The Impedence Problem

[mullett]

Hey guys,
I was just reveiwing the thread with my wife. She pointed out something that is I think very piculiar. If we are to go by the manufacture specs wouldn that mean that we have to read the specs first. She asked because...

And last but not least ... in order to keep this thread moving forward and on-track, at the end of EVERY POST, I ask that you copy and paste this:

PLEASE DO NOT POST TO THIS THREAD UNTIL YOU HAVE THOROUGHLY READ ALL OF THE PRECEDING POSTS IN THIS THREAD!

Joyce(my wife) says
"Typical bunch of guys, wont ask for help or follow the directions men just do what they want to anyway"

If you follow this logic most consumers do what the clerk tells them and don't try to learn any more than that. We have saught out this forum to try to learn more.
Wether I agree with you or not mikey you are trying to do something good...Trying is a big part of it.


Thanks for all the additional info Sydney. You see this kinda stuff as with the QSC amps in a lot of our car audio equipment Do this but not really type hype

PLEASE DO NOT POST TO THIS THREAD UNTIL YOU HAVE THOROUGHLY READ ALL OF THE PRECEDING POSTS IN THIS THREAD!

Still waiting for thet great day
Jeff Flowers

[Ron K]

RE: My 1st post:
#2 Find out about the amp

* However on page 13 in 4 languages:
DO NOT USE 2 OHM LOADS.

Chin up Mikey ( so your head is above the hype )

Hey Syd. you took that out of context! That quote is made regarding Bridge mono use not stereo use. Stereo or dual channel use can easily handle 2 ohm loads.

4 Ohm minimum in Bridge mono which is usually what you see unless you gots a Crown MA10000 running from 480v three phase power which is safe in mono to .5 ohms!

This thread albeit interesting will never have a finite end. Impedance is a complicated "issue" as far as reactance and capacitance are concerned and you haven't even factored in temperature and its affects on impedance nor the other components affecting impedance that may be in the circuit.

Even though there "may" be a frequency that "may" hit a very low impedance in a loudspeaker under given circumstances you still cant guarantee that it will last long enough to cause amplifier problems before it causes the voice coil to smoke from generated heat!There is already a industry wide acceptance of "nominal" impedance and it does apply to additional loudspeakers wired on the same channel.

I think manufactures specs of nominal impedance and add the affect of the cab to it and your good to go.

Until someone shows me test data to prove otherwise I'll stick with what is common practice even if it is wrong; simply because it works and blown drivers or smoked amps are not a norm here!

The proof of the puddin is in the eatin!

FWIW Nominal means close to but not exact "not" average. If it were average the value would be very high as most loudspeakers especially in enclosures have some very high impedance values within the usable passband!

Hence Danleys nominal 4 ohms and minimum 5 ohms! Close but not exact!

PLEASE DO NOT POST TO THIS THREAD UNTIL YOU HAVE THOROUGHLY READ ALL OF THE PRECEDING POSTS IN THIS THREAD!

[Sydney]

I wondered if that issue would come up.
From every example of amp design that I have seen that is bridgeable, you can get more efficiency/power when bridged and certainly better distortion figures and stability ( vs singular ) observing the impedance restriction.
If you are saying that each of 2 channels can handle a 2 ohm load, AND produce more power than bridged then I would have to disagree, that would make the aggregate sum of current/power greater in separate mode than the strapped mode*
* Unless you are accepting a reduced power output for better distortion figures, you still get the inferior 2 ohm distortion figures on each channel as each channel is running closer to instability .

If 2ohm ratings are given with disclaimers of 1% distortion @ 1000Hz for brief periods, then I dismiss the 2 ohm mode as hype

[Ron K]

I wondered if that issue would come up.
From every example of amp design that I have seen that is bridgeable, you get more efficiency/power when bridged ( vs singular ) but with the impedance restriction.
If you are saying that each of 2 channels can handle a 2 ohm load, AND produce more power than bridged then I would have to disagree, that would make the aggregate sum of current/power greater in separate mode than the more efficient strapped mode*
* Unless you are accepting a reduced power output with better distortion figures,
After all why would anyone bridge, if you can get more output unbridged.
If 2ohm ratings are given with disclaimers of 1% distortion @ 1000Hz for brief periods, then I dismiss the 2 ohm mode as hype

I'm not implying that at all and neither do the specs.

Marketing hype aside (which is what the 1K-1% spec largely is you wont get an argument there)it's still a useful spec.By not hiding the fact that distortion increases when operating within these parameters the spec retains some merit and you still get increased output by being able to drive lower ohm loads.Twice the output with half the load.......not in the real world.The ability to drive multiple motors from a single/dual channel/s.... yes, definitely .The ability to have increased output by bridging into a single motor. Yes definitely! Both at the same time. In bridge mono at 4 ohms possible if you hitting 2- 8 ohm loads.The same from a single channel would be less power available to the load!

You have no choice but to accept higher distortion figures as you approach the limits of output ability.The key is to operate with a reduced output but still higher than unbridged or with an easier load.It's never gonna be a direct doubling of output with halving of impedance. In fact what happens when the enclosure impedance figures raise at certain freqs? There goes those power specs right down the hopper.

There's no free lunch beyond the capabilities of any system!

It's a matter of separating the chaff from the wheat!

PLEASE DO NOT POST TO THIS THREAD UNTIL YOU HAVE THOROUGHLY READ ALL OF THE PRECEDING POSTS IN THIS THREAD!

[davygrvy]

Thanks for that IEC ref. David.

No prob. But what the IEC spec doesn't say is a "correct" method to determine what it should be. Do we assume it to be an average of the passband? Why isn't phase angle of the load an attribute as well as that effects the work the amp has to do?

[Sydney]

There's no free lunch beyond the capabilities of any system!

That's for sure...
As QSC says in the same manual under Safe Operating Levels
" With normal ventilation and 4 - 8 ohm loads, the amplifier will handle any signal level - including overdrive...
"However lower load impedances and higher signal levels produce more internal heating.
Into 2 ohm loads frequent or prolonged clipping may trigger protective muting."
I think that constitutes a disclaimer ( avoid more internal heating )

QSC claims 4X peak power and 3x sustained power in bridged mode.

[Sydney]

Attention 2 Ohm Fans:
Are you tired of your interconnects and cabling consuming over 20% of your power...
Try the Crown Belchfire BF-6000SUX
http://www.crownaudio.com/pdf/legacy/be ... asheet.pdf
A Rane PI 14 Pseudoacoustic Infector, coupled by a Jensen JE-EP-ERs Multi-denomial Transpedance Informer to a Crown Belchfire BF-6000SUX amplifier, will yield UNBELIEVABLE results.

* In the tradition of modern advertising: A dramatization, do not drive like this, Side Effects may include...

[Ron K]

Attention 2 Ohm Fans:
Are you tired of your interconnects and cabling consuming over 20% of your power...
Try the Crown Belchfire BF-6000SUX
http://www.crownaudio.com/pdf/legacy/be ... asheet.pdf
A Rane PI 14 Pseudoacoustic Infector, coupled by a Jensen JE-EP-ERs Multi-denomial Transpedance Informer to a Crown Belchfire BF-6000SUX amplifier, will yield UNBELIEVABLE results.

* In the tradition of modern advertising: A dramatization, do not drive like this, Side Effects may include...

OMFG Syd I completely forgot about that one. The RANE one I posted up a little while ago in another thread but that Crown one is the best especially with the jumper cables attached to the outputs!

Not to venture too far off from the main topic which is to establish a way to determine how many speakers can I hook up to my amp.

The real problem lies in the speakers specs!

If you believe "nominal" to be close to minimum than you can stick to the industry standard.

If you believe "nominal" to be an average than your gonna have some problems when doing the math.

"Nominal" by definition does not in any way suggest average! How that became a belief in the understanding of loudspeaker specifications is news to me but I have heard it used in those terms.

I've always believed it to mean close to the minimum value sometimes a tad over or sometimes a tad under! Those tads never seamed to matter at least not yet!

[davygrvy]

The real problem lies in the speakers specs!

I agree. The complexity of the load can not be simply stated with a single number. I just found this article that has a nice intro describing impedance of a loudspeaker.

I like how the author takes the minutia of 4.97Ω and correctly turns that into the IEC spec for nominal impedance: 4.97*10/8 = 6.21 rounded down to 6Ω. I like how he states the phase extremes of +39° and -54°as well.

But why do I concern myself with phase? It has to do with power factor. Phase of the load really screws with power factor as you can read about here.

Unfortunately, amplifier manufacturers don't usually list phase angle limits.

So, does a 2Ω load rated amp work with +/-60° limit and a minimum impedance peak of 1.6Ω (80% of nominal 2Ω)? I just can't say for sure with any certainty.

Sydney is probably waiting for me to mention this, so here it is.. This one guy came up with a metric for determining how difficult a speaker is to drive called EPDR. I tried examining it myself and thought is was wonderful. But there isn't an equivalent rating for the amplifier side and I really wonder how far this method should be examined.

One thing to note about the ESP article is that it is describing class A/B amplifiers. There are many other types these days such as H, G, etc.. and all have different dissipation behavior.

PLEASE DO NOT POST TO THIS THREAD UNTIL YOU HAVE THOROUGHLY READ ALL OF THE PRECEDING POSTS IN THIS THREAD!

[Ron K]

The real problem lies in the speakers specs!

I agree. The complexity of the load can not be simply stated with a single number. I just found this article that has a nice intro describing impedance of a loudspeaker.

I like how the author takes the minutia of 4.97Ω and correctly turns that into the IEC spec for nominal impedance: 4.97*10/8 = 6.21 rounded down to 6Ω. I like how he states the phase extremes of +39° and -54°as well.

But why do I concern myself with phase? It has to do with power factor. Phase of the load really screws with power factor as you can read about here.

Unfortunately, amplifier manufacturers don't usually list phase angle limits.

So, does a 2Ω load rated amp work with +/-60° limit and a minimum impedance peak of 1.6Ω (80% of nominal 2Ω)? I just can't say for sure with any certainty.

Sydney is probably waiting for me to mention this, so here it is.. This one guy came up with a metric for determining how difficult a speaker is to drive called EPDR. I tried examining it myself and thought is was wonderful. But there isn't an equivalent rating for the amplifier side and I really wonder how far this method should be examined.

One thing to note about the ESP article is that it is describing class A/B amplifiers. There are many other types these days such as H, G, etc.. and all have different dissipation behavior.

PLEASE DO NOT POST TO THIS THREAD UNTIL YOU HAVE THOROUGHLY READ ALL OF THE PRECEDING POSTS IN THIS THREAD!

Where does the resistance of the Voice Coil come into play.I seam to remember reading somewhere that ultimatly the coils raw resistance comes into play as its a physical characteristic set up by the coils diameter and length and influenced by how hot it gets.

Power compression??? Increased impedance due to heating???

[davygrvy]

The speaker is supposed to be operated in its linear range for testing. IEC 60268-5 states this:

16.2.2 Method of measurement

16.2.2.1 The loudspeaker shall be brought under normal measuring conditions in accordance
with 3.2.2, conditions a), b) and d).

16.2.2.2 A constant voltage or current shall be supplied, the former usually being preferred.
The value of voltage or current chosen for the measurement shall be sufficiently small to
ensure that the loudspeaker operates in a linear region.

NOTE Measurements of impedance may be strongly influenced by the drive level. If the level is either too low or
too high, inaccurate results may be obtained. The data should be examined for consistency at several drive levels
in order to establish the best conditions.

16.2.2.3 The modulus of the impedance shall be measured at least over the frequency range
20 Hz to 20 000 Hz.

16.2.2.4 The results shall be presented graphically as a function of frequency. The value of
the voltage or the current shall be stated with the results.

It also says, but not in this spot of the text, that the conditions of the space it is tested in be stated as well (ie. half-space outside, room of 2000m^3, or trunk of car with a volume of XXX, etc.). If the bandwidth of the speaker isn't full-range, the impedance graph only needs to show its intended range, not 20-20k.

How a hot voice coil behaves, I'm not quite sure.

[Sydney]

Becoming ( annoyingly ) predictable am I?
I'm not sure if this has been posted:
http://www.transparentsound.com/measure ... ngdahl.pdf
It discusses several of the issues mentioned:
And a very general guideline for determining "Nominal" impedance"
The authors point out that "Nominal" is minimal standard - a rough categorization and guideline for general usage.
Long established Industry practice was to base this value on that portion of a speakers impedance curve, well above resonance ( fs ) where the load was mostly resistive, and below the rising impedance curve.
So the assumption of the impedance being 1.333... X DCR would be close enough. This was relevant for speakers being operated in that frequency range.
Case in point: In measuring numerous used 2206 drivers I found a DCR of 5.2 ohms X 1.333 = 6.9 Ohms Z (calculated). It is sold as an 8 ohm driver. DCR is approx 75% of impedance in the "mid range"
I measured a low of 6.4 ohms @ 0 phase point ( approx 225Hz ). At 100Hz ( 10 ohms ) and at 1000Hz ( 14 ohms ).
The distribution of values in this passband has much less variance than the range below 100 ( where the reactive portion has a much greater influence ).
Since this driver can be used in cabs for bass below 100 or above 100Hz; I would go with the impedance values in the desired frequency range.
If a speaker is operating within a octave of Fs ( I believe ) the electrical phase performance has to be taken in consideration, because of the increase ( current ) stress placed upon a amp when large neg phase angle occurs near low impedance ranges.
The effect of the enclosure ( another set of filters ) really has the greatest influence in the octaves nearest the drivers Fs.
I'm sure it's easy to see that the load placed by 4 15's paralleled ( 2 ohms ) in a badly tunned BR, is going to be different than the same 4 in a sealed cab or a horn. Or 4 8's paralleled ( 2 ohms with the same fs ) is going to have a different phase impact.
Because of the infinite variety of ways that ( 30 - 38 guage ) voice coil wire can be formed in to a coil, it is possible to get a target DCR with a vastly different Le.
http://gboers.xs4all.nl/daisy/home/g3/1 ... ecoil.html

How a hot voice coil behaves, I'm not quite sure.

David: I found some stuff about that if you are interested.

[Ron K]

"First, over the
majority of the range of operation, the voice coil resistance still dominates."

Thats kinda what I thought I remembered hearing and reading about many years ago!

[Sydney]

That's why I made the distinction with the 2206 used as a midrange vs a sub.
Midrange is level and resistive, Near Fs the "lossy" inductor is reactive.

[davygrvy]

So how do we come to a conclusion about "rated impedance" vs amplifier loading?

I like the EPDR technique on loudspeakers, but there is no corresponding amp spec for safe-operating-area intrusion.

We could probably start on the loudspeaker side and translate actual impedance graphs to the IEC "rated" value, then look at amps specs and get to the bottom (haha) of whether minimum rated load means "rated" or actual.

There's this fuzz on both sides and I don't like it, either.