Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Bob Lee (QSC)]

Ted Olausson wrote on Thu, 14 February 2008 07:34

BTW how can an amp that is stated to have 1000w/8ohm and 1100w/4ohm at 0% distortion hade a DF above 1000!? It should be 1000/2000/4000 watt at 8/4/2 ohm...

???

[John Roberts {JR}]

Ted Olausson wrote on Thu, 14 February 2008 09:34

BTW how can an amp that is stated to have 1000w/8ohm and 1100w/4ohm at 0% distortion hade a DF above 1000!? It should be 1000/2000/4000 watt at 8/4/2 ohm...

I understand the logic behind your misunderstanding but it assumes unlimited current capability in the amplifier. Amplifier DF is related to output impedance which is strictly a matter of device characteristics, internal wiring, and negative feedback, or more specifically loop gain margin. An amplifier's maximum power at lower load impedances, is dominated by current capability and only in the margin to source impedance related output stage losses.  An amplifier's close but not perfect doubling, suggest source impedance effects. Wider divergence from 2x per halving is caused by PS or protection circuitry current limiting output stage.

In an ideal world, there would be a strict doubling of power with halving of load impedance but in practice this requires delivering 4x the amp at 2 ohms as for 8 ohms. Customers have routinely voted with their purchases to support a lesser capability, The marketplace provides what customers are willing to pay for.

JR

[Rob Spence]

This is the LF driver I have connected to my DC300
It is mounted to drive into a sealed enclosure (that is, basket side out).

Driver     Fs     Qts     Qes     Qms     Vas     Re     Prated     Xmax     Diam
L12P48    19.5    0.4    0.44    10    9.52    5    150    0.22    12

Note, I don't claim to understand these parameters and didn't choose them. The speaker cabinet maker did that.

[Tim Duffin]

That is not what I was talking about.  A class C "Amplifier" in the sense I am familar is used for unidirectional excitation of particles in water. I think about gyrators for ultrasonic cleaning while you are thinking of gyrators for RF... I never even knew that they had a circuit called a 'gyrator' in RF.  

I remember having to assemble a class-c amp in college and look at its output on a scope-- looked like a bunch of humps with some overshoot at zero crossing.  

T

[John Roberts {JR}]

Tim Duffin wrote on Fri, 15 February 2008 20:22

That is not what I was talking about.  A class C "Amplifier" in the sense I am familar is used for unidirectional excitation of particles in water. I think about gyrators for ultrasonic cleaning while you are thinking of gyrators for RF... I never even knew that they had a circuit called a 'gyrator' in RF.   I remember having to assemble a class-c amp in college and look at its output on a scope-- looked like a bunch of humps with some overshoot at zero crossing.   T

A gyrator is an active circuit to synthesize an inductor from resistor, capacitor, and a gain stage. A gyrator could be operated at RF frequencies.

Class C is a resonant amplifier circuit commonly used in RF.

It is you who seems to be confusing the two in your post here .   http://srforums.prosoundweb.com/index.php/m/290789/30842/0// /361/#msg_290789

Tim Duffin wrote on Thu, 07 February 2008 20:09

Fine, Ill use the quote button. Anyways, the correct answer for my question you did not get. The circuit which is class C is called a "gyrator" and it is not used in audio as per JR's comment.

FWIW, what the output of a class C amp looks like will depend on the rep rate it is excited at.. it could look sinusoidal if excited synchronously with it's resonance. If hit with an impulse it will look like a damped ring .  

I am not familiar with your "unidirectional excitation of particles in water" perhaps a pump? Gyrators and Class C have two pretty well defined and unrelated meanings in audio.

I don't doubt there are sundry other devices called gyrators.

for example

Rush developed a variation of this spinning approach called the Gyrator, a horizontal board on which torpid patients were strapped and spun to stimulate blood circulation. (1) An Irish physician developed his own version of this twirling device: O'Halloran's swing (figure 3). Patients were rotated up to 100 times a minute in it. The centrifugal force drove blood to the brain, theoretically treating mental illness and gaining patient obedience. (5)



JR

[Ted Olausson]

John Roberts  {JR} wrote on Thu, 14 February 2008 18:15

I understand the logic behind your misunderstanding but it assumes unlimited current capability in the amplifier. Amplifier DF is related to output impedance which is strictly a matter of device characteristics, internal wiring, and negative feedback, or more specifically loop gain margin. An amplifier's maximum power at lower load impedances, is dominated by current capability and only in the margin to source impedance related output stage losses.  An amplifier's close but not perfect doubling, suggest source impedance effects. Wider divergence from 2x per halving is caused by PS or protection circuitry current limiting output stage. In an ideal world, there would be a strict doubling of power with halving of load impedance but in practice this requires delivering 4x the amp at 2 ohms as for 8 ohms. Customers have routinely voted with their purchases to support a lesser capability, The marketplace provides what customers are willing to pay for.

Sorry, i was tired. OPAMPS has the same behaviour.
But what other (Known or suspected) factors except DF has a great impact on the relationship between speaker and amp?

[John Roberts {JR}]

Ted Olausson wrote on Sat, 16 February 2008 08:53

Sorry, i was tired. OPAMPS has the same behaviour. But what other (Known or suspected) factors except DF has a great impact on the relationship between speaker and amp?

I am very much a proponent of amplifiers sounding quite similar to each other as long as operated linearly.

Besides voltage clipping which is more common than many operators suspect, there is running out of current. Especially for an amplifier that doesn't follow the approximate doubling of power with halving the load impedance. That amplifier will current clip before it voltage clips at low impedance loads.

Another poorly understood factor is how loudspeakers, especially when using passive crossovers vary their impedance with frequency. Impedance dips with paralleled speakers could drop quite low at spot frequencies.

The most audible differences are frequency response errors, and added signals (distortion). The tiny differences between amplifiers operating linearly are IMO insignificant in the context of the loudspeakers they are connected too, and room modes, and sources like microphones.

JR

[Ted Olausson]

John Roberts  {JR} wrote on Sat, 16 February 2008 16:06

I am very much a proponent of amplifiers sounding quite similar to each other as long as operated linearly. Besides voltage clipping which is more common than many operators suspect, there is running out of current. Especially for an amplifier that doesn't follow the approximate doubling of power with halving the load impedance. That amplifier will current clip before it voltage clips at low impedance loads.

Now we are getting somewhere....
But as you say, the linear area doesnt go all the way to clip, it will deviate much earlier than that. Especially with amplifiers with small PSUs and speakers with high levels of back EMF.

Quote:

Another poorly understood factor is how loudspeakers, especially when using passive crossovers vary their impedance with frequency. Impedance dips with paralleled speakers could drop quite low at spot frequencies.

That is close to one of the secrets behind the mystical speakers i was mentioning...
Their Re is 5.2 but their Z is 11 ohm within their BW.
(One of the amps is also very close to "2xpower" from DC to 80MHz and has perfect stepresponse, the other is a normal amp)

Quote:

The most audible differences are frequency response errors, and added signals (distortion). The tiny differences between amplifiers operating linearly are IMO insignificant in the context of the loudspeakers they are connected too, and room modes, and sources like microphones.

So are the differences between preamps....

However, the speakers electrical Q is made by the amp, and when you let Qes change then FR also changes.
-and so does the stepresponse.

Protection will also affect the sound much earlier than expected, a perfect example of that is QSC EX vs MXa which is the same amp with different protection, and there is a big difference between them even at -20dB.
There is also a numbergame with distortion, THD% is very inaudible but IM% is almost never stated but highly audible.

BTW, i am not saying that amps are important in comparision to other elements, i am just saying that amps arent perfect.

[Andy Peters]

Ted Olausson wrote on Sun, 17 February 2008 06:50

(One of the amps is also very close to "2xpower" from DC to 80MHz and has perfect stepresponse, the other is a normal amp)

80 MEGAHERTZ?

Why does it need bandwidth out that far, if it's an audio amplifier?

-a

[Alexander B Larsson]

Ted Olausson said:
"There is also a numbergame with distortion, THD% is very inaudible but IM% is almost never stated but highly audible."

Well, Labgruppen and QSC certainly displays figures for IM distorsion, at least for the amp models I checked.
Coincidence?  

I DO find it interesting that many of the posters really liked the sound and reliability of the "old, continous power, 24/7" amps. Many of you have also clearly said that you ARE willing to compromise on the bass accuracy to get more power and lighter racks.

And I think nobody argues that the D class technology is still in need of more development, since the IM and load matching issues are still not really solved...

/Alexander