Why 4 ohm?

[Rust]

While reading the news of the Magico M9, a question which has been bothering me for some time came to mind. Rather than highjack that thread, I post this.

Many of the high end manufacturers have gone to 4 ohm nominal impedance for their speakers. What, if any, are the advantages of a 4 ohm speaker?

At a lower impedance the I squared R heating of the voice coils requires more exotic measures to resist this. Inductors in the crossover need to be much larger. Amplifiers have to be much more robust to push current into speakers with impedance curves as low as 1 ohm. This 4 ohm trend also leads to what I consider somewhat disingenuous sensitivity ratings, that is the typical 4 ohm rating is at 2.83 volts which is NOT one watt into 4 ohms. At one watt a 90 dB rated 4 ohm speaker is closer to 86 dB. The only feasible advantage I see is possibly fewer turns of wire in the voice coil for a lighter moving mass which may be offset by a more robust former to resist heating from the higher current.

I see the M9 has somewhat skirted the larger inductor issue by going to a two component external crossover system requiring four channels of amplification, and gone to titanium formers to alleviate I squared R heating of the voice coils. Other high end manufacturers of low impedance speakers have taken various measures to address these issues too.

It's not just statement products either. 4 Ohm seems to have become a thing in recent years which means many relatively affordable speaker will require more robust amplification in comparison to an 8 Ohm speaker.

So what's the knock against 8 ohm and higher speakers? They are more amplifier friendly, especially for tubes. Is it suddenly more difficult to design higher impedance speaker? To steal a quote, inquiring minds want to know.

[dimapro]

Agreeing with you that pushing more current, i.e going lower impedance is not everything in sound terms. Wishing to see more 8 Ohm speakers with even curve as that's what usually tube-friendly. My ultimate dream is to see field-coil speakers to come back so you can adjust the level of "magnet" at your wish.

I guess it's just like the trend going towards exotic driver materials and staying away from paper-based drivers (that usually have more natural tone).

[Rust]

I'll agree with the more natural tone noting that Wilson has moved away from exotic materials to pulp and silk all the way up to their statement product. I would guess the choice is either to go with a material so stiff it has little resonance or a material that effectively self damps resonance.

[Rob]

with direct radiator speakers like Magico, YG, Wilson, Vivid, Rockport, etc. the average impedance is more often dictated by the driver's design. With multi driver speakers they are series/parallel wired to keep the impedance reasonable without resorting to compensation networks, higher insertion loss and the need for even more power. 4 ohms usually isn't the designers goal but rather the lowest impendence reasonable without sacrificing other performance parameters deemed more important than impedance alone.

[atmasphere]

Many of the high end manufacturers have gone to 4 ohm nominal impedance for their speakers. What, if any, are the advantages of a 4 ohm speaker?

At a lower impedance the I squared R heating of the voice coils requires more exotic measures to resist this. Inductors in the crossover need to be much larger. Amplifiers have to be much more robust to push current into speakers with impedance curves as low as 1 ohm. This 4 ohm trend also leads to what I consider somewhat disingenuous sensitivity ratings, that is the typical 4 ohm rating is at 2.83 volts which is NOT one watt into 4 ohms. At one watt a 90 dB rated 4 ohm speaker is closer to 86 dB. The only feasible advantage I see is possibly fewer turns of wire in the voice coil for a lighter moving mass which may be offset by a more robust former to resist heating from the higher current.

I see the M9 has somewhat skirted the larger inductor issue by going to a two component external crossover system requiring four channels of amplification, and gone to titanium formers to alleviate I squared R heating of the voice coils. Other high end manufacturers of low impedance speakers have taken various measures to address these issues too.

It's not just statement products either. 4 Ohm seems to have become a thing in recent years which means many relatively affordable speaker will require more robust amplification in comparison to an 8 Ohm speaker.

So what's the knock against 8 ohm and higher speakers? They are more amplifier friendly, especially for tubes. Is it suddenly more difficult to design higher impedance speaker? To steal a quote, inquiring minds want to know.

There is no advantage to four ohms at all. With any amplifier made the distortion is higher into 4 ohms, but in high end audio, it literally is about keeping distortion down! Many solid state amps and double power into 4 ohms, but getting more power isn't helpful if it comes with more distortion. And don't fool yourself: the ear converts all forms of distortion into some form of tonality. Most solid state amps don't make much distortion, but the distortion they do make is usually the higher ordered harmonics and they are easily heard.

The ear uses the higher ordered harmonics to sense sound pressure so it is more keenly sensitive to them than almost anything else owing to its +120dB range. So when you increase the distortion 'just a tad' with solid state you will get harsher and brighter as those two qualities are assigned to the higher ordered harmonics by the ear.

IOW if you want to have smoother sound with more detail, have your amp drive a higher impedance speaker.

Now if you have a tube amp you also get more distortion even if you are using the correct tap so the power tubes are properly loaded. The output transformer is simply less efficient into lower impedances, so it runs warmer (heat from power made by the output section...) and may lose as much as an octave of bandwidth off the bottom.

The other problem you run into is that 4 ohms is far more critical of the speaker cables. Their DCR is more significant than it is at 8 or especially at 16 ohms. IOW the higher the impedance of the speaker, the less critical your speaker cables become.

Now some speakers simply wind up being 4 ohms and are really pretty good. But you always have to ask- how much better could they be if they were 8 or 16 ohms instead?

Class D amplifiers have the lowest output impedance of all amps and so can act as a nearly perfect voltage source. But they have a filter at their output and that filter has a 'Q' value (Q standing for Quality, which is partially a function of the geometry of the inductor used in the filter). The designer of the class D has to set that Q to be optimized at some impedance. When you run lower impedances, the Q goes down, which means that the filter frequency is more broad and the phase shift covers a wider range. A self-oscillating class D can compensate for the phase shift since they have so much feedback, but the residual (what's left over from the switching frequency) will be higher amplitude. Its a lot easier to optimize the filter if the load impedance is 8 ohms or more!

The bottom line is this: If sound quality is your goal, your amplifier dollar investment will be best served by a speaker of higher impedance, all other things being equal. If sound pressure is your goal then you have a weak, 3dB argument for 4 ohms as opposed to 8.

Because solid state amps so often double power as impedance is halved, a very practical issue on paper is that of apparent drivability. This is literally why the sensitivity spec came to be in the 1980s and now has entirely eclipsed the older efficiency (1 watt/1 meter) spec.

Its a money thing.

Back in the 1950s when transistors were proposed, one advantage amp manufacturers saw was no output transformer and no filament supply, so they could make an amp for about 10-15% of the cost of a tube amp, yet still charge about 90% of the same money. When tubes were king, speakers were high efficiency to take advantage of the tube power, which is expensive. But higher efficiency requires more precision and so costs more. When cheap solid state amp power came along, speaker manufacturers realized a similar profit opportunity- they could make a less efficient speaker that cost 10% to make but price it just as high. To make it look more attractive to those that don't know what the spec sheet is saying, 4 ohms became a lot more popular! That's what is behind four ohms. Sorry, its not pretty but this sort of thing in audio is commonplace- Its rarely about performance and often about the bottom line.

[Slowgeezr]

Mep, I was just wondering. I always thought most loudspeakers have varying impedances. I hadn't really concerned myself with it, except to make sure my amps could handle the lowest impedances on speakers I've owned. Do most of the eight ohm impedance speakers stay near eight ohms over the entire or most of the frequency range? I thought if someone listed a good eight ohm rated speaker, I could then read about it and see how the impedance varies for it. Thanks.

[mep]

Slowgeezer-Thanks for the reply. Real world speakers don’t have a flat impedance that never varies from their rated nominal impedance. That’s why the stated impedance of speakers is rated as 4 ohms nominal or 8 ohms nominal. Having said that, an 8 ohm nominal speaker presents an easier load to drive than a 4 ohm nominal speaker. Speaker sensitivity is another variable that factors into the equation.

[Rust]

"The bottom line is this: If sound quality is your goal, your amplifier dollar investment will be best served by a speaker of higher impedance, all other things being equal. If sound pressure is your goal then you have a weak, 3dB argument for 4 ohms as opposed to 8."

Higher SPL by itself is not the goal, higher quality SPL is. On a beer budget I try to achieve if not champagne results, at least sparkling wine.

[1morerecord2clean]

Very interesting topic for me. I have close to zero technical knowledge. Interesting because my Piega C711 Coax concentric ribbon driver coupled with radiator woofers are 92db efficient with 4 ohms impedance are being driven by my Pass 150.8 amp. I sit only 9’ from the speakers and in my space the speakers seem very well driven. However I have toyed with trading up to the 250.8 not knowing if the added power will make a difference. And BTW, doesn’t the impedance factor of the speaker wires come into play with this question.

[Slowgeezr]

Surely the impedance factor of the cabling would be added, however small it might be, I would think.

[atmasphere]

Slowgeezr

I've never worried about it, except to have an amp capable of driving low impedance speakers. Could you name a few speakers that are 8 ohm, that sound excellent? Thanks.

Goodness- there are way too many for me to remember!!

My speakers at home are actually 16 ohms, 98dB and flat to 20Hz. They are the T-3.3 from Classic Audio Loudspeakers. They employ field coil drivers so there is no sag in the magnetic field surrounding the voice coil when current is applied to it- so they sound as fast as ESLs and pretty much for the same reason.

Audiokinesis has always made musical and easy to drive speakers. Audiokinesis is probably best known for their Swarm subwoofer system which is a Distributed Bass Array.

A number of B&W's smaller speakers are 8 ohms, such as their 805- any of their 2-way designs or those that employ a single woofer in the cabinet.

One of the best monitors I've heard was made by High Emotion Audio. They are no longer in business. They made a monitor called the Bella Twin that was available in 4 or 16 ohms- we have the 16 ohm version. One of the more revealing speakers I've heard, really fast, really smooth.

Usher has made some nice 8 ohm speakers. So has Coincident, Spendor, Silverline, Tannoy, Eminent Technology, Pro-Ac and dozens I've not thought of (apologies...).

[mep]

Rob-I don’t understand why people wouldn’t see the benefits of high sensitivity speakers.

[Rob]

high efficiency solves a multitude of problems. Many speaker designers Ive got to know over the years believe that high-quality, high-powered amps (almost always SS) are both plentiful and the de facto choice of most audiophiles. We both know that's not the case (for us, anyways).

[Slowgeezr]

Thanks Ralph (atmosphere). This is an interesting topic to me, as I've always been an "if it sounds good, that's what matters" person and never really worried about speaker impedances, except that I didn't want it so low as to affect the amp's capabilities to drive it well. I do like a speaker with relatively high efficiency, though as I think it helps with dynamics.

[Head_Unit]

This 4 ohm trend also leads to what I consider somewhat disingenuous sensitivity ratings, that is the typical 4 ohm rating is at 2.83 volts which is NOT one watt into 4 ohms...4 Ohm seems to have become a thing in recent years...Is it suddenly more difficult to design higher impedance speaker? To steal a quote, inquiring minds want to know.

Yes, more sensitivity. The DC resistance is in the denominator of the speaker efficiency equation-so lower impedances get a leg up even aside from the 2.83V issue. Now it's not quite that simple, because changing the impedance changes the voice coil which changes the wire in the magnetic gap and some amount of mass...but still, if you could build a superconducting zero ohm speaker it would be infinitely efficient, 287 dB powered by the small electrical changes in your brain as you think about the music ha ha. Your points about the difficulty for amps are valid...but the speaker manufacturers mostly could care less. The specialist amp manufacturers don't mind as they can sell you a beefier more expensive amp. It's like Microsoft Windows and Intel-the huger and huger software gave an excuse to sell stronger and stronger hardware.

[rbbert]

The manufacturer's stated "nominal" impedance is often different from the "nominal" impedance suggested by outside testers (e.g., John Atkinson at Stereophile but there are several others as well); it is sometimes higher and sometimes lower. I suspect that in many cases this is determined by the manufacturer to compromise between stating something too low (potentially harder to drive) or something too high (suggesting to many potential buyers that they may need a more powerful amp).

Is there an industry standard about how to determine a "nominal" impedance from the speakers impedance curve? As an aside, many speakers I have owned and/or have been interested in have an advertised impedance of 6 ohms; perhaps a compromise by the manufacturer to fall between the 2 philosophical positions I described above? Or perhaps that is simply where a formula determines the "nominal" impedance based on analysis of the speaker's impedance curve?

A couple of speaker designers with whom I have spoken say their main concerns with X-over design (a major factor in overall impedance) are flat frequency response, low distortion and accurate phase response (although some designers are apparently less concerned with that last, based on testing that indicates one or more drivers out-of-phase with the others, usually to optimize flat frequency response); this often leaves the impedance to simply fall where it may land.

[astrotoy]

From a speaker designer's viewpoint, is there a design advantage (better sound quality for example) of having a lower nominal impedance speaker?

If one starts with the amplifier, then the speaker, I can see that a favorite amp would dictate possible limitations on the speaker parameters. If one starts with the speaker, then that would dictate limitations on the amp.

I did the former - with the lower powered SET amps dictating a very efficient high impedance speaker.

Larry

[lance002]

I have used the same approach.

[adrianwu]

When I asked John Wolff at Classic Audio to provide a set of field coil drivers, he recommended the 16 Ohm 4" midrange, but the 8 Ohm 15" bass. I asked him why not 16 Ohms for the bass, and he told me that the extra windings on the voice coils exact a penalty in sound quality that is not worthwhile. In any case, with 100dB sensitivity and a very benign impedance curve, I don't think having 16 ohms would confer much advantage in this case.

[Steve Lefkowicz]

The designer for brand of speakers I use most of the time (Tekton Design) is a big proponent of low impedance speakers. Not only are most of the Tekton speakers a 4 ohm load, but some models are offered in a 2-ohm configuration. He does offer some models with an 8 ohm option, and the "Perfect SET" line are designed specifically to be SET Tube amp friendly with a stable 8 ohm load. The Perfect SET 15 I use mostly are 8-ohm and have worked beautifully with every amp I've tried, from about 4 watts to 100.

[atmasphere]

From a speaker designer's viewpoint, is there a design advantage (better sound quality for example) of having a lower nominal impedance speaker?

If one starts with the amplifier, then the speaker, I can see that a favorite amp would dictate possible limitations on the speaker parameters. If one starts with the speaker, then that would dictate limitations on the amp.

I did the former - with the lower powered SET amps dictating a very efficient high impedance speaker.

Larry

Many speaker designers like the increased sensitivity they get by using paralleled drivers. The thing is, sensitivity and efficiency are not the same thing. Sensitivity is a voltage measurement, 2.83V at 1 meter (which, if an 8 Ohm load, is also 1 Watt). Efficiency is 1 Watt at 1 meter. So a speaker that is 90dB Sensitivity but 4 Ohms is 87dB using the Efficiency spec. This is because 2.83 Volts into 4 Ohms is 2 Watts -double the power is the 3dB difference.

Steve Lefkowicz Something to keep in mind if you are using tube amplifiers: When you use the 4 Ohm tap as opposed to the 8 Ohm tap, to allow you to drive a 4 Ohm load, the output transformer will be less efficient. It will run hotter (which is your tube output section heating up the transformer- not the best use of power) and you may lose as much as an octave of bandwidth from the bottom end. Driving 2 Ohms (if the transformer even has such a tap) will exacerbate this problem! Of course the amplifier will be higher distortion as I mentioned earlier.

You can measure (and IME, hear) the increase in distortion when any amp is driving a lower impedance. On this account as an amplifier designer, I disagree with Tekton on this issue; if a speaker designer could increase the impedance of his speaker without making any other changes, the immediate result with almost any electronics is smoother and more detailed sound, owing entirely to the reduced distortion. Of course with solid state amps this means a reduction in power; with tube amps the power does not change or might go up a bit (with an attendant reduction in the amplifier temperature, always welcome with tube amplifiers...).