Transistors slew rate...

Duckman · Post by **Duckman** » 19 Apr 2012, 04:06

There's such a thing?

I mean, I think it would, like in every OA, but I'm guessing if there's noticeable effect when you're choosing between them in a buffer level, or a Big Muff, or any transistor dependant device.

culturejam · Post by **culturejam** » 19 Apr 2012, 04:24

I think maybe there is slew rate with any amplifier circuit.

Duckman · Post by **Duckman** » 19 Apr 2012, 04:39

culturejam wrote:I think maybe there is slew rate with any amplifier circuit.

Silly me

Not a proper title I guess... What about "Determining slew rate"? Or "Saving private Duckman?

Ok, I will restate the situation: what about when you have a transistor based booster...
How can I properly determine the SR in that situation?
Any way, the choosed transistor didn't affect at all?

culturejam · Post by **culturejam** » 19 Apr 2012, 04:51

Sorry! Didn't mean to be a smart-ass!

I honestly don't know how to calculate slew rate on any transistor-based boost.

Duckman · Post by **Duckman** » 19 Apr 2012, 18:26

culturejam wrote:Sorry! Didn't mean to be a smart-ass!

I honestly don't know how to calculate slew rate on any transistor-based boost.

No, you didn't!... I'm pathetic

My curiosity revolves around the fact that in the world of OA, a high SR implies a better performance (as far as sound quality and bandwidth concerns) and wondered if this same concept applied where only one or two transistors are doing the work.
Any way, seems to be not very gravitating from the time that I searched for more, nobody talks about that topic.

Thanks, CJ!

culturejam · Post by **culturejam** » 19 Apr 2012, 18:33

Duckman wrote:and wondered if this same concept applied where only one or two transistors are doing the work.

I think it does apply. I recall having a conversation with soulsonic (on Facebook, I think) about this very thing. He seemed to think that slew rate would indeed apply to "discrete" amplifier circuits.

Duckman · Post by **Duckman** » 19 Apr 2012, 19:09

Well, is good to know that I'm not alone

Sounds logic since OA's are a bunch of transistors, resistors and maybe some diodes. Maybe the same concept translated to discrete devices got something to do with how "clean" or "dirty" a booster is.

cropec · Post by **cropec** » 23 Apr 2012, 21:28

I guess transitor's slew rate is related to its input capacitance and base resistance but i don't know how to calculate actual slew rate using this.

blackbunny · Post by **blackbunny** » 24 Apr 2012, 01:23

Slew rate measurement applies more to amplifiers than individual transistors. The characteristics of the individual transistors that you mentioned play a significant part in the overall circuit, as well as interstage impedance loadings and capacitances. If you want a brief insight into circuit design considerations including slew rates, check out Rod Elliot's website:

http://sound.westhost.com/amp_design.htm#slewrate

Duckman · Post by **Duckman** » 24 Apr 2012, 01:58

You never end reading those ESP pages!

And always got something interesting.
Thanks!

mmolteratx · Post by **mmolteratx** » 24 Apr 2012, 08:01

Walt Jung has a number of articles regarding slew rate limiting and it's resultant distortion on his site. Bjorn Juhl recommended him when I asked him some questions about the HBOD design regarding slew rate limiting and feedback.

http://waltjung.org/PDFs/SID_TIM_TAA77_P1.pdf

Contains equations. Basically, slew rate is the output voltage rate of change, measured in V/S (or V/uS). Formula is SR=I/C, where SR is slew rate, I is output current and C is the capacitive load being driven by the amplifier.

Duckman · Post by **Duckman** » 24 Apr 2012, 08:43

blackbunny wrote:Slew rate measurement applies more to amplifiers than individual transistors.

mmolteratx wrote:Basically, slew rate is the output voltage rate of change, measured in V/S (or V/uS). Formula is SR=I/C, where SR is slew rate, I is output current and C is the capacitive load being driven by the amplifier.

So, as a conclusion, is useless to consider some kind of "slew rate" in a single or double transistor booster stage, like we can do in any OA based stage, right?

blackbunny · Post by **blackbunny** » 24 Apr 2012, 10:39

Duckman wrote:So, as a conclusion, is useless to consider some kind of "slew rate" in a single or double transistor booster stage, like we can do in any OA based stage, right?

Well, any time you make a functioning amplifier you can measure it's slew rate. I don't think it's so relevant in a simple voltage amplifier stage, unless you're trying to drive difficult loads, for example, a following stage that has a relatively low impedance and / or very high capacitive loading. Neither of these are usually a problem in stompbox or preamp design, and you mainly have a limited audio bandwidth (80 Hz - 10 KHz is enough for most stompboxes), so slew rate isn't usually worth worrying about.

Duckman · Post by **Duckman** » 24 Apr 2012, 16:53

OK, thanks to all who used some of their time to help me understand.
I'm finding that I have the uncanny ability to focus on what does not matter.

Seiche · Post by **Seiche** » 26 Apr 2012, 10:28

Duckman wrote:a high SR implies a better performance (as far as sound quality and bandwidth concerns)

is this really correct, though? I guess it depends what you define as "sound quality". From what I understand, in the hifi sense this is correct. In stompboxes a high slew rate might not be desirable though. I'm still trying to get behind what slew rate means. Rate of change of Voltage, so it's what the amplifier drives in a certain amount of time. A low slew rate would indicate it's slower to change that certain state?

Is a lower slew rate somewhat reminiscent of "sag"?

juanro · Post by **juanro** » 26 Apr 2012, 12:48

Not sag. More like "an inability to reproduce high frequency without distorting the heck of it".
Like: the Pt2399 has a miserable output slew rate (mostly due to low clock frequency and integrator limits). So you have to filter the high(*) freqs you send to it, or else you get those badly distorted (percibed as crackling noise, hiss, "grainy" sound)

Juanro

(*)"high" in the case of the PT2399 is "anything above 3KHz" - in the better case.

Seiche · Post by **Seiche** » 26 Apr 2012, 12:51

so people prefer the lower slew rate in some circuits because the highs are somewhat more distorted, maybe more compressed?

edit: the picture looks like it makes the whole profile rounder, more sinus curvy. http://en.wikipedia.org/wiki/Slew_rate

juanro · Post by **juanro** » 26 Apr 2012, 12:57

When limited a square wave, yes. But the same occurs when the input is a sine wave.
When slew rate limit is reached, the output goes up or down as a straight line (the slew rate limit), whatever the input is, so it tends to look as a triangle waveform.

EDIT: I don't know why people preffer lower slew rate. I always associated it with "less treble response", so may be that's it.

DrNomis · Post by **DrNomis** » 26 Apr 2012, 13:18

I think any amplifier, whether it be an Op-Amp, Tube/Valve, or Discrete Transistor type, is going to have some amount of Slew-Rate, Slew-Rate is usually quoted as so many Volts/milivolts/microvolts per second, or micro-second, so, an amplifier with a Slew Rate of say 10V/Second, is going to have a slower Slew-Rate than one with a Slew-Rate of say 100V/microsecond....

By definition, Slew Rate is basically the rate-of-change of an amplifier's output (in terms of volts), over a certain amount of time, note that this assumes that the amplifier's output is changing over it's maximum output capability without any signal-clipping.....

I think device parasitic capacitance, and applied negative feedback will have some contribution to an amplifier's Slew-Rate, which will become more noticeable as the input-signal frequency is increased, if the signal waveform was originally a Sine wave, then it will look increasingly like a Triangle wave due to something called Slew-Rate Distortion....

My ten cents worth.....

blackbunny · Post by **blackbunny** » 26 Apr 2012, 15:19

Seiche wrote:
Duckman wrote:a high SR implies a better performance (as far as sound quality and bandwidth concerns)
is this really correct, though? I guess it depends what you define as "sound quality". From what I understand, in the hifi sense this is correct. In stompboxes a high slew rate might not be desirable though. I'm still trying to get behind what slew rate means. Rate of change of Voltage, so it's what the amplifier drives in a certain amount of time. A low slew rate would indicate it's slower to change that certain state?

Is a lower slew rate somewhat reminiscent of "sag"?

juanro wrote:Not sag. More like "an inability to reproduce high frequency without distorting the heck of it".
Like: the Pt2399 has a miserable output slew rate (mostly due to low clock frequency and integrator limits). So you have to filter the high(*) freqs you send to it, or else you get those badly distorted (percibed as crackling noise, hiss, "grainy" sound)
Juanro

(*)"high" in the case of the PT2399 is "anything above 3KHz" - in the better case.

A high slew rate is considered important by many HiFi afficianados, and several "fast" monitoring / reference amplifiers claim slew rates of more than 1000v/us. These amplifiers are often very unstable, tending toward supersonic / parasitic oscillation, and can be extremely fussy about the speaker loads they drive.

In the stompbox world, we have the 4558 op amp family with a slew rate of 1v/us, the NE5532 industry standard audio op amp with a slew rate of 8v/us and the humble TL071 with a slew rate of around 13v/us.

Compare this to the LM4250 precision low power op amp IC (0.002-0.1v/us depending on supply voltage and current set resistor), which is used to good effect in EMG active pickups and Music Man bass preamps.
The EMG circuit uses the LM4250 cleverly: two low impedance, low output coils feed the differential inputs of the LM4250, moderate voltage gain of around 5 or so, and the very low slew rate rounds off high frequencies smoothly, resulting in a clean-sounding pickup with a fairly flat frequency response, low noise and very long battery life. All from a very low slew rate op amp.

Magic? No, just clever design and engineering. High slew rate is simply not important in this application.