a "fast" amp is one that has the ability to output "naturally" a quickly rising transient input from recorded sources like piano. while the "quickness" of an amp is affected by characteristics that include bandwidth, the "objective" measurement of "speed" is generally expressed by slew rate (theoretically, the higher the expressed value, the faster the amp). as with damn near everything in audio, tho, your ears are the ultimate measuring device. once you understand what "fast" means, you'll know it when you hear it. |
Cornfed pretty much summed it up. The three terms that summarize a "fast" amp to me would be rise time, slew rate and ultra wide bandwidth. You can't have really wide bandwidth without having pretty solid measurements on the other two categories though, so it is kind of redundant. If the device can't due at least 100+ Khz with excellent linearity, you will have "side effects" that are audible and measurable within the normally accepted "audio range" of 20 Hz to 20 Khz. This can be seen as rounded edges of square waves, ripples or sags on the tops of the square waves, ringing on the trailing edge of the square wave, etc... Most of these things would be quite visible on a good quality scope with a test frequency of 10 Khz or so if the amp was "too slow". Sean > |
No offense to the two previous posters, but let me try to answer you in less technical terms. Perhaps some will disagree with me here, but this is how I define it. Terms like "fast" and "quick" refer to a component's ability to reproduce the natural dynamics of a sound. Dynamics or transients are the changes from quiet moments in a passage of music to loud moments and back again. Musical instruments when heard live (for example, think of plucked guitar strings, percussion instruments, etc.) have a certain snap to them. There's an immediacy to the individual notes. They sound, well, "live." Recorded music sometimes lacks that natural level of dynamics and--by comparison--can sound like you're listening to the music with a curtain between you and the musician. Instead of a sharp raise of a plucked string going from jet black to instantly there before you, it's as if there is a slow (relatively speaking--we're talking about the speed of sound here) build up of the sound. Good audio systems can come close to capturing that same level of quickness that live music has. So a "fast" or "quick" or "dynamic" component is one that is able to render those soft to loud passages in a realistic manner. At least that's how I've always defined it. |
Sean, in my experience quite a few of the manufacturers also include a low pass filter in their high bandwidth amplifier designs. If an amplifier's pass band extends well above 100KHz it is also capable of amplifing RF signals which may not be a desirable effect. |
Everyone has offered valid points. Perhaps the only thing I can add are some real world examples. In my experience(especially within a manufacturer's line), the smaller, less expensive amps are faster. The bigger ones are stronger. If you have demanding loudspeakers, requiring the power, you need the larger amps. But if you speakers are more amplifier friendly, the smaller amps can give you more musicality(I know I am going to be ripped by some for saying this). In my opinion, it is related to the power supplies. Massive transformers, and coffee can sized caps make a lot of juice available. But it is often slow and unweildy. An analogy would be comparing a dump truck to a Toyota MR2. The dump truck has the torque, but is slow and ponderous. The MR2 doesn't have the power of the dump truck, but is zippy, light on its feet, agile, and FAST. I think that the easiest place to find these differences in audio amps is to compare North American and European amps. Euro amps are smaller, lower powered, and faster. Examples would be Musical Fidelity, Electrocompaniet(not all of them), Jadis, Kora, Audio Analogue, LFD, Cambridge, Pathos, etc. American/Canadian amps are bigger, more powerful, and slower. These brands would be Krell, Mark Levinson, Jeff Rowland, Classe, Bryston, Audio Research, VTL, etc. For Thiels, you need the current of the larger amps. For Triangles, you can use the smaller amps. I have been through a complete metamorphosis over the past decade. I used to be on one side(big amps), now I am on the other. |
trelja: my experience doesn't quite match yours. with relatively inefficient speakers (< 87 db), like those iv'e owned in different versions for the past 10 or 12 years, high power/current amps are required. when i listen to different amps by the same manufacturer, they tend to have better and better transient response (rise/decay) as the output/current increases. brands i've experienced this with include accuphase, boulder and jrdg. |
If Euro amps are indeed faster than American amps it's probably because they've found a way to apply a tax based on the power output or even the diplacement (number of tubes, size of transformers, etc) of amps in the fashion as used for automobile engines. Why do you think there are so many high revving, small displacement engines in Europe? Hot hatch amplifiers? ;-) |
I think that most everybody has hit upon certain design aspects that are all quite valid. Macm brought up some simple yet good observations. Gmkowal also mentioned an interesting viewpoint that has proponents on both sides of the fence. This stance differs with the designer of one of my preamps. His feelings were that the only reason that a pre / power amp has problems with RF is that they are TOO slow to respond to them. He was designing audio circuits that slewed at over 350 v/ Usec and had rise times in the area of .01 Usec back in 1974. Then there is the fact that ANY filter causes phase shifts, impedance bumps, increased noise and circuitry, etc.. to deal with. My thoughts are to make the unit as fast and simple as possible, the theoretical "straight wire with gain" school of thought. Using that school of thought, Trelja's observations that smaller ( and therefore simpler ) amps sound better would be logical. In most cases, i would say that this is PROBABLY true. I think that it boils down to the fact that the larger amps have more circuitry and output devices. This requires a tighter tolerance for all parts involved with much greater care necessary in matching active components. This takes a great amount of time and is therefore probably neglected in most production based situations. The result is good sound with high power but not quite as good as what could be done if EVERYTHING were extremely tightly matched. While this could be said of smaller amps also, there is simply more parts in a bigger amp and more to go wrong in terms of parts tolerances, making the effects more noticeable. Another factor to all of this as mentioned above is speaker efficiency. Modern technology has raised the efficiency of speakers to the point that less power is now required to get the job done. To follow suite, designers have therefore put more attention into making very good sounding small amps. Once that is achieved, they simply expound on that curcuit to increase power output while PROBABLY not paying quite as much attention to parts tolerances. That is the very reason why some of the small to meduim amps sound slightly sharper and more detailed than the bigger amps with more parts and circuitry. Sean > |
I was just going to post a similair thread, GOOD TIMING. I was playing around with different footer and shelf materials and was amazed at how the pace changes with different combinations. I don't get it, I see how the electronics can play a part, but why cones and base material? |
I must admit to placing a bit of a teaser re my question about "Fast Amplifier" I have always felt, and demonstrated by this by listening tests that an amplifier where the bandwidth of the input signal is deliberately limited by a simple two stage passive RC network to be say 1 dB down at 25 KHz, and say 3 dB down at 35 KHz sounds better on actual music program than amplifiers not so bandwidth limited. The RC network should preferably be inside the amplifier, directly at the input, ahead of any active amplification. This network, presuming that the amplifier power output stage is a quality one, and have a respectably high slew rate would serve to limit the bandwidth of the input signal to something that the actual power stages can handle. Overtaking or exceeding the bandwidth of the power output stage will cause switching problems at crossover, and feedback phase problems leading to bad quality sound. Any comments. |
Your findings are only limited by the equipment that you've used and have seen in production. There were prototype amps made back in the early 80's that slewed at 1000 V / uS and had rise times as fast as those mentioned for the preamp in my previous post. Unfortunately, those prototypes were never put into production as the designer changed avocations and went into designing and manufacturing equipment for nuclear power plants. Sean
> |
