You didn’t read the paper, did you? Beyond a certain amount of capacitance for a given power supply rating the additional capacitance is not only not useful, but larger capacitors actually have increased ripple.
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
HOW can bigger amp change sound at lower volume?
- Thread starter SPOautos
- Start date
You didn’t read the paper, did you? Beyond a certain amount of capacitance for a given power supply rating the additional capacitance is not only not useful, but larger capacitors actually have increased ripple.
That's not a myth?!
The caps might be there to Smooth out ripple but they also capable of delivering way higher peak currents than the transformer
(in most cases)
So for Dynamic Music its totally possible to undersized the power supply and oversize capacitors.
But they kind of doping this.
After Rectification ans Smoothing the DC RMS voltage is about 1.4x higher and depending on the load (and capacitor size) it drops.
How low it drops Limits how much power (volt) the amp Can output before distortion occurs.
So with bigger caps the minimal as well as the RMS DC voltage is higer and the Amp can output more power before clipping
Did you understand the Paper?!
Last edited:
In between the peaks output from the transformer, that is exactly what they are doing. In between those peaks, the transformer is supplying lower voltage than the capacitors. There is no energy coming from the transformer at all.
No, I said you said the other guy said it. Got it?
We're done.
And here is an illustration (simplified) of how they maintain peak power. In the graph,the grey lines are the negative and positive rectified voltages out of the transformer. Yellow and Red are the DC+ and DC- rails. Green is an output voltage wafeform needed to drive the speakers at full power. In the first chart you can see that the minimum ripple voltage is just high enough to support the peak power without clipping.
Now see what happens if the capacitance is halved: Some of the peaks of the output waveform will be clipped:
And halved again, clipping even worse - peak power even more limited:
Now see what happens if the capacitance is halved: Some of the peaks of the output waveform will be clipped:
And halved again, clipping even worse - peak power even more limited:
Last edited:
We're done too.
But back to Topic:
If the Amplifier is well made and the peak output power and voltage less then what it can reliably delver There should be no change in sound.
But In general for non Ideal Amps. Amps with more power have less output impedance and some Amps have increased Compression/distortion Way before the reach the claimed output power.
So overrating the amp might lower distortion and increase damping.
But this can't be said as a general statement for every amp.
If the Amplifier is well made and the peak output power and voltage less then what it can reliably delver There should be no change in sound.
But In general for non Ideal Amps. Amps with more power have less output impedance and some Amps have increased Compression/distortion Way before the reach the claimed output power.
So overrating the amp might lower distortion and increase damping.
But this can't be said as a general statement for every amp.
Sure
So that's a clear NO
I have a fair understanding of exponentials, logarithms, and published loudness studies.
- 1 dB is barely perceptible if you tweak the volume that much and requires 1.26x the power; it is about 26% increase in power but a very minor change in loudness. Many volume controls these days work in 1 dB steps as that is about the minimum people can tell; try flipping the knob 1 dB back and forth and see how much the volume changes. For me, very little, but opinions vary.
- 3 dB is about the change most people perceive as "turn it up/down just a little" and requires 2x the power.
- 6 dB is a significant increase in loudness and requires 4x the power.
- 10 dB is twice as loud (in the midrange; equal-loudness curves are more compressed in the lower and very highest frequencies) and requires 10x the power.
And if I'm not mistaken, peaks of 10 dB above the continuous program are common for dynamics, thus the need for at least 10x power headroom to avoid compression, if not audible distortion. Correct?
This Depends on your music but you can easily see this in Audacity or something like this.
I would just make sure that 0dBFS from my DAC don’t Clip my Amp
for example this track with peaks at ~0dBFS has a -11dB RMS loudness
If both the big & small amps are competent (low output impedance, distortion & noise, and flat frequency response, adequate power handling) they should sound identical when played at identical volume through the same speaker, so long as the smaller amp is not asked to do more than it can power-wise.
Geert
Major Contributor
- Joined
- Mar 20, 2020
- Messages
- 1,944
- Likes
- 3,548
Your theory and graphs are correct, but also a bit misleading. As long as you're not driving the amp into maximum output level there's no issue. This is pretty relevant since this topic concerns amplifier sound at low volumes.
Also related but never mentioned is that a hughly oversized capacity bank leads to higher peak currents because energy needs to be replenished in a shorter time span, exactly because of the lower DC ripple. So a linear power supply, and by extention a power amplifier, needs to be right sized.
My memory from decades ago is of research compiled into an AES paper I can no longer access (let my membership lapse since my career took a different path). The result was the peak-to-average power ratio for music was on average 17 dB, a factor of 50x in power. Modern results I have seen online point to far less than that for highly-compressed music, but about the same for well-recorded "dynamic" music. Movies have been shown (various threads mainly on AVS and other anecdotal evidence reported by folk using various SW tools) to exhibit 20 dB to as much as 30 dB, power factors of 100x and 1000x respectively, but since most of the highest peaks were for things like gun shots and explosions I am of the mindset that I can stand a little clipping for those scenarios.
It's been my experience that people tend to greatly overestimate their average power needs and underestimate the peak needs, though it more or less works out. People who think they need 100 W and discover they are only using 1 W on average, have the headroom they need (20 dB) for almost everything with that 100 W amp.
At least in the midrange; more power is generally used in the deep bass so you can "feel" it.
And people tend to neglect the issue of whether the speakers can handle the power, and how much distortion they generate.
For me personally, 80 dB is very loud, so creating a system that can meet the THX etc. guideline of 105 dB per speaker (and 115 dB for subs) is more (way more) than enough dynamic range for me. I tend to focus on the lower end of the scale by building a well-isolated media room that is nice and quiet rather than getting enough speakers and power amps to liquify my body.
All IME/IMO/FWIWFM/YNNV/etc. - my 0.000001 cent (microcent) - Don
This is true to some extend for conventional transformers.
Of Cause "there is no free lunch" the Peak Riple current must go up even from the perspective of conservation of energy
But RMS current should ruffly stay the same.
On the other hand non ideal Transformers have primary and secondary impedance as well as non ideal coupling. limiting the maximum current they can output.
Some Power supply's even have inductors in series with the transformer primary or transformers with a magnetic shunt to limit the peak current and Increase power factor.
For SMPS this is a different story.
In high Power applications theyoften have a Current Limited operating range, a power limited operating rage and a Voltage limited operating range.
Or at least an CC and CV rang or "virtual" output impedance.
For Most Class D Amps a Drop/ripple on the DC rails is not relay bad it matter more how low the voltage drops. "so over charging" the capacitors to a higher voltage makes them store more energy.
But not rigidly regulating the DC Rails keeps strain and peak current away from the power supply.
All this is irrelevant to music at low volume far from the Amps limits.
egellings says:
- Thread Starter
- #36
Well I'm glad Im not just a idiot that doesnt understand....apparently there are a lot of varying ideas on the subject. Many of which are over my head lol
So my Yamaha MX600U (135wpc) has meters and I never see them go over around 20wpc at my normal listening level (50db-60db). I have Dahlquist dq10 speakers which are notoriously inefficient and when testing frequencies at 80hz and below the Db of each frequency is 10Db less than all other frequencies. So while bass is there is quieter. This is NOT due to my amp but rather just speaker design.
I've been told by pretty much everyone 250wpc amp will help the bass a lot and open up the imaging even more....even at the same low listening levels. The imaging is actually already great. I think I'd be better off, rather than buying a new amp, to instead spend that money on 2 powered subs and just keep the volume low where they barely supplement and help out a bit, especially at real low frequencies.
I appreciate everyones views and sharing.
So my Yamaha MX600U (135wpc) has meters and I never see them go over around 20wpc at my normal listening level (50db-60db). I have Dahlquist dq10 speakers which are notoriously inefficient and when testing frequencies at 80hz and below the Db of each frequency is 10Db less than all other frequencies. So while bass is there is quieter. This is NOT due to my amp but rather just speaker design.
I've been told by pretty much everyone 250wpc amp will help the bass a lot and open up the imaging even more....even at the same low listening levels. The imaging is actually already great. I think I'd be better off, rather than buying a new amp, to instead spend that money on 2 powered subs and just keep the volume low where they barely supplement and help out a bit, especially at real low frequencies.
I appreciate everyones views and sharing.
You hit the nail on the head with the idea of buying a couple of subs. You'll be glad you came here and asked your question, probably more than you realize. It's also easier for your speakers to not have to play the low bass frequencies and you will notice your midrange has improved most likely.
lol. i went the opposite direction. (last paragraph)My memory from decades ago is of research compiled into an AES paper I can no longer access (let my membership lapse since my career took a different path). The result was the peak-to-average power ratio for music was on average 17 dB, a factor of 50x in power. Modern results I have seen online point to far less than that for highly-compressed music, but about the same for well-recorded "dynamic" music. Movies have been shown (various threads mainly on AVS and other anecdotal evidence reported by folk using various SW tools) to exhibit 20 dB to as much as 30 dB, power factors of 100x and 1000x respectively, but since most of the highest peaks were for things like gun shots and explosions I am of the mindset that I can stand a little clipping for those scenarios.
It's been my experience that people tend to greatly overestimate their average power needs and underestimate the peak needs, though it more or less works out. People who think they need 100 W and discover they are only using 1 W on average, have the headroom they need (20 dB) for almost everything with that 100 W amp.
At least in the midrange; more power is generally used in the deep bass so you can "feel" it.
And people tend to neglect the issue of whether the speakers can handle the power, and how much distortion they generate.
For me personally, 80 dB is very loud, so creating a system that can meet the THX etc. guideline of 105 dB per speaker (and 115 dB for subs) is more (way more) than enough dynamic range for me. I tend to focus on the lower end of the scale by building a well-isolated media room that is nice and quiet rather than getting enough speakers and power amps to liquify my body.
All IME/IMO/FWIWFM/YNNV/etc. - my 0.000001 cent (microcent) - Don
imho, more power makes a difference. more power(current for bass)available on transient peaks.
feels that way to me but what do i know.
adding a second sub did increase my low volume enjoyment. unbelievable to hear clear, even defined bass at -60db on MY volume indicator.(different on all systems i know)which is low enough that i can only hear it when ambient volume is low. like early in the morning or late at night.
Similar threads
