Amps - voltage/current discussion

[djanci]

Hi Amp discussion Guys,
I've made a new thread for you.
Thanks
Mark.
-------------------------------------

Originally Posted by planet10

A driver with low mechanical Q becomes a candidtae for current drive

Can you explain further?

Originally Posted by planet10

... and the MA drivers crtainly have no issues with open, clean, and dynamic.

Never put that in question or else I wouldn't be a lucky owner of MA drivers.
But the question is if we have two driver with identical electrical and acoustical but drifferent mechanical parameters, would that with less damping (>Qms) has more clean and open sound?

[planet10]

Originally Posted by djanci

Can you explain further?

Qm is largely a measure of the shape of the impedance peak at resonance. The low frequency response of a loudspeaker driven by a currnt amp highly depndnt on the impedance curve. When impedance rises the power output of the amp increases, so to keep flattish LF response you need a speaker system with a flattish impedance curve (ie low Qm, a system Q of 0.5-1 would be ideal)

But the question is if we have two driver with identical electrical and acoustical but different mechanical parameters, would that with less damping (>Qms) has more clean and open sound?

I don't think it is possible to have the former, and since the Qm is a measure that describes the behaviour of the system at resonance, then i would think any effect on clean & open (which is more a MF + HF thing) is indirect, a side effect with no direct correlation to Qm.

dave

[djanci]

Originally Posted by planet10

Qm is largely a measure of the shape of the impedance peak at resonance. The low frequency response of a loudspeaker driven by a currnt amp highly depndnt on the impedance curve. When impedance rises the power output of the amp increases, so to keep flattish LF response you need a speaker system with a flattish impedance curve (ie low Qm, a system Q of 0.5-1 would be ideal)

Amps are not current source but voltage source, so when impedance rises the power output of the amp decreases which is accordant to resonance theory. But, the truth is that high-end amps with low feedback don't "like" high impedance deviation.

Originally Posted by planet10

...clean & open (which is more a MF + HF thing)...

From article: "more clean and dynamic. And when you look at it, you find it is very simple, because they have less loss. The surround is easier to move, the spider is better constructed, they have better air flow, higher sensitivity.". But also stated: "the old drivers (high Qm) were much quicker. They had some resonances, but you could get rid of that in the crossover.". Some resonances? Where? In MF/HF range? If so that would not be good for FR driver.
It seems that MA's choice of Qm~2 is good balance between two extremes.

[planet10]

Originally Posted by djanci

Amps are not current source but voltage source

Most amps are close to voltage sources, but there are solid arguments, that given that a speaker is a current device, they are best driven by current amplifiers. Typical SE amps are often very close to sitting on the border between current sources & voltage sources -- they tend to like flat impedance.

It is vry educational to take a number of spakers & a variable transimpedance amplifier (one that has a knob on it that changes the amp continuaously from voltage amp to current amp. Each speaker tends to have a spot whr it sounds best.

From article: "more clean and dynamic. And when you look at it, you find it is very simple, because they have less loss. The surround is easier to move, the spider is better constructed, they have better air flow, higher sensitivity."

Higher Qms also means that you have a driver that rings more at resonance.

Except for the better sensitivity bit, you have just described an Alpair driver.

What works best is a very complex dance of trade-offs and cannot be considered but as part of a system with the amplifier (and the cable that connects the amp to the speaker).

dave

[djanci]

Originally Posted by planet10

...Typical SE amps are often very close to sitting on the border between current sources & voltage sources -- they tend to like flat impedance....Each speaker tends to have a spot whr it sounds best.

Ufff, all amps are designed to be constant voltage source and likewise all drivers are designed to give flat(desired) freq response when applied constant voltage. There are only "less ideal voltage source" amps. If you would have ideal constant current source connected to speaker there will be deviation in potential difference on speaker in freq domain. Deviation will be proportional to impedance curve so there will be oscillations in freq response - SPL (power dissipation) gain at speaker resonance(s) and toward HF - some kind of equalizer - unless you have speaker with flat impedance curve. Unpredictable if you don't know impedance curve and some of the best loudspeakers could sound terrible with such amp.

[planet10]

Originally Posted by djanci

Ufff, all amps are designed to be constant voltage source and likewise all drivers are designed to give flat(desired) freq response when applied constant voltage.

That is not true at all. First Watt F1 & F2 are 2 examples of SS current amps. We have 2 Transamps here, designed first & foremost as current amps. Most SE tube amps are not voltage amps.

I am working on some current drive speakers so that i can take full advantage of current amps. Joe Rasmussen has some done.

dave

[buzzforb]

Originally Posted by planet10

That is not true at all. First Watt F1 & F2 are 2 examples of SS current amps. We have 2 Transamps here, designed first & foremost as current amps. Most SE tube amps are not voltage amps.

I am working on some current drive speakers so that i can take full advantage of current amps. Joe Rasmussen has some done.

dave

I hope to build an F2 that i have a boards for. I will let you now how it stands against the Aleph J, driving the 10.2's. surprising how little power you need to make these babies sound goooood! Everything that i have read suggest that current driven amps are superior in terms of their control over the driver, which in theory, would give a better more accurate response. Just limited in power. I am no expert though, just going on what Mr Pass has said along with some articles from some other semi-smart people.

Side note; Mark, did you haev to come out with a new driver I already have an itch for the A7.

[djanci]

Interesting. Simple question: What is the output voltage of such amp with applied input signal but without speaker connected?

[pinobot]

Originally Posted by djanci

Can you explain further?
But the question is if we have two driver with identical electrical and acoustical but drifferent mechanical parameters, would that with less damping (>Qms) has more clean and open sound?

It's subjective. The same why some people prefer a Labyrinth or back loaded horn over an acoustic suspension for bass.

[planet10]

Originally Posted by buzzforb

Everything that i have read suggest that current driven amps are superior in terms of their control over the driver, which in theory, would give a better more accurate response.

I interpret what i read a bit differently. Control over the loudspeaker -- as in amp has an iron grip on the speaker -- is actually non-existant. Hence the need for a low system Q (impedance is flat at bass resonance -- or complementary to the bass roll0off). What a current amp does, is since the speaker isn't being asked to be a (poor) voltage-to-current converter, you loose the back EMF that the speaker feeds back to the amplifier. If we ignore that some amps actually become unstable from this feedbackm I see this back EMF as an overlaid signal component that cause an "infinite" regress time-smeared fog that decreases downward dynamic range.

If you haven't gotten ETM's book on the subject well worth a read... some tips in there on making spakers (typically dsigned for voltage drive) for current drive. Nelson's article also good. Have a look at Joe Rasmussen's latest speaker too. He uses an aperiodic TL to control the drivers resonant peak. The lower a driver's Qm the easier the task is.

With your Pensils you may find the need for more damping, maybe even worth getting a bale of Ultratouch, a very good damping material for this kind of thing. If you have the ability to measure impedance (most computer based measure systems), you ar looking to flatten the impedance curves, A rising impedance at the top end may have to be dealt with as well.

dave