|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
62.44.135.7
I'm buying more sensitive speakers for my 2A3/45 Fi monos. I'm concerned that if the increased bandwidth of the new speakers, the increase in their sensitivity might not be enough to lower the distortion and might even leasd to clipping.
Is there any rule of thumb one can follow to determine how large an increase in speaker sensitivity is required (in relative or absolute terms) to offset the increase in bandwidth of the speakers?
Observe, before you think
Follow Ups:
Paul Joppa did a very nice job of explaining some things and tying it together.
I would add, just for clarification, that your amp is already amplifying the signal that's sent to it, regardless of whether your speakers are able to turn that signal into sound rather than heat. "Greater bandwidth" - meaning better reproduction of low frequencies and high frequencies, will turn more of that signal into sound rather than heat.
Given the very low output of your amps, higher efficiency speakers will help, maybe even noticeably.
With regard to the distortions, can you describe the distortions you're hearing? Do they seem to be amplifier "soft clipping", or something else? Here again, the Joppster is right on the mark. Tubes clip softly, almost like speakers do their compression thingy, and analog tape going into saturation.
There is nothing you can do about hum, except to change your amp or try to eliminate a ground loop or grounding problem. I am not an expert on circuitry design, but I wonder if improving the power supply would increase your S/N. Still, I'd look at grounding issues first, and then address improvements to the amp or other upstream components.
Horn loading
My 2A3/45 amp (Fi) plays big orchestral music loud on my closed box 90dB 2-way standmounts. I'm actually happy with the bandwidth of these speakers. I just want to get rid of the distortion/improve the micro- and macro-dynamics on this type of music without a highly sensitive hum amplifying speaker. But bandwidth is never the same on 2 different speakers, if it's even really advertised with any accuracy.
Observe, before you think
Hi,
OK.
Start by reading this thread (below), and associated threads, to get the driver themselves clearer sounding and lower in various distortions.
Bud Purvine's enABL speaker cone treatment seems just like it will do what you seek, easily.
Jeff Medwin
Bandwidth is also not the same as distortion.
We may have a communications issue, since so many terms have very narrow and precise technical meanings that do not match all that well with the broader meanings in non-technical talk. Rather that start a dictionary, I'll try to address the distortion issue:
The distortion you hear with large-scale music, on 90dB speakers, with a 2- or 3-watt amp, is most likely clipping - the amp cannot produce the instantaneous transient pressure peaks, Being a SET with no feedback, the clipping is "soft", which is to say it is gradual and does not produce the nasty sharp-sounding distortion of hard clipping, characteristic of solid state and (to a lesser degree) high-feedback tube amps.
To alleviate this problem, you must use speakers of greater efficiency, or an amp of greater (peak) power capability. As you noted, speakers of greater efficiency will reproduce any hum of other noises more efficiently as well. An amplifier of greater power output will solve the problem, but only if it does NOT produce more hum/noise.
The technical term here is "dynamic range" - the difference between the residual hum/noise and the maximum power before clipping. For example, consider a typical 2A3 amplifier with AC heated filaments and no feedback. It might produce 0.002 volts of hum; that's 0.0000005 watts into an 8 ohms speaker. The same amp will produce typically 3.5 watts at the onset of clipping. The ratio between those, expressed in dB, is 68.5dB. That's the dynamic range of the amp. With speakers at 90dB, it will produce 95.5dB of sound, and with no signal the hum will be about 27dB.
Now most audiophiles want a minimum of 102dB peaks, so if you are among them you'd need 96dB speakers. The hum level in your room with the same amp would then be 33dB. Whether you can hear that hum depends a lot on how quiet your listening room is.
This IS the high efficiency forum; be aware that many people here feel that an increase above the 102dB I mentioned above is well worthwhile.
Hope that helps.
Apparently, the bandwidth issue was a non-sequitor, or may apply just to the speaker changes I was discussing with the mfr. I've asked him to clarify.
If it's clipping, it's very, very soft, even with the 45s. The speakers I have are very easy to drive in the frequency domain of its bandwidth and my room is small 12m^2. I do have larger amps - 8W and 12W, and the differences I hear are not major on, for example, Sibelius' 4th.
Fortunately, I can audition the speakers, but another concern is they will light up my room.
Thanks again.
Observe, before you think
Subjectively a wider bandwidth speaker will sound louder playing abroad band signal like music at the same power as a narrower one. How we hear loudness is not the same as how we measure it.
Each time you raise he sensitivity by 3dB, you have increased the loudness produce with a fixed power. If you increase it 10dB, you have had an effect similar to making your amp 10X more powerful.
What makes all this more difficult is that when we use a sound level meter or make a judgment about loudness, we are dealing with a short term average level and we don’t hear peaks as the loudness.
Music may require short term peaks 10, 100 or even 1000 times the average level in order to reproduce the dynamic components so very often a high sensitivity speaker will sound more dynamic than a low efficiency speaker.
We can’t hear very short clipping as a flaw but it does take away from the dynamics (when present).
The loudness wars are a way to get more average level out of a power / loudness limited system, the peaks are squashed, the average level increased and the quiet parts made much louder.
The speaker bandwidth has no effect on the amplifier's ability to deliver power. If the speaker sensitivity/efficiency is the same then the same amp will produce the same sound level in the midrange; a wider bandwidth speaker will just produce more sound level at the frequency extreme(s).
How does that square with what djk says?
Observe, before you think
It's directly related to the size.
To keep the efficiency the same and increase the low frequency an octave requires an increase in size of 8x.
To keep the bandwidth the same and increase the efficiency by 3dB require the size to increase by 2x.
Efficiency, size, low frequency bandwidth; pick two and the other is fixed.
Help me here. Size of what, measured in what? Power(W)? Driver size(?).
Observe, before you think
But you're mistaking bandwidth as a key driver of amplifier power. The speaker that hits 80Hz only doesn't consume any less power at 20Hz than the one that hits 40Hz (assuming equal sensitivity). It just wastes more power heating the voicecoil and moving the cone for no meaningful output.
You need to pull "bandwidth" out of the equation. Hoffman's iron law, as djk says, efficiency, size, bass, must be balanced, but the speaker with less bass doens't demand less from the amp unless it's highpassed, which is exceedingly rare.
Bass is supposed to sound big. 6.5" is not a woofer size.
The speakers I am looking at have a bigger box by far (and do have more bandwidth as an incidental factor).
Observe, before you think
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: