Do I need an amp?

[jostenmeat]

0.7 ohm? I hope it is a typo, or your amp is as big/heavy as a welding machine.

Actually, I've always had the cheapest amp section that I am aware of among persons that own Summits. Its now a Cambridge 840a, A to B class switching, IIRC at 10w or something, and goes from published 200w to 300w at 4 ohms. The minimal is hit at 20khz which is exotic indeed. Still, electrostats have been known for wanting solid amps. I just decided that the budget would be better managed if I could fit in some treatments. I've spent more on treatments than the integrated ($950) at this point.

If the dips happen in a wide enough spectrum including the mid bass area, a reputable company such as Klipsch would not have rated it 8 ohms nominal.

Never specifically came across this point before. Thanks.

 

[mtrycrafts]

.. It saves me some money that I could use for bills, lol.

Bills should come well before a hobby When you run out of bills, then, by all means

 

[adwilk]

Not to beat this to death but i have a real world application...

I have the RF-7's and a modest onkyo HT receiver. My buddy has a yammy 2500s or something like that.. even at insane SPL's we couldnt notice anything that warranted me changing to separates. Hell, even the sonic impact t amp we had sitting around rated at like 7 watts kept up for the most part. There were some low end response issues and audible distortion for a brief moment... i dont have a sub and i have plenty of the low stuff from my ht receiver..

 

[jostenmeat]

If the dips happen in a wide enough spectrum including the mid bass area, a reputable company such as Klipsch would not have rated it 8 ohms nominal.

I thought about this a little more, PENG, and I was wondering . . . there are a lot speakers out there that are rated as 8 ohm nominal, but have minimal impedances of 3 ohms or thereabouts. I think of most of the 800 BW series, or many of Focal's Be or Profile line.

OTOH, my PSB Images are rated as 6 ohm nominal, yet without as low of a dip, namely 4 ohms.

Do you believe that these "reputable companies" are stating such specs in regards to any decently "wide spectrum of lower impedance"?

Should I take it that such a "lower spectrum" is wider with the PSB than it is with the BW 802D? Sorry to bug you again, but I'm curious for your thoughts on this idea . . .

 

[fmw]

I thought about this a little more, PENG, and I was wondering . . . there are a lot speakers out there that are rated as 8 ohm nominal, but have minimal impedances of 3 ohms or thereabouts. I think of most of the 800 BW series, or many of Focal's Be or Profile line.

OTOH, my PSB Images are rated as 6 ohm nominal, yet without as low of a dip, namely 4 ohms.

Do you believe that these "reputable companies" are stating such specs in regards to any decently "wide spectrum of lower impedance"?

Should I take it that such a "lower spectrum" is wider with the PSB than it is with the BW 802D? Sorry to bug you again, but I'm curious for your thoughts on this idea . . .

In the industry impedance ratings for both speakers and amplifiers are "nominal" which you can define roughly as "average." All speakers have variations in impedance along the frequency spectrum-some more than others, of course. That's normal. Don't worry too much about minimum (or maximum) impedance levels. It is the nominal that should guide you when choosing hardware.

If the minimum impedance were presented to an amp continuously, that would be a different matter. But these minimums are usually present for very short periods of time - not long enough to overheat a power supply. If the speakers are nominal 4 ohm, then any amp with a 4 ohm rating should drive it without working up a sweat.

In a typical home environment where amps work at a small fraction of their potential power output, even the nominal impedance rating isn't that big a deal. A 100 wpc amplifer running at 1 or 2 watts continuously doesn't encounter the problems that one would encounter running 50 watts or 100 watts continuously. Music peaks might cause distortion but they don't really cause much overheating. It is the continuous applications that do that.

I've seen an AV receiver drive wild electrostatic speakers loudly and continuously with no undue heating. The speakers were rated at 4 ohms nominal and had dips as low as 1 ohm. And that's a receiver with no more than a 6 ohm rating. If you were to try to fill an auditorium with sound with that arrangement you might throw the amp into protection mode. In a home, however, it doesn't take much power to play loud and the receiver didn't mind it at all. Yes, we connected a scope to check for clipping. No clipping. Yes, it is my very own receiver.

Keep in mind that specifications and measurements can be like statistics. You can make them say whatever suits your fancy or interpret them in a way that can guide you in opposite directions. And audiophiles revel in making mountains out of molehills. It's always been that way.

 

[jostenmeat]

fmw, did you answer my question? If you did, are you saying that reputable companies are stating nominal+minimal however they want, and not somewhat accurately as PENG was saying? Or what am I misunderstanding?

As far as I have learned nominal didn't ever mean much (which is why I'm interested in PENG's thoughts, for it appears that this may not be true after all).

For example, speaker wire gauge is a function of minimal impedance and length.

Or I have learned from some others, such as TLS guy, that it's the minimal impedance that is the only thing that matters in regards to the amp chosen.

From Merriam Webster:

nominal (adjective)

Definition: being something in name or form only

Related Words: so-called; phantom, virtual

Antonyms: actual, real, true

 

[adwilk]

This is a pretty sweet little read.. the discussion is based on what speakers might be appropriate to be driven by tubes in terms of using a speakers published electrical specs, namely impedance...

Click Here!

 

[mtrycrafts]

fmw, did you answer my question? If you did, are you saying that reputable companies are stating nominal+minimal however they want, and not somewhat accurately as PENG was saying? Or what am I misunderstanding?

As far as I have learned nominal didn't ever mean much (which is why I'm interested in PENG's thoughts, for it appears that this may not be true after all).

For example, speaker wire gauge is a function of minimal impedance and length.

Or I have learned from some others, such as TLS guy, that it's the minimal impedance that is the only thing that matters in regards to the amp chosen.

From Merriam Webster:

nominal (adjective)

Definition: being something in name or form only

Related Words: so-called; phantom, virtual

Antonyms: actual, real, true

There is an two part article at the AH university link at the home page that may help on this:
here is one part
http://www.audioholics.com/education/loudspeaker-basics/loudspeaker-power-handling

 

[PENG]

I thought about this a little more, PENG, and I was wondering . . . there are a lot speakers out there that are rated as 8 ohm nominal, but have minimal impedances of 3 ohms or thereabouts. I think of most of the 800 BW series, or many of Focal's Be or Profile line.

OTOH, my PSB Images are rated as 6 ohm nominal, yet without as low of a dip, namely 4 ohms.

Do you believe that these "reputable companies" are stating such specs in regards to any decently "wide spectrum of lower impedance"?

Should I take it that such a "lower spectrum" is wider with the PSB than it is with the BW 802D? Sorry to bug you again, but I'm curious for your thoughts on this idea . . .

First of all, I do believe 'reputable companies" including companies such as PSB, B&W and klipsch would factor in the impedance dips when they rate their so called "nominal" impedance. If those dips cover a wide spectrum, say for a good part of the audible spectrum (20 to 20,000 Hz), then they would not (or should not) have rated their speakers 8 ohms nominal. That may be why PSB rated yours 6 dB because they have those 4 ohm dips and most likely those dips cover a significant portion of the speaker's frequency response spectrum. I thought fmw sort of answered your questions the way I would have, except that I do not agree nominal impedance = average impedance. Let's say speaker X measured 2 ohms flat between 20-10,020 Hz and 14 ohms between 10,021-20,000 Hz, then do we want the manufacturer of speaker X rate this speaker 8 ohms because the average is (2+14)/2=8 ohms for the spectrum 20-20,000 Hz? I certainly hope not! He is right about the fact that impedance varies in such a way that you cannot rate it as a single value, so you have to call it something else hence the term "nominal". It is not necessarily (most likely not) the average value. It should, however, be a value considered typical enough within the frequency response spectrum. And that should mean that an amplifier rated to drive speakers rated for this nominal value can be safely used to do a decent job with such speakers. Of course I can be wrong, but I would bet I am not far off in the case of speakers made by "reputable" companies.

Before I comment more on your second question, I would like to comment on the article linked by auwilk.

You will always find articles such as this one that talked about how meaningless "specs/numbers" are. TLS seems to be big on that too and I respect that just the same. Some of those articles were presented by people who sell amps, and in such cases we have to wonder if their motives are not a little complicated. That being said, I did not find anything factually wrong in that article from the electrical theory stand point. I do find it futile to talk about differentiating sensitivity and efficiency. To me, it is like trying to differentiate "effect" vs "affect", "currently vs presently" etc., mostly common errors. The only difference is, efficieny and sensitivity are technically totally different except that when Polk Audio (I also consider them reputable) specified their speaker's efficiency 88 dB we could easily guess that they actually meant "sensitivity" of 88 dB @ 2.83V 1 meter. Polk actually would say something like efficiently (AKA sensitivity). I don't like it neither but I know what they meant. The article also use a 91dB sensitivity speaker that has the what he referred to as "For large spreads of the audible bandwidth this speaker measures under 4 ohms." characteristics as an example to show why it is clear how misleading sensitivity can be when selecting a tube-friendly speaker.

While there is a lot of truth in what the author said and that 91 dB sensivity may be considered high, the OP's Klipsch has a published sensitivity of 99 dB @2.83V 1m. They did not say whether that 99 dB is anechoic or in a so called typical room, but either way there is no doubt the RF-5 can be driven to very high SPL in rooms that most ordinary people can afford. The author of the article seems a little extreme in trying to tell people that sensitivity numbers are meaningless. Again, he is right in the specifc examples he cited but in my opinion he is wrong in a general sense.

Now take a read of this:

http://www.polkaudio.com/education/article.php?id=4

In this article Polk Audio said:

"A highly efficient loudspeaker might be 90 dB. Each 3dB increase in efficiency doubles the sound output for a given power input. So a 100-watt, 90 dB self-powered speaker and a 200-watt 87 dB unit would produce exactly the same sound output."

This, on the other hand seem a little simplistic. It would have been more accurate if Polk had qualified the statement with something like "......for speakers having the same electrical characteristics, e.g. impedance & phase angle vs frequencies......................

And this, from klipsch website (note that, like it or not, they also use "efficiency" and "sensitivity" interchangeably):

http://www.klipsch.com/support/faqs/Default.aspx

"A. The output of a speaker (measured in decibels (dB)) for a given amount of amplifier power is termed its efficiency. Speakers with high efficiency will require less amplifier power to produce a given sound pressure level. Live music and movie soundtracks have peak levels as high as 120 dB. To attain just a 3 dB increase in volume, an amplifier's power must be doubled regardless what the speaker efficiency. Therefore, the amount of amplifier power needed if you wish to attain the sound pressure levels of a live orchestra or a large explosion in a movie is greatly dependent on the efficiency of your speakers. Here is a chart that shows two different speakers, one with 88 dB efficiency (common) and one with 100 dB efficiency, and the amplifier power need to produce given sound pressure levels:


The above charts show that the typical speaker with a sensitivity of 88 dB requires 1024 watts just to get to 118 dB! A more efficient speaker (example here is our KLF-30) requires just 64 watts to hit those levels of live music and Home Theatre. Therefore, the amount of amplifier power that a person needs is determined by the efficiency of the speaker that person is using plus the desired dB levels he or she wants to be able to produce. Other variables which impact the choice of amp power are the size of the listening room, the absorption characteristics of the wall and floor surface materials and the distance between listener and speakers. In a smaller room and with high efficiency speakers, 50 watts may be enough to provide full dynamic range. In a large room with low efficiency speakers, even 500 watts may not be enough. Higher speaker efficiency always helps deliver the most from whatever power you choose."


Finally, you last question, yes I would think in that case your PSBs must have lower impedance over a wider part of its frequency response spectrum than the B&W 802Ds do. For B&W to rate the 802D 8 ohms nominal with 3 ohms minimal, they have to be sure that amps rated for 8 ohm loads can be used to power their speakers sucessfully.

 

[PENG]

It's possible many speakers do this, you may never aware. Just avoid those 20hz test tones, and Resident Evil: Apocolypse if you can.

It may be possible, but I do not believe "many speaker do this." I would say such occurrences are rare.

Think about this, at 0.7 ohms, even if it is purely resisitve, i.e. phase angle=0 (cos0=1)

At 2.83V, current=V/I=2.83/0.7=4.04A. Power=I^2*R (for resistive load)=4.04^2*0.7=11.44W.

So at this resistance (worse if inductive or capacitive) level, the amp would have to be 11.44 times more powerful than if it only has to drive an 8 ohm load. How many amps can do that at decent to high SPL level in a not too large room of say 12X20X9 ft room?

Another way to look at it, the $12,000/pair 802D has a minimum impedance of 3.5 ohms. Speakers at that price range typically would not compromise their design by balancing performance and cost, and they ended up with 3.5 ohms minimum so what the odds must be low for any manufacturers ended up with super low 0.7 ohm speakers minimum or not.

I understand some of the electrostatic speakers (by the sound of it jostenmeat may have a couple) present such low impedance at 20 kHz, but at that frequency the current demand would not be too high so it should not stress the amp.

 

[jostenmeat]

This is a pretty sweet little read.. the discussion is based on what speakers might be appropriate to be driven by tubes in terms of using a speakers published electrical specs, namely impedance...

Click Here!

Thanks, I will definitely give that a look when I make some time.

There is an two part article at the AH university link at the home page that may help on this:
here is one part
http://www.audioholics.com/education/loudspeaker-basics/loudspeaker-power-handling

Yeah, I remember that when it was first linked at this site. I read/skimmed it, and wasn't extremely interested in power handling. Somewhat related, but didn't really answer my question precisely. More to do with handling/dissipation of heat is seems. I was just specifically curious about mftr's stated nominal and minimal published specs. I'm going to give that a re-read for the hell of it soon. Thanks.

First of all, I do believe 'reputable companies" including companies such as PSB, B&W and klipsch would factor in the impedance dips when they rate their so called "nominal" impedance. . .

Finally, you last question, yes I would think in that case your PSBs must have lower impedance over a wider part of its frequency response spectrum than the B&W 802Ds do. For B&W to rate the 802D 8 ohms nominal with 3 ohms minimal, they have to be sure that amps rated for 8 ohm loads can be used to power their speakers sucessfully.

An eloquent post, but its the bookends to your writing I was interested in. The stuff in between I believe I already understood, or at least had already inferred. Thank you very much. Nevertheless, I'm sure the meat of it will be very helpful to others here.

It may be possible, but I do not believe "many speaker do this." I would say such occurrences are rare.

Think about this, at 0.7 ohms, even if it is purely resisitve, i.e. phase angle=0 (cos0=1)

At 2.83V, current=V/I=2.83/0.7=4.04A. Power=I^2*R (for resistive load)=4.04^2*0.7=11.44W.

So at this resistance (worse if inductive or capacitive) level, the amp would have to be 11.44 times more powerful than if it only has to drive an 8 ohm load. How many amps can do that at decent to high SPL level in a not too large room of say 12X20X9 ft room?

Thanks. I think I learned about this math firstly from TLS, I can't remember from which thread. Can you teach me how to execute this math with any particular phase angle? (You might've already done that . . . with some Infinity speakers?). Thanks.

Thanks to all.

p.s. I swear there's a moratorium placed on my ability to dish out rep. sheesh, I swear its been 24 hrs.

 

[mtrycrafts]

Think about this, at 0.7 ohms, even if it is purely resisitve, i.e. phase angle=0 (cos0=1)

At 2.83V, current=V/I=2.83/0.7=4.04A. Power=I^2*R (for resistive load)=4.04^2*0.7=11.44W.

So at this resistance (worse if inductive or capacitive) level, the amp would have to be 11.44 times more powerful than if it only has to drive an 8 ohm load. .

But, this resistance happens at 20kHz, no? How much power is delivered into tweeters at such frequencies in the first place? Perhaps, very little. Or, even at lower frequencies, say 16kHz or 10kHz? Many/most such tweeters are rated to about 20 watts at most. Would that amp put out 11 watts into that .7 ohm load for a dynamic passage?

 

[jostenmeat]

This is a pretty sweet little read.. the discussion is based on what speakers might be appropriate to be driven by tubes in terms of using a speakers published electrical specs, namely impedance...

Click Here!

So I read this article, and I found it pretty interesting. I'm not planning on ever using tubed amps, but the applications or "matching" to certain speaker characteristics was interesting.

I do have a question or two. Resistive = flat, or zero angle, right? Inductive = positive phase angle, and capacitive = negative?

I was under the impression that any severe angles are tough, right?

If that's true (which it might not be), why does the article state, "A resistive load translates to a flat impedance, and an inductive rather than a capacitive phase angle"?

Is it because there is nothing for discussion's sake that is purely resistive, and flatter impedance curves are considered to be benignly inductive . . . or something like that?

Thanks!
-jostenmeat

p.s. I was already second guessing my thinner 16 g speaker wire due to the recently linked audioholics article. Now I have to wonder about EMF.

 

[poutanen]

p.s. I was already second guessing my thinner 16 g speaker wire due to the recently linked audioholics article. Now I have to wonder about EMF.

Not that this has much bearing on anything but when I opened up my Yamaha SW118's, the wire going from the connector to the sub was 18 gauge... and these things are meant to run lotsa power for extended periods of time. Mind you they're only running up to about 150-200 Hz...

 

[PENG]

But, this resistance happens at 20kHz, no? How much power is delivered into tweeters at such frequencies in the first place? Perhaps, very little. Or, even at lower frequencies, say 16kHz or 10kHz? Many/most such tweeters are rated to about 20 watts at most. Would that amp put out 11 watts into that .7 ohm load for a dynamic passage?

You quoted only part of my post, if you read the last paragraph you will see that we are saying the same thing. That is, no, 0.7 ohms at 20,000 Hz is not a big deal. I wasn't sure if Seth=L was referring to 20,000 Hz when he said "many speakers do this......". He also said "Just avoid those 20hz test tones...." May be it was a typo, perhaps he meant 20 KHz.

 

[PENG]

So I read this article, and I found it pretty interesting. I'm not planning on ever using tubed amps, but the applications or "matching" to certain speaker characteristics was interesting.

I do have a question or two. Resistive = flat, or zero angle, right? Inductive = positive phase angle, and capacitive = negative?

I was under the impression that any severe angles are tough, right?

If that's true (which it might not be), why does the article state, "A resistive load translates to a flat impedance, and an inductive rather than a capacitive phase angle"?

Is it because there is nothing for discussion's sake that is purely resistive, and flatter impedance curves are considered to be benignly inductive . . . or something like that?

Thanks!
-jostenmeat

p.s. I was already second guessing my thinner 16 g speaker wire due to the recently linked audioholics article. Now I have to wonder about EMF.

You are right, for a purely resistive load the impedance versus frequency curve will be flat and the current produced by the EMF (voltage) will be in phase with the applied (across the load) voltage meaning the phase angle is zero. This is true up to a point because at very high frequency, even resistance will begin to increase with frequency due to the so called 'skin effect'. For a purely inductive load, the current will be 90 degree lagging the applied voltage while for a purely capacitive load the current will be leading the applied voltage by 90 degree.

You also understood correctly that a highly inductive or capacitive load would result in a large phase angle between the applied voltage and the resulting current. The larger the phase angle, the more difficult it is for the amp because it would have to output a higher voltage to push through a higher current in order to produce the same 'real' power output (watts) that the amp would be able to develop with a resistive load at a lower voltage. Basically the power formula is P=VoltXCurrentXcos(phase angle). Cos is the short form of the trigonometry term 'Cosine'. Cosine(0)=1 and it decrease as the angle increase, to a minimum (in absolute value) of Cosine(90)=0. As you said, practically speaking there is no such thing as pure resistive, inductive or capacitive load so you don't have to worry about a 90 degree phase angle. However, cosine(60)=0.5 so a combination of 8 ohm impedance and a phase angle of 60 degree is as bad as a 4 ohm impedance and a phase angle of 0 degree. In fact if the phase angle is 0, you can consider the 4 ohm impedance as 'resistance' because it is purely resistive.

I don't know what you meant by "..wonder about EMF"...
EMF is electromotive force and its unit is volt, (V being the symbol). In that article you just read, the way the author talked about the counter EMF effect made me wonder if he got it the wrong way. Counter EMF is supposed to oppose, hence 'counter' the applied voltage, not aiding it. It is possible that I misunderstood the wordings he used though.

 

[mtrycrafts]

You quoted only part of my post, if you read the last paragraph you will see that we are saying the same thing. That is, no, 0.7 ohms at 20,000 Hz is not a big deal. I wasn't sure if Seth=L was referring to 20,000 Hz when he said "many speakers do this......". He also said "Just avoid those 20hz test tones...." May be it was a typo, perhaps he meant 20 KHz.

Yes, of course, you know what you wrote Me, must have stopped short of that last sentence, or lost my train of thought by then,

 

[jostenmeat]

You are right, for a purely resistive load the impedance versus frequency curve will be flat and the current produced by the EMF (voltage) will be in phase with the applied (across the load) voltage meaning the phase angle is zero. This is true up to a point because at very high frequency, even resistance will begin to increase with frequency due to the so called 'skin effect'. For a purely inductive load, the current will be 90 degree lagging the applied voltage while for a purely capacitive load the current will be leading the applied voltage by 90 degree.

You also understood correctly that a highly inductive or capacitive load would result in a large phase angle between the applied voltage and the resulting current. The larger the phase angle, the more difficult it is for the amp because it would have to output a higher voltage to push through a higher current in order to produce the same 'real' power output (watts) that the amp would be able to develop with a resistive load at a lower voltage. Basically the power formula is P=VoltXCurrentXcos(phase angle). Cos is the short form of the trigonometry term 'Cosine'. Cosine(0)=1 and it decrease as the angle increase, to a minimum (in absolute value) of Cosine(90)=0. As you said, practically speaking there is no such thing as pure resistive, inductive or capacitive load so you don't have to worry about a 90 degree phase angle. However, cosine(60)=0.5 so a combination of 8 ohm impedance and a phase angle of 60 degree is as bad as a 4 ohm impedance and a phase angle of 0 degree. In fact if the phase angle is 0, you can consider the 4 ohm impedance as 'resistance' because it is purely resistive.

I don't know what you meant by "..wonder about EMF"...
EMF is electromotive force and its unit is volt, (V being the symbol). In that article you just read, the way the author talked about the counter EMF effect made me wonder if he got it the wrong way. Counter EMF is supposed to oppose, hence 'counter' the applied voltage, not aiding it. It is possible that I misunderstood the wordings he used though.

Oh, so that's what skin effect is. I always thought skin effect had to do with those cool-looking, snake-skin wrapped speaker wires filled with snake oil for liquid effect!

I'm giving away how little I know about math, but how did you come up with cosine(60)=0.5? What does cos(45) equal, for instance?

Absolute value for phase angle, whether pos or neg, ok. I like how you used the current's "point of view" in explaining this. That current leading is negative phase, or lagging means positive phase. So "postive phase" could be stated from the voltage's "leading/positive" position, correct?

Thanks PENG. I never thought I would ever care to learn about this stuff, but when I run into someone who is both knowledgable and more than willing to simply explain, I just can't help myself.

OHHHHH, I'm an Audioholic General ! ! !

 

[PENG]

but how did you come up with cosine(60)=0.5? What does cos(45) equal, for instance?

I can tell you cos(0)=1, cos(30)=0.866, cos(45)=0.707, cos(60)=0.5, cos(90)=0 and sin(0)=0, sin(30)=0.5, sin(45)=0.707, sin(60)=0.866, sin(90)=1.

Given these commonly known values you can use the compound angle formula to calculate cos(15), cos(75) easily.

sin(A + B) = sinAcosB + cosAsinB
sin(A - B) = sinAcosB - cosAsinB
cos(A - B) = cosAcosB + sinAsinB
cos(A + B) = cosAcosB - sinAsinB
cos2A = cos2A - sin2A

Example:
cos(75)=cos(30+45)=cos(30)cos(45)-sin(30)sin(45)
so that cos(75)=0.866*0.707-0.5*0.707=0.2588

You can also derive the sine, cosine, and tangent functions for any angles using infinite series. Anyone with a college or university science/engineering degree would have gone through all these things but that does not mean everyone can remember things they learnt.

If you have not done trigonometry and basic principles of electricity (usually covered in university physics) and are interested in the topics, then I would suggest you go and spend so time in a good library. Of course you can also google the topics, but in that case choose what you read carefully.