How to accurately gauge speaker sensitivity to match with tube amp?

Fritz wrote:

" Hi David- Here’s the impedance and phase curves of the series crossover Carbon 7 SE Mk2 loudspeakers.
http://fritzspeakers.com/Images/speakers/Carbon7mk2ScanDome.jpg "

In case anyone doesn’t understand the implications of what they see at the link, that impedance curve is SUPERB!

I had never taken the time to look that deeply into Fritz’s speakers. I have heard them consistently sounding disgustingly magnificent at audio shows, to the extent that I have absolutely NO DESIRE to compete with him. And now looking at his impedance curve, I can see why: He’s cheating. He’s making life so easy for the amplifiers that they all simply love his speakers.

Tube amplifiers may well love his speakers the most: They will reward you with more output in the bass region. This is about as close to a "free lunch" as it gets.

Duke
competing in very different market segments from the one Fritz owns

http://www.fritzspeakers.com/DreamWeaverFritzImpedancePage.asp

And a bunch of comparative measurements.

Hi David- Here's the impedance and phase curves of the series crossover Carbon 7 SE Mk2 loudspeakers.  
http://fritzspeakers.com/Images/speakers/Carbon7mk2ScanDome.jpg

" A high damping factor will provide very good control of the bass drivers."

Hmmmm.

For a quick summation of my thoughts, skip to the last two paragraphs. Apologies for getting fairly nerdy in between here and there.

In practice, any series resistance in between the amplifier and the woofer’s voice coil effectively ADDS TO the amplifier’s output impedance, and correspondingly reduces the damping factor.

Let’s run some numbers. Supposes our speaker has a nice 2.5 kHz second-order crossover, which calls for a 1 mH inductor in series with the woofer. Power handling requirements are easily met by an 18 gauge air-core inductor, which can handle 300 watts before saturation. The series resistance of this inductor is .51 ohms.

And let’s suppose we have an uber-amplifier with a damping factor of one zillion. Or one zillion zillion. Or one zillion to the zillionth power. It won’t matter.

After the signal passes through that inductor, our uber-amplfier’s amplifier’s effective damping factor is now about 17. And this is assuming only the one series inductor, and ignoring any other wiring.

So in most cases it really doesn’t matter how high the amplifier’s damping factor is. The series resistance in the crossover (and/or speaker wires) dominates.

Okay, but what about this "very good control of the bass drivers" that we’re apparently missing out on?

Well, turns out that it’s not nearly as dramatic as the wording implies. It all shows up as a change to the electrical damping of the woofer’s motor - the electrical system Q, or Qes.

Assuming a typical high-quality 8-ohm woofer in the example above, the series inductor effectively raises the woofer’s electrical Q by about 7%. So if the woofer’s electrical Q was .28, the series inductor effectively raises it to about .30. This could EASILY be an improvement!  We'll get more bass with a higher Qes, but the designer should take it into account by sizing and tuning the box based on our modified Qes of .30. And if he hasn’t, this difference can still be largely compensated for with a few handfuls of stuffing material.

I think amplifier marketing departments may have oversold the benefits of having a high damping factor.

Or to put it another way, in my opinion, super-high damping factors are, in most cases, of academic interest only. I certainly would not trade off anything that really matters in order to get a high damping factor.

Duke

Again, David, keep in mind that the QS’s damping factor is higher than it is for many and I’m pretty certain most high quality tube amps. Assuming the spec is accurate, for the 4 ohm tap a damping factor of 20 means an output impedance of 4/20 = 0.2 ohms, not too far from solid state territory.

Although many audiophiles believe differently, Ralph ( @atmasphere ) and some other technically knowledgeable members (e.g. @kijanki) have explained in past threads that no speaker in existence requires a damping factor greater than the mid to high double digits, and perhaps even lower. Perceived differences in bass response and control between damping factors of say 100 and 2000 are due to other differences in the designs of the amplifiers, IMO and theirs.

Best regards,
-- Al

@mammothguy No bubble burst; I knew their on-paper damping factor but reports about their abilities were pretty positive. And they do a very good job, but I’m totally experimenting with speaker choices here and it cost me nothing to learn something. But I think that damping factor and these speakers provides something of a data point going forward. Appreciate your thoughts and please take care.

Adcom amps typically have a very high damping factor.  They are not necessarily the "sweetest" or "smoothest" on the upper-mid and top-end range, but they definitely drive hard and make for some impressive bass control.  A high damping factor will provide very good control of the bass drivers.  That is the very taught bass experience that you described.  I clicked on the link you provided for your amp and read through the specs.  It indicates that your amp has a damping factor of 20.  If the printed specs are accurate that is a very low damping factor and is likely the reason your speakers sounded as you experienced.  

I don't mean to burst your bubble on that amp, but that is what I see in their specs and that is what I understand (and have experienced many times) with a low damping factor.  If you really wish to keep that amp then a different set of speakers is likely in order.  Some of the options and choices mentioned above certainly make good sense.

I wish you good luck in your quest.  Be safe and be healthy.

Thanks J. Chip. I could hear some locally. Not the Triton 2s?

OP

Dont know your budget but you may want to check out the Golden Ear line of towers. I drive my Golden Ear Triton R1's with 50 watts of tube power which proves more than enough. Most GE speakers have a high spl (mine are 92 spl) and are equipped with class D power amps to handle the bass (each of the Triton R1's are equipped with a 1600 w  class D amp.) This allows relatively low powered amps to easily drive the mid range and higher without strain. Well regarded line of speakers.

J.Chip

great thread gentlemen. with tube amplification I've always used loudspeakers with 8 ohm nominal specs. I agree running the 4 ohm taps takes something out of the bottom end but I didn't know how to explain it like some of you techically oriented guys

@fuzztone

I'd call Mike first see what he thinks.

I wrote him. Only had a couple things for me — advises a 4 ohm tap and to get sensitive speakers.

@ieales
I figured my amps would have the power to drive towers. I have heard the ML Motion 60's are easy to drive; and other towers, too. I suppose if I cannot find a tower that has a friendly enough response, bookshelves will be fine. I do have a REL sub that's big enough to do some of what you suggest. (Crossover, gain, and phase.)

@djones @grinnell Thanks for the suggestions!

@jchiappinelli58 I had this very thought. I was conscious of making things equal but it occurred to me that I was just "eyeballing" ("earballing"?) it. An SPL meter is on my purchase list! Thank you.

To the OP

Maybe you did this but....

In order to compare the qualities of one amp to another, it is absolutely imperative that the sound levels be matched. Otherwise, the amp that provides a slightly louder output will seem to have the better sound. A simple Radio Shack SPL meter is a must have tool for anyone looking to compare components, sources, etc.

J.Chip

Ref 3A especially the de capos
Klipsch old school run the Quartets with 4 watts
Omega
Audio Note
Zu
Pi4

Klipsch heritage line works well with tubes. 

Smaller mains and a pair of subs:

Relieving the amp of the necessity of driving the low end roller coaster effectively doubles its power, allowing less efficient unported mains. [I've never heard ported low end I liked].

Getting smaller mains up on [sand or shot filled] stands solves several problems.

Subs like the Martin Logan ML---x series have ARC built in which can do wonders for bottom end coherency. Subs must have multi-slope crossover, phase and polarity controls. Delay is icing on the cake. The MLs are quite light and will stroll, so be prepared to weight them.

Just don't get swarmed by Mr Carbon.

From the link above [ https://www.dhtrob.com/overige/tubefriendly_lsp_en.php ] this seems helpful:

"[F] the purposes of determining tube-friendliness, sensitivity is essentially meaningless. Many tube gurus insist that one must use speakers with sensitivity ratings of 90 db or higher. This is simply not true. Many speakers with high sensitivity ratings are poor choices for tube amps, just as many speakers with low sensitivity ratings often work quite well….

Tubes like a resistive load - it’s really that simple….[F]rom a tube’s point of view is as follows: tubes like consistency. They prefer a speaker that doesn’t swing from 8 ohms to 2 ohms and back again….Tubes don’t like to pump out current at the drop of a hat. They prefer a consistent and predictable load.

Naturally it is sometimes difficult for speaker manufacturers to keep all the aforementioned variables in their ideal ranges. If phase angles must be capacitive, impedance should be high. The lower the impedance, the flatter it must be, and the less capacitive the phase angles must be. All three variables (slope of impedance, value of impedance, value of phase angle) together determine how good a match a speaker will be for a tube amp.

For this reason, the argument that only speakers with an impedance of 8 ohms or higher will work on tubes is incorrect. There are many tube friendly speakers whose nominal impedance is 4 ohms. If the speaker’s impedance is relatively flat and consistently hovers around 4 ohms, and if the phase angles are only slightly capacitive, or better yet inductive, there is no reason why a 4 ohm speaker cannot perform well on a tube amplifier. Some of our favorite tube friendly speakers are 4 ohms!"

The article concludes by affirming that both metrics matter AND listening matters:

"You can easily tell if a speaker isn’t performing well on tubes by... listening to it! Certainly we encourage you to gather as much data as possible such that you don’t waste your time auditioning speakers which are clearly not appropriate choices for tubes, but when it comes down to it folks the only way to find the right speaker is to listen to it anyway. A speaker that doesn’t perform well on tubes will typically be characterized by one or more of the following sonic traits when connected to a tube amp:

• Dual-mono imaging, i.e. the speakers do not "disappear"
• Unusually poor imaging or soundstaging
• Lumpy, bloated, or "one note" bass
• Lack of control in the bass - excessive boominess or thumpiness
• A "suckout" in the midrange
• Whenever the bass kicks in, the mids go away
• Whenever the bass kicks in, the highs get harsh and strained
• When you hook the speaker up to a solid state amp, the bass gets tighter, the midrange comes back, and the highs mellow out"

The last bullet point here tells me that the Focals are a bad match BOTH for my solid state AND my tubes -- because the Focal's bass sound mediocre on my tubes, and while the solid state tightens up the bass, the mids get grainy and the highs get harsh.

Thanks @pesky. 1 and 2 on your list sound like the key, here. “How far” the deviance is, and for how wide a range it spans, seem like the key to making a guesstimate. Erik gave some clue in his comment above. I am now searching the web for the meaning of phase angle beyond what you stated to see how that might be applied to my search. Came up with this so far:
https://www.dhtrob.com/overige/tubefriendly_lsp_en.php
I'm not sure how good the information is, but will try to cross-check.

Thanks johno...will look for that email exchange.

The absolute impedance means little if you have appropriate taps on your transformers. What does matter is:
1. How far the speaker’s impedance deviates from its nominal value (and where and how steep the dips and peaks are). Nasties in the bass region are a real no no.
2. What phase angles they generate (and where). Tube amps are generally less tolerant of speakers with high phase angles (where the voltage waveform is no longer in phase with the current demands) than well designed solid state alternatives.
Some fairly low sensitivity speakers with moderate impedance curves and phase angles can sound quite decent with tube amps if you’re not into Motorhead. If you are I would suggest looking at 92db+. They sound really excellent on Klipsch Quartets with only 35W on tap.

You asked ?But what else should I look for to help make a guesstimate about whether the amp will drive the speakers with the kind of control they are capable of?" 

This is a great question and I recently received an e-mail from the Audionote dealer in India, regarding the same question. The email was called "What Are Tube Friendly Loudspeakers?"

There detailed explanation was the best I've seen to date it unravels how to read those stereophile graphs of frequency vs impedance vs phase angle. 

Contact Audionote India, if you want a decent explanation
Audio Note India <info@audionote.in>

Chees Johno

Thanks, @yogiboy ! There are local audition options for those.

@hilde45
You should check out the Spendor A line. I have used Spendors with my Quicksilver amps and the match was outstanding!
https://www.spendoraudio.com/a-line-loudspeakers/

@yogiboy 
1.Room size?
430 sq. feet. 14x26 or so; listening on the long wall, about 9 feet away.
2.Type of music you listen to?
classical, jazz, rock
3.Budget?
4-5k tops
4.Floorstanders or Stand mount?
I'm open.

Yes, Fritz's speakers sound very good. Been trying his Carbon 7s and will try Salk SS 6M's. A lot of folks were advising trying floorstanders, too so that's the phase I'm in now.

Of course, this presupposes that the speakers you've become interested in have been measured reliably and that an impedance curve has been published.

Fritz's standmounts do very well with low-powered tubes (I heard them with a Modwright SET headphone amp!!) but I'm not aware of published curves for the various models.

1.Room size?
2.Type of music you listen to?
3.Budget?
4.Floorstanders or Stand mount?

@atmasphere Thanks for your answer. You're adding more technical information that is pertinent to my narrowing down the candidates -- the comment that "the speaker should not go much below 8 ohms especially in the bass region" is very useful, indeed (at least, to me). 

These speakers are trials. They're going back. No cost to me because I can drop them off. The question is, "what's next." Thanks for your help with that project.

@erik_squires 

 The more variation, the more the impedance curve takes command of the frequency response….

…but look how broad the under 4 Ohms is. That’s what I’d listen for. For the most neutral presentation from these speakers you are going to want a beefy solid state amp.

Great! Very helpful. FWIW, I only looked at the curve after I knew I was hearing something and wanted to understand why. I had seen such graphed curves before, but now they are connecting with my listening experience; they have meaning, now.

@miller

The only measurement you need is sensitivity greater than 90. Then from there you pay attention to listening impressions. Then when you find people describing a sound that matches what you like look for people who are hearing that sound with tube amps…..But on the other hand, maybe you don't care about such things as musical involvement.

LOL! If I didn't care about musical involvement, I'd be a fool. I hope I've not given off that impression with my question. I'm not a specs fanboy by a longshot, and your suggestion is most appreciated. People's taste (and hearing) can vary a lot, and so sometimes that's an X factor which is hard for me to glimpse in the many posts mentioning speakers in which I might be interested. Another tool (like the sensitivity number) was what I was seeking here, but I'll remain open to the possibility that you're right — that more than just sensitivity measurement is worthless and I should read up on listening impressions almost exclusively. (And, of course, all of this is not replacing listening, but getting whittled down the number of things to which I might go to the trouble of auditioning.)

@yogiboy  Thanks— good suggestion about the taps. They're currently on the 4 ohm taps, though.

@almarg That's right — I'm trying to use specs to help delimit a rather wide range of choice. The power of the amps are similar, but I know their damping factors are very different (20 vs. 100, at least). I will look into what other people's anecdotal reports say, too. I never ruled that out, but the metric of speaker sensitivity has been very helpful in narrowing down my choices and that's why I'm looking into related metrics, now.

The dynamic jumps in symphonic music (I spent a day just listening to symphonies) were not distorting with the tubes.

The Focal isn't crazy low for tubes but it will give them a hard time. You should use the 4 ohm tap with this speaker.
Generally speaking though if you want to get the best out of a tube amp then the speaker should not go much below 8 ohms especially in the bass region! If the amp has output transformers (and most do) its common for the output transformer to lose bandwidth by as much as an octave on the bottom when using the 4 ohm tap (and that's even when its properly loaded with a 4 ohm load). This can cause the amp to have additional phase shift and that is interpreted by the ear as a loss of impact.


So I see two ways out of this if the 4 ohm taps don't do it for you and you want to keep the amp. Either get a set of ZEROs (www.zeroimpedance.com) and see if they can help you out (by allowing you to use a higher tap on the amplifiers) or get a set of speakers that are an easier load.


Its really not a good idea to make an amp work hard for a living. They make a lot less distortion if they have it easy, and that translates directly to 'sounding more real'.

Your advice can help me eliminate candidate speakers that would pose similar challenges to my amp as these Focals have.

Yes, David (@hilde45), as you realize and as I have said in a number of past threads that is precisely the main potential usefulness of specs and measurements. **Ruling out** candidates that are likely to be mismatches with either user requirements (e.g., desired peak volume capability) or with other parts of the system, thereby narrowing the field of candidates as well as reducing the likelihood of expensive mistakes.

Your Adcom and QS amps have similar power ratings, so it seems that for your purposes speaker sensitivity is not an issue, as long as it is in the area of 90 db or so or higher. If you already haven’t, though, do make sure that the brief dynamic peaks of recordings having wide dynamic range (i.e., large differences in volume between the loudest notes and the softest notes, such as in the case of many classical symphonic recordings) are reproduced cleanly, without noticeable distortion.

What should I notice in those curves? Just that they don’t dip down as you described -- down to... 2 ohms? For how far across the mid bass?

Erik has given you excellent advice re impedance curves, and as he alluded to it’s hard to provide a more specific answer that would have general applicability. But I agree that in this specific case the 936’s impedance curve raises a lot of concern about how optimal it would be for use with tube amps. Although the relatively high damping factor/low output impedance of your QS (for a tube amp) will tend to lessen the degree of adverse impedance interactions with this speaker than would be the case with many other tube amps.

The suggestion of researching what kinds of amplification others have used with specific speakers is of course also a good one, as I’m sure you realize, as is the suggestion of changing the output tap that is being used. Especially if you have been using the 8 ohm tap, which perhaps you have been doing since the speaker has an 8 ohm nominal rating.

On another note altogether:

... as Roseanne Rosanna Danna used to say, never mind.

That was actually said by Gilda Radner’s Emily Litella character. Her Roseanne Roseannadanna character’s famous line (and also the title of a book Gilda wrote, which I believe dealt extensively with her battle with cancer) was of course "it’s always something."

Regards,
-- Al

Try changing the output taps on the amp and see if that helps!

I think you may be better off asking people which amp they use with their Focal 936s or Focal speakers in general.  There may already be threads on this topic that you can consult.

You have a 60 WPC tube amp with KT88s, virtually the same amp I have run for decades. https://systems.audiogon.com/systems/8367 This is a Melody 50 wpc. Before that was an Aronov very similar 50 wpc. Between the two ran four very different speakers JBL L7, Linaeum Model 10, Vandersteen something or other (very briefly) and Talon Khorus. All these speakers were around or a little more than 90 dB. All played plenty loud in that 17x24 room. All sounded great, for what they were. You have plenty of power.  

Don't even ask me what the impedance curves looked at for any of these speakers. I haven't the foggiest. Used to pay attention to such things. Then I realized impedance curves, like frequency response and all the rest, are at best second derivatives. In other words there is the final result, what you hear, and then there are all the things that go into that. The sound you hear is derived from sensitivity, dispersion, impedance, etc. They are not what you want. They are only things we have learned to measure and derive from what really counts, the sound we get.  

Now considering you are having problems I have never had in 30 years you might want to stop and think about it. If a guy has never had a problem in 30 years, always picked great satisfying speakers, what is he doing? How is he doing this? It ain't by looking at curves on charts.

The only measurement you need is sensitivity greater than 90. Then from there you pay attention to listening impressions. Then when you find people describing a sound that matches what you like look for people who are hearing that sound with tube amps. 

This takes no more time than searching through impedance charts. Its no more easy to describe the sound than measure the impedance. Look how many times JA and others go into excruciating detail about the technical aspects of impedance. Then at best all you have is one more extremely crude measure of performance. It tells you nothing about how deep and wide and inviting the sound stage will be.

But on the other hand, maybe you don't care about such things as musical involvement. Maybe you do only care about tiny little technical details like how tight the bass is between 146.9 and 237.8 Hz. If so then, as Roseanne Rosanna Danna used to say, never mind. 

OP:

I wish I had a better answer for you. Ideally, keeping the impedance above 4, as well as not varying too much. You also would like to see minimal peaks.

The more variation, the more the impedance curve takes command of the frequency response.

Based on all of this, I’d be surprised if you liked this combo:

https://www.stereophile.com/content/focal-aria-936-loudspeaker-measurements

but on the other hand, you might like how it sounds. You’ll get a brighter, more airy presentation, but look how broad the under 4 Ohms is. That’s what I’d listen for. For the most neutral presentation from these speakers you are going to want a beefy solid state amp.

With tubes you'll get a lot of color here, but who am I to tell you if you'd like it??

What should I notice in those curves? Just that they don't dip down as you described -- down to... 2 ohms? For how far across the mid bass?

Volume is fine with the speakers -- they get loud without a problem. It's that control, tightness, I'm looking for.

The Focals though, like a lot of modern speakers, have dips in the mid-bass which can really depress the output of a tube amp.

In the case of tube amps, I'd look more at the impedance curves (or minimum).  60 W is probably enough to drive them to fun levels in terms of volume.