Question for the engineers/audio sages [Archive] - Audio & Video Forums

PDA

View Full Version : Question for the engineers/audio sages



kexodusc
02-04-2005, 08:37 AM
Oh wise ones, in my never ending quest to learn more about all things audio, I have stumbled across a term I've heard frequently, but understand little of it's significance.

"Slew rate"...my limited understanding is that this is the ratio of how quickly the output of an amplifier can track its input, measured in V / usec.

How relevant (in real terms) is this slew rate? Are most amplifiers at a point now where the slew rate is so high that any increases produce diminished returns on quality of sound (ie: subtle nuances, dynamics, etc)? Why isn't slew rate frequently published as a spec for comparison?

jneutron
02-04-2005, 09:24 AM
Oh wise ones, in my never ending quest to learn more about all things audio, I have stumbled across a term I've heard frequently, but understand little of it's significance.

"Slew rate"...my limited understanding is that this is the ratio of how quickly the output of an amplifier can track its input, measured in V / usec.
While we wait for some wise ones to answer, I'll give it a try..

The generic term "slew rate", means the rate at which something can change. For amplifiers, it is usually defined as the maximum rate the output voltage can change, expressed in volts per microsecond.

If you attempt to force an amplifier to move faster than this rate, the amplifier, internally, may not be very happy. Sometimes, one of the semiconductors will be forced outside of the normal operational limits, and saturate or turn off entirely..the problem is not that a zistor does this, but it may take a little time for it to recover back to normal operation.
If you keep the same voltage output, but increase the frequency, the slew rate of the signal will increase...double the frequency, double the slew rate. At some point, the amplifier will be unable to keep up with the input, and the sine will begin to turn into a triangle.. This defines the maximum frequency sine wave that an amplifier will still be able to produce it's rated output.

How relevant (in real terms) is this slew rate?
If it is too slow, the high frequency performance will suffer.
[QUOTE=kexodusc] Are most amplifiers at a point now where the slew rate is so high that any increases produce diminished returns on quality of sound (ie: subtle nuances, dynamics, etc)?
To my understanding, typical amps are capable of full power up to 20Khz..I am unaware of amps that cannot provide full power at 20Khz, I am sure there are some out there, though.
As for subtle nuances, dynamics...if the signal can exceed the amp capability, the amp may not behave as predicted.

Why isn't slew rate frequently published as a spec for comparison?
I make the assumption that since everybody can exceed full power operation at 20Khz, it is no longer considered important.

It is also a difficult tradeoff, very high speed amp design vs stability into a capacitive load, so excess may not be always good.

My other post referred to current slew rate..I have quite a bit of experience in attempting to test this in low impedance circuitry, so have been relating the problems associated with it. To my understanding, amplifiers are not spec'd for current slew rate, but the assumption is that it is there because the voltage into a "resistive" load can meet the speed specification...My experience says that the "resistive" loads in use are incapable of drawing the current at a very fast rate of change, but using standard test equipment in standard ways can fool the person testing..

Unfortunately, all voice coil driven speakers rely on the current to develop a force within the speaker, so measuring the voltage is an indirect method of measuring the current. Measuring the slew rate of the voltage, does not really provide measurement of the slew rate of the current because of errors inherent with fast changing magnetic fields.. If the slew rate of the current is compromised by either the amp or the speaker cables to the extent that some of the signal is shifted, or delayed, to the tune of 1 to 5 uSec, then it broaches the realm of audible to humans.

Cheers, John

PS..Owner/operator: I am very unhappy that this site appears to be tossing data miners onto my computer and would prefer these cookies be left out.

sam9
02-20-2005, 11:54 AM
Does not relate directly to input. The raw number is simple: what is the maximum rate of voltage change that the amplifier can achieve at the output. (There is an internal figure as well, but that is really of interest only to the designer.) The higher the power rating of an amplifier, the greater the slew rate that is required. You would think that you could simply ensure that the number was great enough to reproduce a 20kHz signal at max output accurately. However, designers like to build in a bargin of safety, so they will often want to exceed the requirement. Up to 100W, it takes very little extra effort to design an amp with more than adequate slew rate. It might actually take more extra effort to design an amp that was INADAQUATE in this respect! As power increases the concern and difficulty increases.

On top of this there is a concern for TIM (Transient intermodulation distortion), where it is claimed there are errors caused by an inadequate slew rate at frequencies much higher than 20kHz that generate audible artifacts at frequencies below 20kHz. Not everyone believes this form of distortion is real, but in recent years it was a "hot button" in the high-end community. Even some designers who do not believe it is a real will design a high facor into their product just so make reviewers happy.

(I have no opinion whether it is real or not.)

hermanv
03-09-2005, 10:44 AM
If musical instruments have harmonic content that streches all the way to the limits of human hearing, generally defined as 20KHz (not true, many people can hear much higher frequencies). If we now electrically mix and sum a full orchastra of instruments many of which produce these same harmonics what is the highest peak instantaneous rate of change?

Having spare slewing ability is a lot like extra horespower in a car. The low powered car will get you there but the higher powered car has more acceleration (slew rate) it is more adaptable and felxible because there is more reserve available. Either car can get you a traffic ticket.

TIM distortion is a complex subject but think of a case where the error signal (feedback) can not keep up with the difference in slew rate on the input and output of an amplifier. An incorrect error signal will cause ugly things to happen. This is one strong reason why global feedback circuits have a deservedly bad reputation.