why is slew rate important?

[paper]

the opa 227 is generally considered quite decent, even though it has a slow slew rate. my experience is that the 227 seems to make notes a little more resonant, which i like quite a bit. is that due to a slow slew rate? what is slew rate? and what are transients?

 

[Clutz]

Quote:

Originally Posted by paper the opa 227 is generally considered quite decent, even though it has a slow slew rate. my experience is that the 227 seems to make notes a little more resonant, which i like quite a bit. is that due to a slow slew rate? what is slew rate? and what are transients?

Slew rate is the change in voltage per unit time, typically described as dV/dS.

 

[rickcr42]

When considering a "slew rate" you must also consider "at what fequency ? At what voltage ? Does this vary with load ? Vary with voltage ? stable into a reactive or capacitive load ? "
when everything is said and done anything over 25V/uS is adequate for line level audio, 2V/uS or so at the microphone/phono stage microvolts signal level and around 50-100V/uS at the power amp level though limiting the upper frequency bandwidth to say 80khz has the effect of making the stage have "unlimited bandwidth" even though a square wave will look like crap on a scope.Would be important if we listened to square waves but we don't so not

 

[RobDD]

Quote:

Originally Posted by rickcr42 When considering a "slew rate" you must also consider "at what fequency ? At what voltage ? Does this vary with load ? Vary with voltage ? stable into a reactive or capacitive load ? " when everything is said and done anything over 25V/uS is adequate for line level audio, 2V/uS or so at the microphone/phono stage microvolts signal level and around 50-100V/uS at the power amp level though limiting the upper frequency bandwidth to say 80khz has the effect of making the stage have "unlimited bandwidth" even though a square wave will look like crap on a scope.Would be important if we listened to square waves but we don't so not

Exactly. Did you ever look at a sq. wave at 50 hz through a cap. slew rate is the least of the problems. Maybe that's why I have a way to bypass mine which I do most of the time.

ps. I like the 227 too. I think it's the best chip for portable equipment. Love cellos through my AKG340 phones on it.

 

[RobDD]

ps.ps I built a SPICE model for the 227 and it shines in simulation too.

 

[nikongod]

if it sounds good and measures poorly, you are probably measuring the wrong things. someone else said that.

likewise the opposite is sometimes true too.

 

[kdaq]

Is it possible to quantify (even if only subjectively) how different slew rates affect the sound (all else being equal)?

 

[Clutz]

I think the difficult, and key part of that question is 'all else being equal'. There are lots of differences between different opamps other than just differences in slew rate.

 

[kdaq]

Yeah, I hear ya. This seems like a job for someone strongly versed in theory, as opposed to anything that would be practically possible in the realm of subjective hearing tests.

 

[Garbz]

i think slewrate is subtly linked to detail of a chip. I.e. the OPA604 has a slow slew rate compared to the OPA627 and the detail is clearly audible. Even more so for the OPA627 / 637 which has 3 times the slew.

I also think this is why balanced amps have potential to sound better. Double the slew.

 

[paper]

been reading up on the topic. in order to reproduce a full throttle sine wave at 20khz, the slew rate only needs to be 2.1 v/us. the 2227 makes the cut. so do tube amps, which max out at 10 v/us. i remember seeing an article about some guys making audio amplifiers out of video op amps, which have ridiculously high slew rates, close to the 4000 v/us range.

but then, the op amps top speed says little about how it drives. yeah, that's right. i just mixed a metaphor.

 

[Garbz]

and human hearing only goes to 20khz, so it makes little sense raising the breakup frequency of a tweeter from 27khz to 70khz yet that's exactly what B&W engineers have done, and it makes a fantastic difference.

Maths can't always explain how things sound, and that's a major problem for us!

 

[Francis_Vaughan]

Again, all other things being equal. Raising the breakup of the tweeter from right on the edge of audibility to much higher might have all sorts of in-band consequences for which the higher breakup is really simply an easily measured (and marketed) metric that has little to do with the real issue.

Slew rate means a great deal in a feedback amplifier. Which means all op-amp amplifiers. If any stage goes into slew rate limiting at any time the feedback loop opens - and really bad things happen. There are lots of ways this might happen - including instability (bad anyway) but even very low level oscillation, outside RF pickup (from either the input or output leads, or even the power - a good reason for snubbers and/or fast recovery diodes in the PS, and so it goes.) Don't assume that just because the audio signal is bandwidth limited that the environment the amp lives in is. And lets not even go near the issues with the NOS DAC crowd.

 

[tangent]

Take any op-amp of your choice that comes in both a compensated and an uncompensated version. (OPA627 & 637, NE5532 & 34, etc.) The primary effect of adding unity gain compensation is to lower the op-amp's slew rate; all else should be as close to equal as you can ask for in a practical system, so it makes a good test subject. Listen to both. If you can't hear a difference, go forth and happily ignore slew rate issues.

 

[Tomo]

Hey,

I was watching OPAMP specs a while back and I saw something on most of high speed opamps. ...

Spectrums (frequency respose diagram) show a small peak right before high-end 3dB drop.

Does this mean anything?

I made a High Bandwidth (High Speed) amp similar to Rod Elliot's THS6012 amp. I am pretty happy. But this high speed thing didn't do something significant as I like my META-like amp and YAHA amp (w/ <10uV/s opamp) also. But I have to say high freq sound great. (Bad recording sound really bad with it.)

Tomo

P.S. Do anyone know the slew rate of 74HCU04's? I gotta amp with it and four AA's. Driven my HD580 wonderfully yesterday, freaked me out!