Current Driven Loudspeakers and Tranconductance Amplifiers

[Tube_Dude]

Recently i have found a Italian site where they made great claims about transconductance amplifiers (amplifiers with very great output impedances)...they claim that they are better suited for driving dinamic speakers,as they say that they are current operated devices.
Wath do you think of this,guys??...current or voltage???...

Regards

Jorge

PS: the site ishttp://www.ultrasound-hifi.com/

[hagtech]

Sounds like a copy of the Aleph. Current output, voltage feedback.

I've always loved transconductance amps. Good for fast sample-and-hold circuits, gated amps, programmable filters, etc.

jh

[mrfeedback]

See http://www.ultrasound-hifi.com/Us_wh..._frame_en.html for gushing reviews that would put Stereophile/TAS to shame.

Regards, Eric.

[DarkOne]

I think that operating in current domain might be succesfull. The coil in speakers is moved by the current not by voltage so current drive is theoreticaly better (but in real world ???). Anyway, I'm gonna check it out.

[Havoc]

Well, there has been a lot of basic work done on it by MOJ Hawksford. He did theoretical as well as practical work on it, 2 great articles in JAES. The thing is that current drive cancels some of the non-linear terms in the transfert function of dynamic speakers. He measured up to 20dB reduction in distortion of a speaker. Wich is great, concerning that speakers are the weakest link in this respect. On the other hand, JL Hood (I think, it might be D Self) argues that unless everything is extremely well made, a good voltage amp can reach the same levels of total distortion. It can only work with active speakers, no x-overs in the way! Do not forget that you loose all damping of the woofer, the damping factor of these amps makes that of a tube amp look great. Think of output impedances of 20k-100k.

I have build a small 3-way speaker with 10W transconductance amps. No bad, but it were no top speakers either. More a proof of concept. Still have to finish them. Only one is made and has filler applied. It even needs to be sanded.

[GRollins]

Saying that a speaker is a 'current driven' device is semantically null. Without voltage, you will have no current; it takes both. Any equation that shows current as a variable can be restated to show voltage. For instance, the old standby for power I**2R can also be phrased V**2/R--note the lack of any (direct) referent to current. At all times and in every way, Ohm's Law will prevail.
Now, to the extent that an amplifier might have a high output impedance, I wouldn't be surprised to find that such a circuit 'sounds' different. For one thing, the damping factor will be low, by definition. This may work better with some speakers than others. There will very likely be other factors that come into play depending on the circuit topology. All in all, the amplifier could very easily have a distinctive sound.
Whether that sound suits you is another question entirely.

Grey

[capslock]

This thread really belongs into the loudspeaker forum. Will one of the moderators move it and give it a name like "current driven loudspeakers"?


While it is true that a speaker is both current and voltage driven, the terminology is still ok. Voltage driven means that the feedback loop keeps the output voltage proportional to the input signal while current driven means that the current is proportional to the input signal.

Any constant current source will try to keep the current to its ideal value no matter what the load resistance is. By definition, R_out = delta V_out / delta I_out. As the deminator is almost zero, the output impedance approaches infinity.

This does not mean we loose electric damping, it just means that the definition of damping factor is only meaningful for voltage drive. In a voltage drive configuration, the back-EMF generates a voltage that will be shunted to virtual ground by the low R_out of the amp.

In a current driven situation, ideally, the back-EMF causes an error current and the amp corrects for this by outputting an extra current of opposite polarity.

[pinkmouse]

Capslock

It seems a bit half way between camps to me!

Still if it's gone quiet in SS, why not try livening it up in Loudspeakers

[planet10]

It would be nice to light up this thread.

Thorsten mentions current drive at least a couple times in the gainclone thread, and i had a long telephone converstaion with an old RCA engineer who figured that hifi had gone down a blind alley by moving to voltage drive.

He figured the amp should have the same output impedance as the speaker. He was building a small number of bullet-proof PP Class A amps for theatre use (he was in ill health and just disappeared off the map so i was never able to contact him again).

dave

[ThorstenL]

Hi,

Saying that a speaker is a 'current driven' device is semantically null.

True, however what it attempts to state in electrical tems is sound. it implies that the speaker is a current actuated device, meaning that the force being applied to the cone is proportional to the current flowing in the voicecoil and ONLY, STRICTLY and ABSOLUTELY to the current.

Given that Speaker has a voicecoil that is filled with a solid iron plug (causing eddy currents to be induced) and given that the voice coil gets hot and is wound with copper or aluminum (instead of constantan) and thus increases it's resistance it should be PAINFULLY obvious that driving a moving coil speaker from a modulated voltage source (Amplifier with low output impedance) will MAXIMISE the system wide distortion and will cause significant levels of short and long term compression.

Some measures can of course be taken to aleviate the problems in driver design, but in the end there are strict limits as you always trade efficiency for these improvements and thus you loose a significant proportions of the gains when we take a specific given SPL as reference.

Sayonara

PS, driving moving coil speakers designed to require external electrical damping will not work of course, we must in effect design drivers with a Qm in the region of 0.3 to 1 as the only possible damping is mechanical.

A good choice would be a seamless Aluminum coil former, this would both kill some of the eddy current problems in the polepieces and reduce the drivers voicecoil inductance drastically and it would give inherent damping to the driver that would in effect appear as very low Qm. Of course, such a speaker when driven from a normal amp would have absolutely no bass...