Print

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

[Ivan Beaver]

I understand.
My peecker sound amps are rated 565w @ 8 ohms and 900w @ 4 omhs. And guys from peecker sound tech support told me that i should not use them at 2 ohms.

This factor of changing impedance of a driver, does it variate from a driver to a driver? Is it anywhere in the specs? And how is it related to frequency?
If i have an 18 sub rated at 8 ohms, what can i expect from it. Impedance jumps at higher or lower frequencies?

All that means is that those amps are not rated for 2 ohm operation.

But many other amps are.

I never said the impedance changed.  I said it varies with freq

Look at any impedance graph of a loudspeaker and you will see what I mean.

The "rating" is the simple number that the loudspeaker is closest to.

At some freq it will be correct, but at other freq it will be wrong.

Some speakers that are rated for 8 ohms could be 50 ohms at some freq. and 6 ohms at others.

Some speakers may have the impedance above the rated impedance and never go as low as the rating.

It varies speaker to speaker.

The impedance CURVE can give a lot of information regarding the tuning of the cabinet and so forth.

The "simple number" is only a rough indicator of the actual load.

[Scott Holtzman]

All that means is that those amps are not rated for 2 ohm operation.

But many other amps are.

I never said the impedance changed.  I said it varies with freq

Look at any impedance graph of a loudspeaker and you will see what I mean.

The "rating" is the simple number that the loudspeaker is closest to.

At some freq it will be correct, but at other freq it will be wrong.

Some speakers that are rated for 8 ohms could be 50 ohms at some freq. and 6 ohms at others.

Some speakers may have the impedance above the rated impedance and never go as low as the rating.

It varies speaker to speaker.

The impedance CURVE can give a lot of information regarding the tuning of the cabinet and so forth.

The "simple number" is only a rough indicator of the actual load.

This discussion happens, in part, because we use the term "impedance of a speaker, or driver"  The correct term is nominal impedance.  It is a value that nominally represents the average impedance under average conditions.  It is fine for back handed calculations of amp and speaker compatibility. 

What it is not is a constant to be used in formulas for sizing amplification requirements.  For that you need an integral of the curve that Ivan mentioned.   

In engineering we use the term Monte Carlo analysis to describe complex multi-parameter systems.  You analyze all the possibilities you find the limits.  Then engineer your system for your safety factor at those limits.

Frankly, it's a lot to digest for anyone with high level math background.   So many moving numbers. 

It all starts with the variability of the source material then branches out from that point.

We all want neat answers, what speaker should I use, what amp.  The truth is every decision is tainted by practicality.  Tom's  "the speaker has to fit through the door" principal. 

Then you add in the economic factors and lastly the emotional ones (purple driver syndrome.  great article BTW) and frankly our answers come out sounding like they were written by product liability lawyers. 

Your music, your room, your gear, your money and your fingers on the knobs.   

That's when you have to get real.  If you are blowing shit up you need to change the way you are doing things.  If your customers are paying you what you want for the gear you bring and are happy with the service you provide our opinion really doesn't matter.

I am starting to swerve off course, I think I made my point.

[Eugen Jeličić]

I see, but these graphs show drastic variations in impedance.
How do these impedance changes affect the amp? Would a driver with less drastical impedance jumps and drops over the frequency range sound better and why?
I didn't see things this way. I tought the amp will give a spaker roughly the same amount of power over the entire frequency range if the input voltage at all frequencies doesn't change.

[g'bye, Dick Rees]

I see, but these graphs show drastic variations in impedance.
How do these impedance changes affect the amp? Would a driver with less drastical impedance jumps and drops over the frequency range sound better and why?
I didn't see things this way. I tought the amp will give a spaker roughly the same amount of power over the entire frequency range if the input voltage at all frequencies doesn't change.

I assume the graph represents sine wave response whereas program material contains more frequencies/harmonics, so I would further assume the curve would appear less drastic.

[Ivan Beaver]

I see, but these graphs show drastic variations in impedance.
How do these impedance changes affect the amp? Would a driver with less drastical impedance jumps and drops over the frequency range sound better and why?
I didn't see things this way. I tought the amp will give a spaker roughly the same amount of power over the entire frequency range if the input voltage at all frequencies doesn't change.

The term "power amplifier" is mostly wrong.  But that doesn't keep people from using the term-just like phase when it should be polarity

Think of the amp as a VOLTAGE amplifier-NOT a power amplifier.  Because it IS a voltage amp and NOT a power amp

IF it was a power amp-then the output voltage (and resultant power) would be jumping all over the place due to the different impedance at different freq.

But for the most part (until we hit the limits) it is a constant voltage amplifier.

In fact with the higher impedance peaks-it actually "likes" the load better.  It is easier to "drive"-produces less heat-has more headroom and so forth.

It is also related to another incorrect term when talking about sensitivity 1W.  NO manufacturer applies 1 watt when doing this measurement.  They apply 2.83V (which happens to be 1 watt @ 8 ohms)

However if they were actually applying 1 watt-then the voltage at the freq where the impedance is higher than 8 ohms would HAVE to be higher (in order to get 1 watt)

This would produce a higher output at those freq and the response would get even more ragged.

But they apply a constant 2.83V-and the power will vary with freq

So at some freq-there is much LESS than 1 watt being applied.

So the term 1 watt (when talking about sensitivity) is technically wrong.  The CORRECT term is 2.83Volts-because that is what is being applied-NOT power.

[Eugen Jeličić]

So if we have 2 drivers rated at 4 ohms nominal and one of them is having changes/jumps and at almost 50% of the frequencies it reproduces it hits almost 8 ohms impedance.
Then we have the second driver which is also rated at 4 ohms nominal but for for the most part it stays near 4 ohms and doesn't have that many peaks at higher impedances.
Then the amp is going to have an easier time driving the first driver but the second driver will be using the amps power better and be louder over most of the frequency spectrum.
Assuming that both drivers have the same sensitivity.

And yes i know a power amp is basicali a voltage amp and that 2.83V is used for sensitivity measurements.
But this also means that when you have two drivers rated at 100dB sensitivity at 1khz. Unless i take the impedance graph from both of them and compare the actual power going into both drivers under those conditiones i am not doing a good side by side comparison?

One of the drivers might be a 10ohm load at 1khz, the other one might be an 8 ohm load at 1khz even though they are both rated at 8 ohms nominal?

[Ivan Beaver]

So if we have 2 drivers rated at 4 ohms nominal and one of them is having changes/jumps and at almost 50% of the frequencies it reproduces it hits almost 8 ohms impedance.
Then we have the second driver which is also rated at 4 ohms nominal but for for the most part it stays near 4 ohms and doesn't have that many peaks at higher impedances.
Then the amp is going to have an easier time driving the first driver but the second driver will be using the amps power better and be louder over most of the frequency spectrum.
Assuming that both drivers have the same sensitivity.

And yes i know a power amp is basicali a voltage amp and that 2.83V is used for sensitivity measurements.
But this also means that when you have two drivers rated at 100dB sensitivity at 1khz. Unless i take the impedance graph from both of them and compare the actual power going into both drivers under those conditiones i am not doing a good side by side comparison?

One of the drivers might be a 10ohm load at 1khz, the other one might be an 8 ohm load at 1khz even though they are both rated at 8 ohms nominal?

I have never seen a speaker that has a sensitivity rated at 1KHz.

That could be a pretty useless measurement.

If 1Khz was a couple of dB higher than everything else, then the speaker would have a rating that would be false and people would expect it to be louder than it is.

A sensitivity rating should be an average across the intended freq range.  NOT a peak or a dip in the response.

I would worry about the power an amp is delivering in terms of looking at different speakers impedances.

An amp will not be any louder because of some impedance peaks.  It will simply run a tad bit cooler when reproducing those freq.

The impedance ratings are a "nominal standard" rating.

And the TRUTHFUL impedance (some sort of an average) may be quite different than the "simple number" says.

But people just don't seem to understand anything other than 2-4-8-16 ohms.  If you give a speaker a real rating of 6 ohms, it confuses lots of people because they can't find an amp rating for 6 ohms.

So people will give the closest "standard" number to avoid confusion-but it is not "correct".

[Eugen Jeličić]

Okay so sensitivity measurements should be taken as an average over the entire spectrum.

How? If a speaker is rated 8 ohms nominal. And it has a peak at 11ohm@500hz. That means that the amp won't be able to give the speaker the same amount of power at 500hz as at some frequency where the impedance is lower. And that means that it will run quieter ar 500hz because of the impedance jump.

Unless you have an amp powerfull enough and can boost the 500hz fader on the EQ a little bit.

On the other hand if we have another 8ohm driver but this one is keeping the 8ohm impedance at 500hz. That driver will run louder at 500hz then the previous one because it will pull more power from the amp?

[Ivan Beaver]

Okay so sensitivity measurements should be taken as an average over the entire spectrum.

How? If a speaker is rated 8 ohms nominal. And it has a peak at 11ohm@500hz. That means that the amp won't be able to give the speaker the same amount of power at 500hz as at some frequency where the impedance is lower. And that means that it will run quieter ar 500hz because of the impedance jump.

Unless you have an amp powerfull enough and can boost the 500hz fader on the EQ a little bit.

On the other hand if we have another 8ohm driver but this one is keeping the 8ohm impedance at 500hz. That driver will run louder at 500hz then the previous one because it will pull more power from the amp?

NO-The amp produces VOLTAGE.  Power is the RESULT of that voltage across a certain impedance load.

If you run a sweep of freq-the output voltage will be the same-NO MATTER what the impedance is.

The POWER will be different because of the different impedances.

You DO NOT boost freq that have a higher impedance.

The impedance has almost nothing to do with the freq response.

You can have points that have a high impedance and they can (but may or may not be) be LOUDER than those with a low impedance-even though there is less "power" going to the driver at those freq.

DO NOT try to put the sensitivity/freq response and the impedance curves together. 

Or to put it another way-if a speaker should "happen" to have flat impedance curve-that DOES NOT mean that the freq response will be flat.  It could vary quite a bit.

A flat impedance curve is no better or worse than one with a lot of peaks.

The impedance curve is the RESULT of the final design-including tuning of the cabinet-crossover design-driver design etc

[Eugen Jeličić]

You didn't understand me.

I realize that the amp just regulates and rises voltage and the power that is going through the coil is determined by the impedance at that moment.

BUT, We have Driver 1 and we have driver 2.

They both have exactly the same sensitivity curve.
They are both rated at the same RMS power.
They are both rated at the same nominal impedane of 4 ohms.
And they are both connected to an amp that shouldn't run under 4ohms and it's power is the same as the RMS power of the driver. So it's slightly underpowered for those drivers.

Now, the only thing that is different between these drivers is that driver 2 stays consistantly around 4 ohms over the entire frequency spectrum but driver 1 has peaks up to 8 ohms at many spots.

This basicaly means that driver 1 will run quieter of this amp generally.
Because it will be using it's power not as much as it could at all frequencies.
If we had a more powerfull amp, we could compensate for drivers impedance jumps by boosting power at frequencies at which the driver has higher impedance. Because if the driver can take 500w rms at all frequencies. And the amp needs let's say a 50v output at four ohms to use all of drivers headroom then if the impedance jumps to 8 ohms the amp needs to rise the voltage in order to use all of the drivers headroom.

So in a way, the impedance graph will have a connection with how loud the driver can get unless you have an overpowered amp and can push it as hard as possible at all frequencies. 

[ProSoundWeb Community]