Current source speaker amps

[fonograph]

Why dont we use current source speaker amplifiers?

Speaker coil drivers work with magnetic force that depends on amount of current,why are then all amplifiers voltage sources with low output impedance and voltage feedback?

Wouldnt it be better if amps were current sources,that is output impedance as high as possible and use current feedback  so the current waveform inside the coil is as close as possible to the music waveform?

My thinking is that voltage is less important than current  for speaker driver,since the acceleration force on the diapragm depends on current,not voltage,then I think current source style amp with current feedback would provide less  harmonic distortion and less time domain ringing.

I know there is issue of driver damping due to nonlinear impedance of the driver,but there are planar magnetic drivers that have impedance that is almost completly flat.Tell me what you think,am I missing some key piece of knowledge?

[cs.dk]

since the acceleration force on the diapragm depends on current,not voltage

I think the word here is power J or W. If you raise either of them the power will raise.

[fonograph]

Well,watt is just joule a second.I see you like to nitpick about my choice of word force,I dont see how this is beneficial to this thread in any way but anyway,my choice of word "force" was not wrong since its the magnetic force that moves the driver mass.

[cs.dk]

I know what a watt are. No nitpicking though.
I see I should be more precise.. I meant power drives the coil. Raise either the voltage or amps will give more power.

[Kleinstein]

The modern speakers are made to be driven by a voltage. In the old tube radio days there where speakers made to be driven by a current.
If driven by a current one gets more low frequency resonance of the speaker and thus a less predictable frequency response. The low frequency resonance gave an slight extra boost to the low frequencies, that was kind of welcome for the lower power tube amps, despite of not giving an accurate reproduction of the music, unless the amplifier is matched to the speaker.

To a large part the force is needed to move the speaker mass, not so much in moving the air.
The voltage also include the velocity of the coil moving and this is a little closer to the sound emissions.

With normal speaker systems there is also the cross over network, that is made to be voltage driven. So this would be the first part to be changed.

The lack of damping to the mechanical resonance of the speaker cause way more ringing with current drive. The better way here would be to have the cross over network before separate amplifiers, to get an even lower impedance drive to the speakers.

Harmonic distortion is tricky in speakers. AFAIK there are two main contributions: one is an possibly inhomogeneous magnetic field seen by the coil. This would effect current and voltage drive in a different way.  A smaller field would cause less magnetic force and this way effect the current drive. In voltage drive a lower field gives less induction voltage to counteract the drive and this could allow a higher velocity.

However voltage drive is not perfect as there is still the ohmic resistance and thus more in between ideal velocity control and force control. Its more like force control in the very low frequency range and more like velocity control in the high frequency range.
Due to the opposing effects chances are that voltage drive would be more linear, especially in the low frequency range, where large excursions occur.

The second nonlinear effect is from filed weakening: the current flow in the coil causes a small change in the field of the permanent magnets. With modern magnets this is usually a small effect. It was possibly more significant with the old AlNiCo magents. In some cases there are extra turns around the magnet to compensate for this - though this was more a thing if the past. Anyway this should effect current and voltage drive about the same, maybe again slightly less for the in between case.

[spec]

Hi fonograph

Kleinstein has the answer.

Thought you may like to know that, with similar thoughts to you, I once made a current audio power amplifier to see what it would sound like driving a speaker.

With small LowFi speakers it made little difference: they sounded as bad as ever. But on decent speakers the sound seemed to vary according to the type of enclosure.

With an infinite baffle the sound was bass heavy but not unpleasant and the top seemed to be unchanged.

Transmission line cabinets were more base heavy and base reflex speakers were even more base heavy.

The general sound, apart from the base enhancement, was a clean effortless sound that was somehow unnatural.

Hope that makes a bit of sense.

At one time I think Phillips produced some speakers that had built-in current amps, but they were part of a servo loop that used a feedback transducer attached to the speaker cone. The servo speakers didn't catch on though.

[Audioguru]

The resonance of an undamped speaker does not simply increase low frequencies, instead it increases ONE frequency (or only a few) like a bongo drum.
I think a planar magnetic driver is normally coupled to a large panel that also has a bad resonance. The very low output impedance of a voltage amplifier cannot properly damp the panel resonance.

[boB]

Mackie HR-824 powered monitors work this way for the woofer rather than using a feedback coil.

It works very well, too.

boB

[bson]

Why dont we use current source speaker amplifiers?

Some do.  Current-feedback amplifiers are becoming more common; I have a Marantz MM7025 in my lab (https://www.us.marantz.com/us/products/pages/ProductDetails.aspx?CatId=avseparates&ProductId=MM7025).  There's a book, Current-Driving of Loudspeakers by E. Meriläinen which covers it in detail, although IMO with some serious misunderstandings (especially regarding the presence of back-EMF), but the basic concept makes perfect sense.  I looked around a bit and can't recall where I bought my copy, but it's an interesting read (maybe not so exciting nowadays since CFB topologies are more commonplace).

Edit: oh never mind, Amazon has it.  Not sure how I failed to find it.  https://www.current-drive.info/

[SiliconWizard]

Audio current amplifiers do exist. As I recall, they are more often used to drive electrostatic loudspeakers.

I remember the "The current amp" project in the Elektor magazine in the 90's, it was designed to drive "ribbon speakers".