Impedance Curves

[Deang]

In the following RB5 review:

http://www.audio-ideas.com/reviews/loudspeakers/klipsch-rb5.html

There is this:

"The impedance curve shows woofer and port peaks, the latter at 50 ohms, a low point between them of 5 ohms, with a high of 29 ohms in the midrange. Maybe this speaker isn't quite the answer for single-ended fans after all, since frequency response will tend to track the impedance curve with these amps in such circumstances."

What does the latter statement mean, and does this statement also apply to a push-pull triode design like my AE-25?

------------------

deanG

[mobile homeless]

First of all, that Ad at the top of this page is painful. I try to block it out but it is like bad solid state overdriven through a Klipsch midrange horn!

deang, as for your question, it is refering to the fact that the dips to 5 ohms and back up to 29 would put undue strain on SET amps that like relatively BENIGN impedance curves, surely none that take dramatic dips, especially in the low end of the spectrum. Most SET amps generally just dont do as well with speakers that dont produce an easy load. Yes, sensitivity plays a part too, but these impedance dips are also a BIG challenge and are not always handled with aplomb. Some do it better than others however.

As for your push-pull Superamp, this should not be quite as much a problem. Still, it does not help matters. Since you HAVE the RB-5 speakers, just relax and let your ears do the measuring since this is ultimately the bottom line. Speaker and amp interactions can be a funky thing with many variables coming into play. Dont let that review fool your ears.

That being said, others with the RB-5 might take the measurements into account before opting for the ultra-low watt SET solution. Still, listening is the best answer - Always try to audition gear if at all possible. OF course, this is not always that easy to do.

kh

ps- Do you have the RB-5 or RF-5? Actually, I forgot which you use.

Phono Linn LP-12 Vahalla / Linn Basic Plus / Sumiko Blue Point

CD Player Rega Planet

Preamp Cary Audio SLP-70 w/Phono Modified

Amplifier Welborne Labs 2A3 Moondog Monoblocks

Cable DIYCable Superlative / Twisted Cross Connect

Speaker 1977 Klipsch Cornwall I w/Alnico & Type B Crossover

system one online / alternate components / Asylum Listing f>s>

This message has been edited by mobile homeless on 03-12-2002 at 02:21 PM

[Deang]

you mind as give up trying to remember who has what on this site

i actually use the RF7's

------------------

deanG

[justin_tx_16]

quote:

---

Originally posted by deang:

you mind as give up trying to remember who has what on this site

i actually use the RF7's

---

did you confuse the RB5 with the RF7?!

------------------

-justin

SoundWise Support

A technical help site created by me and my fellow Klipschers

I am an amateur, if it is professional;

ProMedia help you want email Amy or call her @ 1-888-554-5665 or for an RA# 800-554-7724 ext 5

Klipsch Home Audio help you want, email support@klipsch.com or call @ 1-800-KLIPSCH

RA# Fax Number=317-860-9140 / Parts Department Fax Number=317-860-9150s>

[mike stehr]

Get it now!!!

get it now!!!

get it now!!!

[Deang]

Justin,

Are you asking me or mobile?

I never forget what i have I just figured since everything in the Reference line uses similiar drivers and crossover components/points -- I thought the impedance curves might also be similiar.

Mike,

You should buy a set. They have them at Brandnamez.com for $529.

This message has been edited by deang on 03-12-2002 at 02:49 PM

[justin_tx_16]

quote:

---

Originally posted by deang:

justin,

you asking me or mobile?

i never forget what i have

---

i as asking you LoL... i thought you were wondering about the impedance curves on your RB5's and you really were wondering for them on the RF7's... LoL sorry for the confusion I as just amazed that someone would get the two mixed up haha I should have known better

------------------

-justin

SoundWise Support

A technical help site created by me and my fellow Klipschers

I am an amateur, if it is professional;

ProMedia help you want email Amy or call her @ 1-888-554-5665 or for an RA# 800-554-7724 ext 5

Klipsch Home Audio help you want, email support@klipsch.com or call @ 1-800-KLIPSCH

RA# Fax Number=317-860-9140 / Parts Department Fax Number=317-860-9150s>

[Ray Garrison]

deang,

Just to expand a bit...

The issue here is that the frequency response of an amplifier can be affected by the relationship of the output impedence of the amp and the input impedence of the speaker.

An amplifier that cannot provide an unlimited amount of current will be happier driving a load whose input impedence is quite a bit greater than the output impedence of the amp. Picture it as giving the amp something to "push against". The closer the output and input impedences come, the more current delivery the amp will be called upon to provide. A tube amp, most of them anyway, are not very good at providing large amounts of current, so the effeciency of the power transfer of the amp will be highly dependent upon the input impedence of the speaker. If the speaker's impedence varies from fairly high values to very low values, the amps frequency response will also vary. A solid state amp, or a tube amp that can deliver a lot of current, will be less impacted by the varying impedence of the speaker, so the frequency response variations will be less severe. (Actually, before I get corrected here, what we're talking about is the output impedence of the amp, not the current delivery capabilities, but for all practical purposes one implies the other; high current delivery capability usually implies a very low output impedence.)

For example, take a look at this.

Here is a graph showing the frequency response measured by Stereophile of the Cary Audio Design 300B powered integrated amp. Note the severe frequency response variations into Stereophile's "dummy speaker load", a circuit that mimicks a "typical" speaker's impedence curve:

Here is the response of a Wavelength Gemini 2A3 powered amp, under the same test conditions: note that the anomolies are less severe:

And, just to show that SOME tube amps can be very linear into real world speaker loads, here's the McIntosh MC2000 into the same loads:

Hope that helps...

------------------

Music is art

Audio is engineering

Ray's Music System

This message has been edited by Ray Garrison on 03-12-2002 at 05:25 PM

[mobile homeless]

So the McIntosh Mc2000 sounds much, much better than the Wavelength Gemini, right? This is the problem with graphs like that and with the assumptions taken from as much, especially when dealing with easy to drive speakers. Ray, is that the dummy load that mirrors real world operating conditions?

There is actually so much more going on here that dictates sonics and how your amp will actually sound with real music and various types of speakers. One must look at what the dummy load signifies.

kh

[mike stehr]

Dean, I was posting about the Monster Wire Ad.

All that purdy PCV.

Never mind!

[Ray Garrison]

Mobile,

Granted, the frequency response of an amp into some arbitrary load is not going to tell you how it sounds any more than measuring, say, the maximum lane change speed of a car is going to tell you how it drives. It is simply one factor that tends to get measured because it can be measured. And, it does give you an idea of how much interaction there will be between the perceived frequency response and a speaker load.

I was just answering Dean's question about what the statement meant regarding frequency response variations due to the input impedence.

The exact parameters of the dummy load can be found on Stereophile's website, for those who are curious.

Ray

------------------

Music is art

Audio is engineering

Ray's Music System

[Ray Garrison]

Mdeneen,

Info like this (again, using the Cary 300) is available on the Stereophile site:

Fig.7 Cary CAD-300SEI, THD+noise vs frequency at (from top to bottom at 200Hz): 4W into 2 ohms, 2W into 4 ohms, 1W into 8 ohms, and 2.83V into dummy load (right channel dashed).

Comments / observations?

Ray

------------------

Music is art

Audio is engineering

Ray's Music System

[mobile homeless]

Load it up, Ray! Throw the softball pitch right down the middle! High and slow...heh.... Indeed.

It is widely known how divided ole Stereophile is concerning SET amplification. You have Atkinson against most of the others, all having left, leaving Stereophile grasping for straws. Even ole Robert Harley waxed lyrical on the Cary despite the specs, not to mention Olsher, Skull, and many of the old diehards.

I will admit, the Cary 300SEI is one of the worst measuring SETs around. It is a VERY soft sounding SET as well, and I dont believe near the sound quality of their other offerings, at least extension wise. It is the most rolled off SET I have heard, but still very musical, even given it uses Solid State rectification.

The Wavelength Gemini via the 2A3 is half the watts of the Cary 300SEI integrated, but uses a better circuit and FAR FAR better iron.

Still, I can only shake my head at those that live by the measurements regarding audio gear. Hell, that was the one thing Holt pioneered; LISTENING to the gear! He was the first to really develop this aspect.

If interested in some of JGH's musings, check out this article written about 10 years ago. It addresses the subjective vs Objective viewpoint and goes through some of the vocabulary that Holt and others pioneered. Well worth a read.

http://www.stereophile.com/showarchives.cgi?50:0

kh

[Randy Bey]

Interesting you mention this mdeneen, this distortion thing. One of the main things that got me going on tubes, and SET in particular, was the low distortion that the circuit allowed.

I am certainly not an EE and would probably cringe if poked at, but from what I know the distortion of a tube goes down as output goes down -- I think this is a restatement of what you said earlier. In the 1watt region, the distortion is almost unmeasurable. What I heard concerning SS is that they always reference distortion at 100% power since that is when their distortion is lowest. Never any mention of distortion in the first watt.

That is the furthest I have ever gone into measurements, besides doing some in-room SPL curves and addressing the peaks and valleys there. It's like finding a beautiful (to you) woman and deciding that you would marry her if she would have you. Others may not think she's all that pretty, but so what?

[Deang]

I want to revisit this subject. It has come up again in my mind, and I have a question regarding some of the comments.

The thread centered around the following comment by a reviewer of the RB5.

"...Maybe this speaker isn't quite the answer for single-ended fans after all, since frequency response will tend to track the impedance curve with these amps in such circumstances."

I have some additional questions revolving around the following comments.

Mdeneen said -- "What's not plainly discussed therein (the review), is the fact that what is important when speaker meets amplifier, is the resultant ACOUSTIC POWER CURVE, or SPL. Naturally the speaker has this wavering impedance curve because it is a reflection of the mechanical and electrical resonances within the entire speaker system - drivers and crossovers. When a driver (or circuit) is at resonance, not only does the impedance rise, but the acoustic output rises and given the fixed voltage of the amplifier, the amplifier power "drops", but the SPL is level. In other words, the driver is most efficient at the zenith of resonance and least at the nadir. And accordingly, the amplifier "power" is the opposite. If this were not true, speakers and amps of all types would simply not be listenable! The amplifier power output measure is simply E^2/Z. So, if you put a 4-ohm to 100-ohm pot on an amplifier and set the amp to 1V out and twist the pot back and forth, the power will rise and fall to the tune of E^2/R. And so, speakers have this wavering impedance curve, but when mated to almost any amplifier the ACOUSTIC POWER CURVE (SPL) - the part you hear with your ears - is generally very level and very independant of amplifier type (given the amps are operated in their rated ranges.)."

I say -- I see a bit of a contradiction here. How can the acoustic power or SPL remain "level", if the impedance rising causes an increase in "acoustic output" (or inversely, the impedance dropping causing a decrease in acoustic output). It would seem to me that at any frequency where the impedance drops very low, or goes very high -- the frequency response is going to show a "notch" (either up or down) at that point, and the overall signature of the sound is going to reflect those fluctuations. Now, I'm told the only way of avoiding this, is to make sure the amplifier can deliver the necessary output voltage into any impedance. Anyone want to take a guess at what kind of amplifiers can do this, and what kinds cannot?

Ray Garrison said -- "The issue here is that the frequency response of an amplifier can be affected by the relationship of the output impedence of the amp and the input impedence of the speaker. An amplifier that cannot provide an unlimited amount of current will be happier driving a load whose input impedence is quite a bit greater than the output impedence of the amp. Picture it as giving the amp something to "push against". The closer the output and input impedences come, the more current delivery the amp will be called upon to provide. A tube amp, most of them anyway, are not very good at providing large amounts of current, so the effeciency of the power transfer of the amp will be highly dependent upon the input impedence of the speaker. If the speaker's impedence varies from fairly high values to very low values, the amps frequency response will also vary. A solid state amp, or a tube amp that can deliver a lot of current, will be less impacted by the varying impedence of the speaker, so the frequency response variations will be less severe...."

I say -- then, it doesn't really matter much if a speaker is efficient enough for tube amplification, but it also has to have stable impedance across the range of it's frequency response. Since tube amps are "current limited", I have to wonder what happens with a speaker driven by a tube amp if the impedance dips to under 4 ohms at say, 7Khz? If current is not available, then the acoustic output at that point will drop, and if the impedance is at 10 ohms, the acoustic output will go up. How do we reconile this with the idea of desiring a "flat" frequency response?

Mobile said -- "deang, as for your question, it is refering to the fact that the dips to 5 ohms and back up to 29 would put undue strain on SET amps that like relatively BENIGN impedance curves, surely none that take dramatic dips, especially in the low end of the spectrum. Most SET amps generally just dont do as well with speakers that dont produce an easy load. Yes, sensitivity plays a part too, but these impedance dips are also a BIG challenge and are not always handled with aplomb. Some do it better than others however. As for your push-pull Superamp, this should not be quite as much a problem. Still, it does not help matters."

I say -- it's complicated stuff.

Kelly finsihes off his post by reminding me to use my ears -- which I'm not even sure I trust anymore:)

I've had conversations with both a source who wishes to remain anomonyous, and Dennis Had from Cary. I spoke to Dennis yesterday at length. Much was discussed pertaining to this thread, which is why I went back to dig it out.

Dennis Had didn't seem too thrilled with the idea of running his gear on a speaker where the impedance/ frequency response curves are unknown. This was his reaction when I related my understanding of this issue to him. He also concurred with the following information I received from the anomonyous source.

"An amplifier is supposed to be a voltage source because speakers are voltage sensitive devices.

If the amp has unlimited current, it can deliver the necessary output voltage (as determined by the volume setting) into any impedance.

As impedance drops, more current is demanded to maintain output voltage.

Tube amps are typically current limited, so when impedance drops the voltage drops and amplitude response changes.

Solid state amps are typically capable of higher current output so they can maintain necessary voltage into lower impedances.

High current amps (think solid state) would continue to double output WATTAGE with every halving of the load impedance, but they don't increase output VOLTAGE; they maintain voltage as impedance drops."

[mobile homeless]

Lord...lovely to drag this up again. It caused a mild headache the first time around, and that was in the more visually appealing FIRST forum, which still was far easier on the eyes if not the skull contents.

I am surprised that Dennis didnt discuss the idea that that impedance dips are much more problematic at the low end of the spectrum, where it really brings something like a SET to its knees, especially one with a mediocre power supply and circuit.

But I loosely agree with all that is said in that conclusion but will add that YOU NEVER KNOW how it will ultimately do until you give it a whirl in your own system. I cant stress this enough as I have been shocked time and time again with combinations that didnt make sense but sounded very good. On the whole, these are good guidelines and excellent info to keep stored in ye olde Gerbil wheel.

IF moving into SET amplification, the power supply and wattage does play a roll as well as the circuit and the type tube. Given two amps with 3.5w output, there WILL be differences in how they sound with a given pair of speakers depending on their design. But basically, you want a BENIGN IMPEDANCE that does not take big dips, just as said. In fact, I remember Had saying this was almost more important than the sensitivity in some ways. In other words, you can get lower on the actual sensitivity reading if you have a benign impedance that does not put more strain on the SET amp. Still, you want both if possible.

It is not a science per say in that measurements will NOT always tell you how your example of SET or tube amp will sound with the speakers. Room size and how hard you have to push the amps will play a role. But when people come on here and give absolutes regarding measurements and specs and what something WILL and WONT do, it should raise a red flag as this just is NOT the case. After years of messing with this, I have found out the exact opposite. Still, there are guidelines.

I had my 3.5w SET amps on a pair of very easy to drive ProAc Mini Towers that had a sensitivity rating BELOW 90 but a VERY benign impedance. They sounded VERY VERY nice. Yes, the Cornwalls did let the amps and presentation BLOOM and breathe a bit more, but the ProAc's did very well.

I am tempted to bring out another story regarding the EICO HF-81 and my friend from Quebec that builds 300B/2A3/V52 SET amps (among others). My hands are tiring but this will have to come out sooner or later.

BTW, I would think that the RF-7 curves can be found somewhere.

kg

ps- As another side, the Klipsch Cornwall response VARIES all over the place higher up the range with 8ohm to 30ohm differences, yet it still sounds very nice. IT does NOT drop below 8ohms really and the dip in the low end is not evident, hence it is not too much a problem. Being near 100dB sensitivity doenst hurt either

[jtkinney]

What about K-horns? There is an impedance curve posted on another forum that is scaled to miss the peak impedance of 55 ohms at 2150 Hz. Wouldn't this also be a bad speaker for a SET amp?

[mobile homeless]

See above in what I just wrote. It depends more on WHERE it happens and is MUCH better to go HIGH than to suddenly drop to 1ohm. Dips in the lows region are VERY problematic to any tube amp but especially SET.

On another note, NO AMP clips as gracefully as a SET amp. I mean NONE. It is one of the most simple, linear, and best amplifcation devices in this way.

That being said, it's not for everyone. I enjoy having one SET system tweaked to gills and one PP tube vintage system for that certain soul that you cant find anywhere else. Only those that have experienced it understand.

kh

[Deang]

Please do not misunderstand me. My post was not intended to imply that "the spec" will be the single determining factor in what sounds good. Instead, it was to bring to light the reality that just because a speaker is efficient -- does not always mean it is the perfect candidate for tube amplification. This is not just limited to SET, but also low powered push-pull triode. These amps, are by their very nature, "current limiting".

Sure, use our ears, and understand that something in the sound that seems "off", may be related to this relationship between the speaker's impedance, what it's doing to the amp -- and what the final result is as far as frequency response is concerned.

Maybe some should be looking at this -- instead of going into "Tweaker's Hell" with cones, 83 different types of interconnects/ speaker cables, 6 different CD players, 9 different preamps, and the like. There is certainly a place for some of this stuff -- I just think the most important thing is getting the right relationship between the speakers and the amp. If that's screwed up, then what's the point in it all.

I predict some elevated acoustic output at 2150Hz for the K-horn with a SET or p-p amp. Of course, if it's only 1 db or a little more, you wouldn't likely notice it.

[mobile homeless]

I dont get this... I thought this was being said in the VERY VERY beginning, at least by me. I think my VERY first posts on SET amplification talked about the importance of the "benign nature" of the impedance curve with speakers, even over the sensitivity rating. IF I remember, it was one of my first main points in what is necessary with SET and low watt tube amps in general. And I believe I usually included it as a cavat with every plug for low watt amplification.

Still, you are making a bit much of it. I dont think it is quite as drastic as you are making it sound. There is some leeway here and other positive factors can make up for some problems here. But this is a given with low watt tube amps, especially SET amplication, which does tend to track the impedance curve, though, once again, it is the DIPS that give the big trouble, especially if the speaker fall in the 3ohm and below.

Still, your system is made up of all the variables in play and as always, if your source, pre, and other variables are sub-par, or mismatched, then poor sonics result.

That being said, I think most of the people here have speakers that are actually FAR easier to drive than the average audiophile offering.

kh