An Amp Sonics side-thread

Split from here http://www.diyaudio.com/forums/showthread.php?s=&threadid=114521

R-Carpenter said:
[Don't think the amp should be a consideration for a choice of a driver.

The speaker should never be considered without knowing what kind of amp you will be using. The amp (or more correctly the amp given the load provided by the speaker) can make a huge difference in how a system sounds.

After all amp (with enough power) will make little or no difference in the outcome.

False. Very false.

dave
 

Brett

Member
2002-01-07 6:02 pm
MaVo said:


In which way does a different amp affect the sound? I suppose you are not speaking about noise levels or the harmonic distortion of tube amps.
Big question, but here is a short answer; you need to puzzle the details yourself.

THD: level, spectrum and how the spectrum changes with level. Also class; AB has worse distortion with decreasing level.
Noise: irrelevant in any well designed and implemented design
NFB; use of or not, level of NFB and how it's implemented
Output Z: can be of use if factored into some designs. Needs to be minimised in others.
Balance in PP CCTs: needs to be maintained under dynamic conditions. How you achieve that will sound different
Clipping behaviour: Amps should never clip, but when they do, how do they behave and how long before they settle.
Component quality: I find it better to design assuming there'll be as little NFB as possible.
Design: Linearity, linearity, linearity. And headroom
Power Supply: it's in series with the amp so make it as inaudible as possible
+ a myriad of other little build details.
 

vasyachkin

Disabled Account
2007-12-31 1:16 am
when it comes to amps i think that all good ones are the same, but not all "hi-fi" amps are good ... in fact most probably aren't.

if you have a solid amp that is flat 10hz - 50khz with 0.01THD, 2 ohm stable and with 10 times the power that you actually need then i think it should be transparent in the system.

such amp does not have to cost an arm and a leg. my QSC Audio PLX seems to be just fine and it costs about 35 cents per watt.

i had some other amps that were simply useless though (a Technics from around 1997 and a NAD probably from around same time). people would have me believe that the NAD will sound good even though its specs are weak simply because its british ... not so. apparently brute force still counts for something. apparently being british cannot make up for having 10 times the distortion, 10 times the noise, 10 times the output impedance, half the bandwidth and 1/10th the power.

i think in practical terms there are 3 problems with powering speakers:

1 - PA amps have loud fans
2 - PA amps are too big to run tweeters actively off of
3 - i like PA amps :)
 

vasyachkin

Disabled Account
2007-12-31 1:16 am
Brett said:
Clipping behaviour: Amps should never clip, but when they do, how do they behave and how long before they settle.

in a commercial application it only makes sense to run the amps full scale, but at home i think we can avoid clipping altogether by simply using a bigger amp. if the amp is something like class H there will not be much penalty in electrical bills.

i would love to use a class A amp in theory, but in practice no serious company (like Crown) will bother producing a class A amp because of how impractical it is. in practice i would rather use a class H amp of an excellent design and low cost than a class A amp of a poor design and high cost.

of course if you want to design and build your own class A amp then clipping will be a concern.
 
Brett said:
True. Very true.


I knew I was going to strike someone’s nerve with “Amp makes no difference” remark.

Unless we are talking about hopeless 1-watt class A tube design, I stand by my point. 99% of solid-state amps out there, given the fact that isn’t driven to clipping (as I mentioned before, “having enough power”) aren’t very much distinguishable in sound. Blind test have proven it number of times. I have been through about 8 blind tests, listening to the amps ranged in the price between $300 and close to $7000. No audible difference was found (5 more people participated in the tests, tests were done on 2 pairs of commercial speakers, ranged between $4500 and $5000 in price). Does it mean they will measure the same? Not at all but our hearing is much less sensitive then an oscilloscope. “Can’t you hear that?!” becomes useless with equipment covered up.
I have 1500 square foot shop and a SPL meter. If you want to prove me wrong, bring it. I will be first one to say, “I was off beam”.

I am not even going to go into solid state vs. tube argument but the point is that Speaker system makes the biggest impact on the sound reproduction! Why? Because unfortunately, there’s no ideal speaker system out there yet. They are all colored and have different frequency ups and downs and of axis response. In fact, room-speaker interaction gives it an infinite number of possibilities in coloration = character of a speaker.

The audible differences between electrostatics, magnetic planar type, horns, open baffle and as we talking about ported vs. sealed enclosures are major. It is first intelligent to consider, which type of coloration you prefer (ported or sealed in this instance) and perhaps think of enough wattage later, based on the efficiency of the driver and the impedance behavior of the speaker.

I always build my speakers with the impedance compensation to make amps life easier and prevent the impedance from dropping down below 3.5-4 ohms and jumping over 20 ohms at cone resonance.
 

tinitus

diyAudio Moderator R.I.P.
2005-11-24 1:47 am
I have heard clear differense between good amps and VERY good amps, although very small differense its enough to make it very important...if quality matters more than quantity

No matter how well a speaker is designed, a good amp may in many cases be the very key to nirvana of sound

But some speakers demand fore more stable amps than others

Sure, you can do a setup where it probably doesnt matter much, but that doesnt make it a general rule

BTW, 5-7000USD worth of amp doesnt automatically mean its any bit better than cheaper ones

Speaker technology has changed rapidly and today its probably much easier to find a good speaker than a good amplifier...but thats only pure speculation
 

vasyachkin

Disabled Account
2007-12-31 1:16 am
Difference between NAD and QSC is like night and day. NAD sounds like FM radio between the stations and yet many people recommend it so its entirely possible to buy an amp that is supposed to be good and wonder why your speakers sound miserable.

just because its easy to make an amp sound transparent doesn't mean that even most amps are made that way.

instead a lot of people don't want to buy amps with op-amps in them (like QSC) so they go for fully discrete designs (like NAD) and in a discrete design there is a temptation to cut corners by throwing out some parts (to save money) and then you're left with an amp that sounds just barely good enough for most people buying it not to notice how bad it is.

if audiophiles could get over their religious beliefs maybe things could start to improve, but i think the 99% figure is way off ... maybe 10%.
 

vasyachkin

Disabled Account
2007-12-31 1:16 am
tinitus said:
Speaker technology has changed rapidly and today its probably much easier to find a good speaker than a good amplifier...but thats only pure speculation [/B]

amplifier technology changed a lot more than speaker technology. after all speakers are copper, aluminum and wood and amps are silicon.

i will venture to say that it will never be easier to find a good speaker than a good amp ... but thats only because speakers are nowhere even close to being "good enough" today.
 
vasyachkin said:
if you have a solid amp that is flat 10hz - 50khz with 0.01THD, 2 ohm stable and with 10 times the power that you actually need then i think it should be transparent in the system.

That is exactly the kind of amplifier that doesn't sound very god on most of my speakers. Damping factor is too high and the efficency of the speakers will have that amp running at its worst.

at home i think we can avoid clipping altogether by simply using a bigger amp.

But that is counter to the general tendency that sound quality is inversly proportional to power output -- hence design of speakers specifically to mate with very high quality 2-10W amplifiers.

dave
 

vasyachkin

Disabled Account
2007-12-31 1:16 am
planet10 said:


That is exactly the kind of amplifier that doesn't sound very god on most of my speakers. Damping factor is too high and the efficency of the speakers will have that amp running at its worst.

But that is counter to the general tendency that sound quality is inversly proportional to power output -- hence design of speakers specifically to mate with very high quality 2-10W amplifiers.

dave

first of all, damping factor can never be too high. ideally it should be ininite.

you do have a point that if you run an AB amp at 0.1% of its power it may have poor distortion and noise performance there ... but at 10% of its power an amp will typically have better performance than at 90%.

as for your last bit the sound quality is only inversely proportional to the power for the same amp. that is to say the same amp will run cleaner at 10% than on 90% but a bigger amp at 10% will run cleaner than a small one at 90%.

yes there is still every reason to build efficient speakers but there is no reason to use 2 watt amps.
 
planet10 said:


That is exactly the kind of amplifier that doesn't sound very god on most of my speakers. Damping factor is too high and the efficency of the speakers will have that amp running at its worst.



But that is counter to the general tendency that sound quality is inversly proportional to power output -- hence design of speakers specifically to mate with very high quality 2-10W amplifiers.

dave


So, what you saying is that a virtually distortion-free amp with gobs of (again distortion-free) power, that is stable in to 2 ohms will make your speakers sound bad?
 
R-Carpenter said:
So, what you saying is that a virtually distortion-free amp with gobs of (again distortion-free) power, that is stable in to 2 ohms will make your speakers sound bad?

How are you measuring distortion? There us no such thing as a distortion-free amplifier. To get gobs of power requires a bunch of sonic compromises... to build a really good high power amplifier costs alot of money (as an example look at the Passlabs amps -- then consider that Nelson still prefers his 5-15W amps for his personal listening)

It isn't that the amp makes the speakers sound bad, it is that they don't work together as a system. My speakers are designed to work with lower output, highish impedance amplifiers (ie simple to build and generally excellent sounding -- when mated with complentary speakers)

dave
 
vasyachkin said:
first of all, damping factor can never be too high. ideally it should be ininite.

That shows your biases and means that you are ignoring a very rich segment of hifi kit. In practise you will never get a damping factor much over 20. You also have to ask how a very high DF is achieved. And you have to ignore some of the best loudspeakers there are. There are very valid reasons for having amplifiers with high output impedance.

the sound quality is only inversely proportional to the power for the same amp

Not at all. In general a small amplifier will sound better than a big amplifier when made with the same level of implementation (actually the low power amp can be built with a bit less attention to detail)

yes there is still every reason to build efficient speakers but there is no reason to use 2 watt amps.

Sure there is. It is called a 45. Many of the very best sounding amps in the world have 1-10W of power, and it is certainly way easier to make a really good low power amp than a big one.

dave
 

AJinFLA

Banned
2005-02-09 4:35 am
Tampa
planet10 said:
To get gobs of power requires a bunch of sonic compromises

Based on what scientific facts?

planet10 said:
That shows your biases and means that you are ignoring a very rich segment of hifi kit.

Is this statement made without any biases?

planet10 said:
And you have to ignore some of the best loudspeakers there are.

In general a small amplifier will sound better than a big amplifier

Many of the very best sounding amps in the world have 1-10W of power

According to who?

cheers,

AJ
 
Dave, it appears to me that you are a hard-core fan of a full range driver speakers. Perhaps there’s something to it. As they say, to each, it’s own.
I am not even going to argue the validity of the claim that less powerful amps sound better then amplifiers with more power. If you like 5W single-ended triodes, god bless you, I think they look really cool.
Fantool is amplifying his speakers for Home Theater, which means that the signal will be very heavily EQ-ed before it gets to his speakers so all voodoo high end 98db efficient speakers aren’t the case here either.
Aside from anything else, my point was that a speaker system and the preference of the owner, secondary the room itself, not the amplifier is the more important links in this chain.
Since you mentioned THD, the question that I am going to elevate now is the audibility of an amplifier distortion. Yes, there’s no distortion-free amplifier, but can you hear it?
:D :D :D
 

Brett

Member
2002-01-07 6:02 pm
vasyachkin said:
in a commercial application it only makes sense to run the amps full scale, but at home i think we can avoid clipping altogether by simply using a bigger amp. if the amp is something like class H there will not be much penalty in electrical bills.
I have my own PA's, all built by me so I understand. Please explain the relevance to domestic reproductive systems.

vasyachkin said:
i would love to use a class A amp in theory, but in practice no serious company (like Crown) will bother producing a class A amp because of how impractical it is. in practice i would rather use a class H amp of an excellent design and low cost than a class A amp of a poor design and high cost.

of course if you want to design and build your own class A amp then clipping will be a concern.
There are many superbly designed and implemented class A amplifiers out there.

Build efficient speakers, then the size and class of the amplification becomes a much smaller issue, with far fewer compromises. By efficient I mean >96dB, preferrably 100dB. Compromise is in size and sometimes cost.
 

Brett

Member
2002-01-07 6:02 pm
R-Carpenter said:
I knew I was going to strike someone’s nerve with “Amp makes no difference” remark.
You seen to be the one upset about it: I simply stated my experiences.

R-Carpenter said:
Unless we are talking about hopeless 1-watt class A tube design, I stand by my point. 99% of solid-state amps out there, given the fact that isn’t driven to clipping (as I mentioned before, “having enough power”) aren’t very much distinguishable in sound. Blind test have proven it number of times. I have been through about 8 blind tests, listening to the amps ranged in the price between $300 and close to $7000. No audible difference was found (5 more people participated in the tests, tests were done on 2 pairs of commercial speakers, ranged between $4500 and $5000 in price). Does it mean they will measure the same? Not at all but our hearing is much less sensitive then an oscilloscope. “Can’t you hear that?!” becomes useless with equipment covered up.
I have 1500 square foot shop and a SPL meter. If you want to prove me wrong, bring it. I will be first one to say, “I was off beam”.
I've done similar. Almost all SS AB amps sound the same to me, whether an old Rotel or yamaha, or the Quest, QSC,or Yamaha high power designs I use in my PA's. Generally, not that great, especially when you're running very high efficiency systems. Distortion is worse at low levels with an AB amp (any type) because nonlinearity increases areound zero. High eff speakers make this much more obvious.

As well as doing tests, I've also designed, built, tested and measured about a hundred amps over the last few years, so I feel I have some insight into why there are differences between them.

R-Carpenter said:
Speaker system makes the biggest impact on the sound reproduction! Why? Because unfortunately, there’s no ideal speaker system out there yet. They are all colored and have different frequency ups and downs and of axis response. In fact, room-speaker interaction gives it an infinite number of possibilities in coloration = character of a speaker.
Little argument there, but I design as a systemic whole. Sometimes amplfiers with a defined O/P Z have a strong benefit overall to the system. Read Nelson's articles on this.

R-Carpenter said:
The audible differences between electrostatics, magnetic planar type, horns, open baffle and as we talking about ported vs. sealed enclosures are major. It is first intelligent to consider, which type of coloration you prefer (ported or sealed in this instance) and perhaps think of enough wattage later, based on the efficiency of the driver and the impedance behavior of the speaker.
Again, little argument generally, except it depends on what frequency band you're talking about. Below 100-150Hz, I couldn't care less what the amp is, so long as it doues the job. For lower eff subs, I just pull out a Quest QA3004 and use that. For midbass and up, I prefer something with less sonic compromise, and I'm prepared to spend some time and effort on building my own amps for this reason: my experience tells me it makes a difference. If it didn't I would quite happily roll the PA rack in and plug it up when it's not working.
 

jdlech

Member
2007-12-25 7:51 pm
I often brought my speakers outside and at partys. The 91db speakers are measured at 1 meter. Sound pressure decreases by the reciprocal of the distance. At 10 meters, the sound volume is down 10db. @ 20 meters, they're down to 71db. My back yard was about 300 feet to the edge of the pond at the time.

My 'home' audio also doubled as a 'pro' setup on rare occasion. And yes, I did 'crank it up 'till their ears bled' on those occasions.

So what does it take to clip a 10W amp connected to 85db speakers? Most popular music has a dynamic range of 6-12db. Let's use 9db for our example. @9db, the average wattage needs to be 8 times less than the 10W maximum, or 1.25W. Anything more clips and distorts. That provides an average of 86db sound pressure.

But that's at just 1 meter. I don't know many people who sit that close to their stereo. Double or even triple that is more likely. So we're talking about 83-80db level maximum before the amplifier starts clipping during the loudest passages.

While that may be plenty adequate for individual listening, you add a dozen people having a good time and it will likely be all but drowned out.

I prefer to have more headroom than I ever need so it never clips.