Head-Fi.org › Forums › Equipment Forums › Sound Science › Slew Rate vs. Distortion
New Posts  All Forums:Forum Nav:

Slew Rate vs. Distortion

post #1 of 13
Thread Starter 
Does an acceptable slew rate, or one that covers the higher bandwidth frequencies, lead way to lower distortion? I always thought that distortion was produced when the headphones couldn't respond fast enough to the change in frequency especially with higher level frequencies, so therefore wouldn't a higher slew rate correspond to lower distortion at the same measured frequency?

Sent from my XT907 using Tapatalk 2
post #2 of 13

I'm not aware that headphones have a slew rate. But slew rate limiting in amplifier circuits does indeed add distortion. I just wrote this article:

 

Slew rate and slew rate limiting

 

--Ethan

post #3 of 13
Thread Starter 
Quote:
Originally Posted by EthanWiner View Post

I'm not aware that headphones have a slew rate. But slew rate limiting in amplifier circuits does indeed add distortion. I just wrote this article:

 

Slew rate and slew rate limiting

 

--Ethan

 

I should have clarified, my fault. I meant does an amp with sufficient slew rate dictate lower distortion, (not considering a headphone's FR impact) but I think you got the idea. I'll take a look at the article listed there. Thanks!


Edited by J Bones - 2/14/13 at 12:50pm
post #4 of 13
Quote:
Originally Posted by J Bones View Post

I should have clarified, my fault. I meant does an amp with sufficient slew rate dictate lower distortion but I think you got the idea. I'll take a look at the article listed there. Thanks!
I think the right question to ask is whether an amp with insufficient slew rate causes more distortion.

I mean, one can always build an amp with arbitrarily high slew rate and arbitrarily high distortion caused by other factors smily_headphones1.gif
post #5 of 13
Quote:
Originally Posted by EthanWiner View Post

I'm not aware that headphones have a slew rate. But slew rate limiting in amplifier circuits does indeed add distortion. I just wrote this article:

 

Slew rate and slew rate limiting

 

--Ethan

That's great but it doesn't address the issue directly. You can build a low pass that behaves in phase and in that case would not limit performance of audio band with out of audio band tayloring. At that point the internal bandwidth of the amp in relation to it's feedbakc loop becomes the inportant slew factor. You've basically just wrote that every tone control is crap. (oh wait, I may agree with that)bigsmile_face.gif

post #6 of 13
Thread Starter 
Quote:
Originally Posted by mich41 View Post


I think the right question to ask is whether an amp with insufficient slew rate causes more distortion.

I mean, one can always build an amp with arbitrarily high slew rate and arbitrarily high distortion caused by other factors smily_headphones1.gif

 

That is a great point and is mostly what I was looking to see. I know high distortion isn't directly related to the slew rate, but more of the reverse, if slew rate needs to be sufficient in order to have any chance at minimal distortion.

post #7 of 13

Distorting regarding slew rate should be looked at in terms of audible distortion.

post #8 of 13

It's an inconsequential measurement if you're trying to get some direct correlation to distortion from measuring the slew rate at the output of an amp. If it slews fast enough to reproduce 20khz at power, the rest is how, where and why it's been slew limited. It's not uncommon to do so to limit transient distortion.


Edited by goodvibes - 2/14/13 at 4:07pm
post #9 of 13
Thread Starter 
Quote:
Originally Posted by goodvibes View Post

It's an inconsequential measurement if you're trying to get some direct correlation to distortion from measuring the slew rate at the output of an amp. If it slews fast enough to reproduce 20khz at power, the rest is how, where and why it's been slew limited. It's not uncommon to do so to limit transient distortion.

 

I wasn't exactly looking for a direct correlation, but if slew rate would even be something you might consider in choosing an amp with minimal distortion at the output. Having read an article the other day on slew rate it was pointed out that more or less slew rate is a figure that one really should not consider in choosing an amp over another, but not why. But I get the feeling that other factors are probably more important in choosing an amp that produces minimal distortion other than the slew rate.

post #10 of 13
Quote:
Originally Posted by goodvibes View Post

That's great but it doesn't address the issue directly. You can build a low pass that behaves in phase and in that case would not limit performance of audio band with out of audio band tayloring. At that point the internal bandwidth of the amp in relation to it's feedbakc loop becomes the inportant slew factor. You've basically just wrote that every tone control is crap. (oh wait, I may agree with that)bigsmile_face.gif

That's not exactly what Ethan's article showed, there was a difference between an RC low pass filter and a slew-limited constant current source.  The article didn't mention the actual distortion slew limiting causes, which used to be known as TIM - Transient InterModulation distortion, or SID - Slew-Induced Distortion.

 

In an RC low pass, while filtering of highs is occurring, the rest of the audio passes untouched except for a bit of phase shift.  However, while the amp is slewing and doing so at its maximum rate, it can do nothing else, including pass other audio, so during the slew period the signal that causes the amp to go into slew limiting is modulating everything else in the signal.  That's the distortion mechanism.  It's a form of intermodulation distortion that occurs under very specific conditions.  

 

The interest in TIM distortion and slew limiting started in the early 1970s with early IC Opamps that were very slow.  When those chips found their way into power amp front ends and preamps, the whole TIM thing came to light.  With the introduction of IC Opamps that would slew between 7 and 13V/us in the late 1970s and early 1980s, the whole thing pretty much went away.  More recent research has shown actual TIM/SID to be very limited in audible detriment, and it has all be played down a lot lately.

 

What's important is to realize that most (and by most I mean almost all) amps used for audio these days are far faster than required, and won't be pushed into slew limiting by audio signals. The bad combination is high frequency AND high voltage swings combined with slow slew rates.  You need all three aspecst to hit slew limiting.  Take away one, and you're out of the woods.  In the case of headphone amps, we don't have very high voltage swings, and likely have fast amps, so no problem.  

 

Including slew rate figures in published specs has always been rare, mostly because while higher numbers are always impressive, the average spec-reader has no idea that once an amp is fast enough, faster makes no difference at all.  In fact, very fast amps come with stability issues, which are much worse than TIM, so in many ways, slower (as long is it's fast enough) is better.  

 

The mechanism for slew rate limiting doesn't occur in transducers.

 

I recall the NE531V opamp - fast as heck, very hard to design with.  Mostly you made it stable by forcing it to slew slower.  

post #11 of 13
Quote:
Originally Posted by EthanWiner View Post

I'm not aware that headphones have a slew rate. But slew rate limiting in amplifier circuits does indeed add distortion. I just wrote this article:

Slew rate and slew rate limiting

--Ethan

Ethan,

Thank you for the link.
I appreciate your articles and always learn something from you.

Regards,
CJ
post #12 of 13
Thread Starter 
Quote:
Originally Posted by Chris J View Post


Ethan,

Thank you for the link.
I appreciate your articles and always learn something from you.

Regards,
CJ

 

Took me awhile to understand the article Ethan, but that article and jaddie's comment make it much clearer to understand the difference between the two variables. Thank you for your input!

post #13 of 13
Quote:
Originally Posted by jaddie View Post

That's not exactly what Ethan's article showed, there was a difference between an RC low pass filter and a slew-limited constant current source.  The article didn't mention the actual distortion slew limiting causes, which used to be known as TIM - Transient InterModulation distortion, or SID - Slew-Induced Distortion.

 

In an RC low pass, while filtering of highs is occurring, the rest of the audio passes untouched except for a bit of phase shift.  However, while the amp is slewing and doing so at its maximum rate, it can do nothing else, including pass other audio, so during the slew period the signal that causes the amp to go into slew limiting is modulating everything else in the signal.  That's the distortion mechanism.  It's a form of intermodulation distortion that occurs under very specific conditions.  

 

The interest in TIM distortion and slew limiting started in the early 1970s with early IC Opamps that were very slow.  When those chips found their way into power amp front ends and preamps, the whole TIM thing came to light.  With the introduction of IC Opamps that would slew between 7 and 13V/us in the late 1970s and early 1980s, the whole thing pretty much went away.  More recent research has shown actual TIM/SID to be very limited in audible detriment, and it has all be played down a lot lately.

 

What's important is to realize that most (and by most I mean almost all) amps used for audio these days are far faster than required, and won't be pushed into slew limiting by audio signals. The bad combination is high frequency AND high voltage swings combined with slow slew rates.  You need all three aspecst to hit slew limiting.  Take away one, and you're out of the woods.  In the case of headphone amps, we don't have very high voltage swings, and likely have fast amps, so no problem.  

 

Including slew rate figures in published specs has always been rare, mostly because while higher numbers are always impressive, the average spec-reader has no idea that once an amp is fast enough, faster makes no difference at all.  In fact, very fast amps come with stability issues, which are much worse than TIM, so in many ways, slower (as long is it's fast enough) is better.  

 

The mechanism for slew rate limiting doesn't occur in transducers.

 

I recall the NE531V opamp - fast as heck, very hard to design with.  Mostly you made it stable by forcing it to slew slower.  

We're on exactly the same page. Thanks for better defining.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Slew Rate vs. Distortion