven 4ohm DVC is clearly audible, and bridged amplifier connection to voice coils in series (8ohms) sounds even worse.


Hi Sreten,

I do not like the idea of series resistors in LS leads where it can be avoided, or unless they are fitted for a purpose.
I know that Quad recommended this decades ago for their Current Dumper, but DVCs are normally used for sub-bass drivers in highly resistive or highly reactive cabinets where thick wire connection is essential to maximise damping and minimise losses for techno music.


Cheers .......... Graham.
 
Hi Graham,

I still don't see why 2 amplifiers playing the same signal should be inherently superior to a single amplifier. If the outputs are accurate, the only voltage difference between the channels should be caused by noise and possibly a minute DC offset.

To say that the differences are clearly audible in a filtered subwoofer application suggests that there's something seriously wrong with the 2ohm amplifier that you heard. If anything, it's likely that the amp was distorting, as I really doubt that any subtle difference in damping factor could be easily audible.

Cheers,
 
Hi Ceramicman,

Two separate parallel circuits are better than one, like a V-twin being better than a single pot. Are you going to try and persuade Harley-Davidson owners that their tourer engines cannot be any more low down competent than a single ?

You need to think about more than just the amplifier circuit alone !!!

It is the entire amp-LS driver-enclosure interface, and the way the cone-enclosure reactions are controlled by the way the DVC is connected. It is best when the *dynamically energised* system impedance reactions are shared by two NFB amplifiers, connected one per voice coil.
(Hope you are not thinking in test bench sine terms here, because no difference would be noticeable then !)

Do please try this for yourself before implying that others do not know what they write about, or you suggest incompetence through incorrect conclusions based upon flawed testing.

Actually my tests were with 1ohm capable amplifiers asked to perform no worse than 2ohm duties, and still the bass was better with the 2x 4ohm DVC connection (of 1ohm capable amplifiers) due to the two parallel connected voice coil-NFB loops sharing control of the LS system generated back-EMF !!!


Cheers .......... Graham.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.