Speaker sensitivity question...

[greggorypeccary]

I have a question about speaker efficiency. I understand that using a more efficient speaker can increase your volume. I'm wondering who here has replaced a speaker in a low-watt (say 25 watts or less) combo and what the results were. I'm not bringing up speaker brands or types of music here because I'm just interested in some real-world, anecdotal evidence to support (or refute) the theories and general wisdom with respect to volume. Is the difference really noticeable or just a little? Thanks for your input!

[Tremo]

Depends! If you're going from a fairly inefficient speaker to a very efficient speaker, the difference will be considerable. Perhaps 6 dB, that's like 4x the power. If you put a real efficient speaker in a DR, for example, they can become rather loud.

[EMan]

Speakers are generally rated in Sound Pressure Level, or SPL. A "perfect" speaker that could convert 100% of the power applied to it int sound would have a SPL of 112dbP (the "P" is reference to Pascals).

Every 10dbP increment less than 112dbP is a effiency reduction of a factor of 10. Most research submits that a 10dbP increase, or decrease in SPL equates to about a halving or doubling of perceived volume.

So, a 102dbP SPL speaker would be 10% efficient at converting electrical energy to sound, 92dbP SPL is 1% efficient.

To your question - as Tremo said, it depends on how efficient the speaker is that you want to replace.

I replaced a 12" OEM Oxford in my '72 DR with a brand new Celestion V30, and at least doubled the volume.

One last point is that not all speakers are rated SPL using the same technique, and YMMV.

[greggorypeccary]

Originally Posted by EMan

One last point is that not all speakers are rated SPL using the same technique, and YMMV.

That's what I feared. So comparing specs from different manufactures may be pointless, or not, who knows?

[Dunxy]

I have done a lot of experimenting with different speakers.I have never looked into SPL,even if i did, it would be impossible for me to garner ratings for some of my speakers as they are mostly 50+ years old.

I'm only passing on my perceptions here, not saying anybody it right or wrong.

My main box is a 4 x12 loaded with 4 x "goodmans" alnico speakers.I have 2 differant pairs in it, one pair are labled at a whopping 10w rms,the others are slighlty higher rated, but not labeled, i guess 15-20w. So my whole cab would be say 50-60w rms tops, as for efficiency or SPL, i have no idea. I have tried my 2 main amps (40w and 70w) with many other cabs, greenbacks,v30 and other higher rated Celestion ceramic offerings. With the Celestion loaded quads,I find a large loss in overall gain, and when the amp's are not pushed to the max, a loss in responsiveness and volume.This is really apparant on the 40w amp,its almost useless in 15w triode mode through any cab over 70w rms.

I allways put the differances down to close speaker matching rather than speaker efficiency or SPL but after reading this thread i am not so sure..

Love to hear what you think about my experiences.

[Dacious]

You can compare speaker frequency plots to get some idea of where the speakers' response is.

Celestion Blue.

The famous 'Blues' as in the Vox AC30 have a strong upper-mid spike probably about where your guitar high-E string is. So it is going to be a shrieker-speaker with tons of top end and 115db efficiency at over 1khz.

Celestion G65

The G65 has a more even response, a little jump around the upper mid but mostly around 97-98 db from 100 hz which is about fret 2-4 on your low E to around 5khz where it starts dropping off. That is a fairly even plot which would provide a relatively even response depending on the amp.

4-5khz is about as high as your electric guitar makes meaningful sound. Voices or acoustic instruments can get to higher than that. Unless you're a dog you probbly can't hear past 10Khz.... The reason it drops off at top and bottom is because speaker impedance varies with frequency from say 8 ohm minimum at around the middle of its' designed response to near-infinity at either end of the scale.

Speaker plots are important - speaker response and SPL at different freqs need to be considered. Like your 12v car stereo that puts out a claimed 400 watts - but only at precisely 2,552khz and 30 watts everywhere else. It will not sound 400 watts loud.

Your guitar amp similarly will put out different power at different freqs.

[David Barnett]

Originally Posted by Dacious

You can compare speaker frequency plots to get some idea of where the speakers' response is.

Celestion Blue.

The famous 'Blues' as in the Vox AC30 have a strong upper-mid spike probably about where your guitar high-E string is. So it is going to be a shrieker-speaker with tons of top end and 115db efficiency at over 1khz.

Celestion G65

The G65 has a more even response, a little jump around the upper mid but mostly around 97-98 db from 100 hz which is about fret 2-4 on your low E to around 5khz where it starts dropping off. That is a fairly even plot which would provide a relatively even response depending on the amp.

4-5khz is about as high as your electric guitar makes meaningful sound. Voices or acoustic instruments can get to higher than that. Unless you're a dog you probbly can't hear past 10Khz.... The reason it drops off at top and bottom is because speaker impedance varies with frequency from say 8 ohm minimum at around the middle of its' designed response to near-infinity at either end of the scale.

Speaker plots are important - speaker response and SPL at different freqs need to be considered. Like your 12v car stereo that puts out a claimed 400 watts - but only at precisely 2,552khz and 30 watts everywhere else. It will not sound 400 watts loud.

Your guitar amp similarly will put out different power at different freqs.

Comparing those two curves the same trends are visible -- the main response of the cone, the beginnings of the voice-coil inductance rolloff above 1K, then the response from the dustcap. The 65 seems to have a slightly more even response below the inductive rolloff, but its dustcap response looks raggedy as all heck. I would expect that speaker's top end to be a little "ratty". The Blue has a gentle rise across its main response, a stronger dustcap resonance, but the dustcap response seems more even, leading me to believe that it will have lots of top but it'll be "sweeter" up there.

[greggorypeccary]

Wow, great info guys! David, you're explaination of the plot has helped a lot, I've always wondered why there is always a dip around 1k, now I know. Not that a graph can really tell what it will really sound like but it helps when comparing different speakers. Then again, I'm kind of strange, I like graphs!

[Dunxy]

Great info Dacious and David.

So comparing those two graph's you have posted, the Blue has the higher SPL,does that make it the louder speaker?

Why do you think it has a higher SPL than the g65? Is it the different power handling,magnet material ,voice coil dia,magnet strength or cone thickness, or a combination of all those variables?

[David Barnett]

Originally Posted by Dunxy

Great info Dacious and David.

So comparing those two graph's you have posted, the Blue has the higher SPL,does that make it the louder speaker?

Why do you think it has a higher SPL than the g65? Is it the different power handling,magnet material ,voice coil dia,magnet strength or cone thickness, or a combination of all those variables?

The 65 will go a lot louder than the Blue Alnico, because it handles much more power. But at the same power, the Blue would be louder. I mean, if you set an amplifier up that was putting out exactly 10 volts and hooked up both speakers, assuming both are 8-ohm speakers, more sound would appear to come from the Blue. If you set up an amp that is putting out exactly 25 volts continuosly, the 65 will be louder, because the Blue Alnico would fail in a few seconds.

[Dunxy]

Originally Posted by David Barnett

The 65 will go a lot louder than the Blue Alnico, because it handles much more power. But at the same power, the Blue would be louder. I mean, if you set an amplifier up that was putting out exactly 10 volts and hooked up both speakers, assuming both are 8-ohm speakers, more sound would appear to come from the Blue. If you set up an amp that is putting out exactly 25 volts continuosly, the 65 will be louder, because the Blue Alnico would fail in a few seconds.

Ahh thats pretty well what thought anyway,just confirms what i have experienced when using the higher power rated cabs.

So in short, if you have per example a 15w rms amp, it will be louder with a 15w rated bluedog than say any higher rated speaker, correct?

[Roli]

Dacious, David - you both know your stuff, very interesting comments. I learned a lot, thank you!

What's that peak just above 30 Hz in the 65's curve? Is it audible at all when driven by an amp?

[Dacious]

Tight voicecoil gap 'just right' dimensions and a flat cone are why the 'Blue' is the 'Blue'. Power handling is 15 watts RMS or probably 20 peak watts. In that range not much is going to live with it, but it won't kill on bass - not it's mission. Celestion produced other products with better low-freq handling like the 'Silver' 25 watts used for AC50s.

The Tone Lizard

[EMan]

"So in short, if you have per example a 15w rms amp, it will be louder with a 15w rated bluedog than say any higher rated speaker, correct?"

Power handling doesn't necessarily indicate a speaker's loudness. For one thing, speaker power ratings can be derived in different ways. Some use VC burnou, and others use actual cone/spider/surround damage.

SPL plots are "supposed" to be run at 1WRMS/1 Meter. That levels the playing field. Later, it was decided that the 1W ought to maintained across the frequency range of the test, and they change the number to something like 3.24VRMS. Seems like this is just a P = E^2/R thing.

Since you like graphs, here's a 'typical' Impedance plot for a speaker, courtesy of Weber VST - http://www.webervst.com/pics/zplot.gif

Further, I reckon you could look at your amp. What kind of tone circuit does it have? Typically, even with the tone controls on 10, there's a pronounced 'mid cut'. How would this affect the SPL 'peak' of a specific speaker.

You amp's damping factor (associated with output impedance) has something to do with it as well. Tube amps generally have miserable damping factors as compared to SS output sections. Tube amps typically run around 3 - 5 ohms, compared to SS that's closer to 10ths of ohms ( output impedance ).

I guess I'd have to re-state my opinion, and that would be that you really have to just try a speaker on your own rig for an accuate assessment (as mentioned above), and use the numbers as general guidlines.

[Drak]

Like E-Man, I had an old Valco amp from the early '50's that came with the original speaker, very small voice coil (1" I believe it was).

I replaced it with a Weber P12N and the volume -at least- doubled, if not more, it was a HUGE difference in volume, and gave a great voice to a great amp that was seriously being underrepresented by it's stock speaker.

[David Barnett]

Originally Posted by Roli

Dacious, David - you both know your stuff, very interesting comments. I learned a lot, thank you!

What's that peak just above 30 Hz in the 65's curve? Is it audible at all when driven by an amp?

That's a good question. I don't know the free air resonance of that model, but would imagine it's higher than that.

[Dacious]

The spike around 30hz would not be audible - you're right into rolled-off territory. The 65 has a free resonance of 50hz IIRC, below that you've pushed it past its' limit. Most guitar speakers are 'full-range' which means 50-5Khz approx. Your lowest note unless you have a baritone is going to be around 70-80hz, and highest about 4Khz unless it's a mandolin.

E-man is exactly correct about preamp design - all the low-end speaker capability in the world will not bring bottom end to an amp if it has a low-end rolloff designed into it i.e. if the tone capacitors cut off low freqencies.

Some guitar amp preamps, to avoid the output transformers/speakers farting out, keep the power section alive or just to avoid unpleasant noises do the same thing. I have little Diason 15 watt 6GW8 amp with a 12" speaker - 6GW8 is like a EL84 with half a 12AX7 in it. It sounds very like a Vox AC15. Replacing the whimpy stock speaker makes almost no difference to low-end response, because the preamp kills almost everything down low, even with bass turned up. The 6GW8s just don't produce a big bottom, and trying for it makes the amp fart out. If you persist trying you can get low-end clipping and melt something. Needs more power and more substantial output transformer iron.

Same as the .047, .033 or .022 tone cap on your guitar - it's there to keep the flatulence out.

[Dunxy]

All interesting stuff...

The conditions i described in my original post were not just observed on 1 amp,3 different amps all with different pre-amps and slightly different power sections.They all sounded better driving my cab vs a 100w+rms jobbie.

I know what you meen about the small pentodes farting out when driven hard, i also have a little 6gw8 amp so am more than familier with them.When i did my comparos, i used real amps.2 were kt-66 one was el-34.

I have even done an a/b between a Greenback and a really nice Aussie blue dog clone in identical 1 x 12's with a stereo line 6.So I was running both speakers at the same time with the same amp, any advantages the greenback had were only apparant when you where driving harder than you would ever drive the Blue dog/clone.When things were kept sane,IE no speaker distortion,the blue dog clone had it in spades, the greenback sounded like it had a wet blanket over the top in comparison.I also did the above test with a v30 in place of the greenback,results where not the same but close, i still favour my alnicos anyday.

For the record i was offered in trade 4 vintage greenbacks for the one alnico speaker i used in the above tests.I spent time trying to justify the trade to myself, 4 speakers for 1! But i couldn't bring myself to do it.

[David Barnett]

Originally Posted by Dacious

The spike around 30hz would not be audible - you're right into rolled-off territory. The 65 has a free resonance of 50hz IIRC, below that you've pushed it past its' limit. Most guitar speakers are 'full-range' which means 50-5Khz approx. Your lowest note unless you have a baritone is going to be around 70-80hz, and highest about 4Khz unless it's a mandolin.

E-man is exactly correct about preamp design - all the low-end speaker capability in the world will not bring bottom end to an amp if it has a low-end rolloff designed into it i.e. if the tone capacitors cut off low freqencies.

Some guitar amp preamps, to avoid the output transformers/speakers farting out, keep the power section alive or just to avoid unpleasant noises do the same thing. I have little Diason 15 watt 6GW8 amp with a 12" speaker - 6GW8 is like a EL84 with half a 12AX7 in it. It sounds very like a Vox AC15. Replacing the whimpy stock speaker makes almost no difference to low-end response, because the preamp kills almost everything down low, even with bass turned up. The 6GW8s just don't produce a big bottom, and trying for it makes the amp fart out. If you persist trying you can get low-end clipping and melt something. Needs more power and more substantial output transformer iron.

Besides, the low E on a guitar in standard tuning is about 82Hz.

Originally Posted by Dacious

Same as the .047, .033 or .022 tone cap on your guitar - it's there to keep the flatulence out.

The tone cap in a guitar is a low-pass filter, not a high-pass. It kills the highs when you turn the tone control all the way down.

[Roli]

Originally Posted by David Barnett

That's a good question. I don't know the free air resonance of that model, but would imagine it's higher than that.

Originally Posted by Dacious

The spike around 30hz would not be audible - you're right into rolled-off territory. The 65 has a free resonance of 50hz IIRC, below that you've pushed it past its' limit. Most guitar speakers are 'full-range' which means 50-5Khz approx. Your lowest note unless you have a baritone is going to be around 70-80hz, and highest about 4Khz unless it's a mandolin.

Thank you! Gotta love this stuff.

[nonvintage]

Greggory, Here's what I found out! My 1956 Gibson GA 30 is a tweed sounding amp with 5y3 rectifier 2 6v6's and 3 preamp tubes 2 12ax7's and for the phase inverter a 12au7. I've tried 4 different speakers in this amp. Here's my story for the differences in these speakers. An alnico Weber 12A125S 30 watt, a reissue Jensen P12R 25 watt, an Eminence ceramic taken from a Fender Deluxe Reverb reissue and a Eminence Legend Alnico 122P 20 watt speaker. The weber has a big fat low end with medium smooth break up. The Jensen has a lot less bass and is crunchy/ raunchy on the top end. The Eminence from the DRRI was maybe the loudest but didn't have the bass of the Weber or the top end of the Jensen. The Eminence Legend 122P had almost as good bass as the Weber and the nastiest tweedy tone overall of all these amps. The sensitivity SPL of the Weber is not published, the Jensen is rated at 93.6 SPL, the Emi from the DRRI is SPL 99 and the Emi Legend 122P has a SPL of 97. The amps loudness was pretty much the same with each speaker. Nothing really blew me away in that regard. But my favorites is pretty obvious. #1 Emin Legend 122P, #2 Weber 12A125S ( I think that a 12A125 would have been better.), #3 Jensen P12R (the P12Q would have been better.) #4 The DRRI Emi Legend 1258. IMHO YMMV YADDA YADDA!

[telbert twang]

How much the efficiency measuring may vary from manufacturer to manufacturer ? Sometimes one´s 95 may be another´s 100 db ?

How important is the cone´s linear movement ability (i mean, some are only 0.5 mm, some are 1.5mm...) ?

In 22W Deluxe Reverb , would the 101db spl Emi Screamin Eagle be noticeably louder than a Jensen MOD-12-70 ?

[telbert twang]

bump

[vanguard]

when i went from a 96 db p10r to a 102 db red fang, the volume increase was mind-blowing. my amp sounded at least 3 times as loud. then i went down to a 98 db gold g10, and it seems a bit less than twice as loud.

i read somewhere that every 2 db's in sens. will result in a perceived doubling in volume. this has held true in my experience.

[JD0x0]

Originally Posted by vanguard

when i went from a 96 db p10r to a 102 db red fang, the volume increase was mind-blowing. my amp sounded at least 3 times as loud. then i went down to a 98 db gold g10, and it seems a bit less than twice as loud.

i read somewhere that every 2 db's in sens. will result in a perceived doubling in volume. this has held true in my experience.

not true 2db is barely noticable, some people may not even hear the difference of 2db as 3db is the average audible change. going from a 50 to 100 watt amp is about the same difference, 2-3db. A perceived "double in volume" is about 6db which you would usually need 4 times the power to achive, which is why people often just get more efficient speakers.