Blind Listening Tests & Amplifiers

[IanHarvey]

FYI,:

Slew rate and frequency response can't be used to calculate the output voltage swing. The equation (slew rate = 2 * pi * Freq * Vpeak ) is used to calculate what slew rate you need to deliver a sine wave (of amplitude Vpeak at frequency Vpeak) without slew-induced distortion.

In other words, if you get an amplifier to deliver ever higher-amplitude and higher-frequency sinewaves, it will either run out of output voltage headroom, or it won't be able to swing the output fast enough. Which one happens first depends on whether the slew rate is respectively greater or less than 2*pi*Freq*Vpeak.

Cheers
IH

[mikek]

Originally posted by mrfeedback


Ok, so you are an idiot then.

..You should talk

[mikek]

Originally posted by IanHarvey
FYI,:

Slew rate and frequency response can't be used to calculate the output voltage swing. The equation (slew rate = 2 * pi * Freq * Vpeak ) is used to calculate what slew rate you need to deliver a sine wave (of amplitude Vpeak at frequency Vpeak) without slew-induced distortion.

In other words, if you get an amplifier to deliver ever higher-amplitude and higher-frequency sinewaves, it will either run out of output voltage headroom, or it won't be able to swing the output fast enough. Which one happens first depends on whether the slew rate is respectively greater or less than 2*pi*Freq*Vpeak.

Cheers
IH

Slew rate of itself, can most certainly not be used to calculate an amps. Voltage swing....i think this development may have arisen out of a misunderstanding of post #975, where i questioned pan's suggestion that an 800V/uS slew rate was prima facie a 'good' thing with respect to sound quality.......

.....for this to be the case, an amplifier with a norminal slew of 800V/uS would have to swing well over 6Kv at its output for slew rate to become a cause of audible impairment.

[mikek]

Originally posted by Pan


Maximum ouput voltage swing has NOTHING to do with slewrate.
Still don´t know what you talk about.

An opamp or video buffer with several hundreds of MHz BW and slerate of 2000V/uS must be able to swing several kV as well if I get you right???

I guess Elantec, BB and AD is just making those numbers up then..

/Peter

No..you 'get' me wrong......
...see post #1023

[mikek]

Originally posted by IanHarvey
to this forum, and it's good to see a range of opinions in the subjective/objective debate. Here goes:

1. It's important to be able to say "I can't hear a difference". Michelson & Morley presumed the existence of the ether which they were trying to measure, but had the courage to say their experiment had failed. From it we learned a lot.

I assume we've all made many upgrades and tweaks to our systems over the years, and if you're like me some changes made a night-and-day difference, and others were not. Personally, I want as many of the former and none of the latter as possible.

Ultimately, being able to say "it makes no difference" is essential to avoid being taken for a ride by charlatans.

2. I'm very happy - it's my normal practice - to let a subjective decision be the final arbiter rather than objective measurements. I'm even happy that this can be an unreliable and unrepeatable experience, and accept that this sometimes leads to flattery instead of accuracy.

Here's the 'but', though: there's a tendency to attach technical statements to subjective judgements e.g. "negative feedback makes your amplifier sound bad". This sort of statement implies there's an underlying physical mechanism at work, and this *must* be the subject of scientific inquiry. To make technical statements, but be unable to justify them technically, is not something I can stomach.


Cheers
IH

...I have absolutely no difficulty with this gentleman's approach....

...spot on in all respects....cheers

[Pan]

Originally posted by mikek




...I have absolutely no difficulty with this gentleman's approach....

...spot on in all respects....cheers

I agree with you on this one mikek, good post and I feel about the same as this gentleman. Makes me feel warm inside to see that you and I can share the same views

/Peter

[Pan]

Originally posted by mikek


Slew rate of itself, can most certainly not be used to calculate an amps. Voltage swing....i think this development may have arisen out of a misunderstanding of post #975, where i questioned pan's suggestion that an 800V/uS slew rate was prima facie a 'good' thing with respect to sound quality.......

.....for this to be the case, an amplifier with a norminal slew of 800V/uS would have to swing well over 6Kv at its output for slew rate to become a cause of audible impairment.

Not so sure I did suggest anything, only said that the best sound I´ve heard has been from power amps with very high "raw" slewrate and BW.

LC Audio Patriot and Zapsolute 4MHz/1MHz, Patriot having 1 degree phase shift at 20kHz and slew 800V/uS.

Dynamic Precision more than 250V/uS

The cheap LM3886 even though it in a simple circuit lacked some qualitys, still was very good.

I do not suggest that 4MHz BW and 800V/uS is necessary for an power amp to be "perfect", however I have never heard any other amp handle high frequencies as good as the
no-feedback 100W class A Patriot V100.

/Peter

[fdegrove]

Hi,

Peter,

High time for you to move into the thermionic camp...

From what I read so far, I think you'd like the way they can sound...you seem to be the perfect candidate for an OTL amp.

Cheers,

[purplepeople]

Since infinite slew rate would be a perfect amplifier (zero signal delay at output => input phase = output phase), I have to assume that slower amplifiers are less desirable.

I'm still trying to figure out if there is a way to measure the linearity of the slew rate and then correlating them to changes at the output. The value has a slope when graphed so the linearity must be measured in V/microsec^2.

Several questions come up...

1) Does the slew rate stay constant for the rated bandwidth or is the measured value some kind of average or median?
2) Does current draw (speaker impedance @ specific frequencies) affect slew rate? What other factors?
3) What is the minimum value for 20-20K?
4) How do tubes, transistors, and power opamps compare?

Also, I was curious so I checked the specs on the Bryston 7B. This is considered to be a very good amplifier and it seems that 60V/us is more than enough.

http://www.bryston.ca/7bsstspec.html

IM and THD+n are the same at .005% @ 8ohms 20-20K
Noise floor is 110dB 20-20K
Slew rate is 60V/us
Damping Factor of 300

:)ensen.

[SY]

pp, I think you're confusing time domain and frequency domain parameters. It's meaningless to ask what a slew rate is at a given frequency.

For the record, slew rate is a time domain parameter, measured by inputting a step function and measuring the slope of the amplifier's time domain response. V/us, not us^2. The slew rate can be calculated from an amp's bandwidth, too.

Better slew rate = better amplifier is true to a point. What that point is is a matter of some discussion. Measured slew rates of real-world audio signals are surprisingly low. IIRC, Nelson Pass published some results on this, indicating that the power amp was being asked to slew at most in the single digit V/us range.