what is slew rate?

i have been reading various datasheets for example op amps and max232 and and low/high slew rate is mentioned alot but after googling alot i dont really understand what slew rate is and why it matters so does anybody have a simple explanation that i will understand?

hi L,
Its basically the rate at which the output voltage can change with respect to time with reference to the input.
E
https://en.wikipedia.org/wiki/Slew_rate

i saw the wiki one earlier but didnt get it, now i understand, what really confused me was some chips bragged about a high slew rate and some brag of a low one so it was hard to work out wether it was good or bad to be high or low, now i see it depends on what the chip is used for
so for example a opamp would be better with a low slew rate?

hi.
Found you a nice tutorial on OPA's look at page #15 for Slew Rate.
E

If I wanted to send video through a opamp I need fast v/us.
Audio can use a much slower slew rate.
Fast amplifiers usually use more power, and have more distortion.
To make a 10 volt signal compared to a 1V signal you need faster slew rate.

This is much like a car. How fast do you need to get from 0 to 60mph? It has nothing to do with maximum speed.

thanks for that chaps! i will read the pdf , next question is a bit odd, i normally only use DC so i dont know much about AC hmm hang on will do a new post on it

Hi,

Slew rate refers to how fast a signal can change from one level to another level.

The main problem with op amps is that there is a lot of circuitry that makes up the op amp and the circuitry can not change instantaneously to an input that changes. It takes some finite time for the output to go from one level to the next required level.

For a simple example, say we have an op amp with a gain of only 2 set up with appropriate resistors.
And lets say that we have 0v on the input. That means we have 0v on the output.

Now we change the input from 0v to 1v. That means we should get an output of 2v because the gain is 2, but what happens at first is the op amp has to have time to respond to this new excitation which is now 1v not 0v anymore. So it is said to "slew" from the first level to the next level, which for our example would be from 0v to 2v. So it starts to slew when the input changes, and it finishes the slewing process after it reaches 2v. So it is viewed as a ramp, from the first level to the next qualitatively.

We also like to be able to calculate this for a given device we might purchase from the web somewhere, so we need to be able to calculate the time it will take to slew for our given input and output.
Since our example amplifier has a gain of 2 and we change the input from 0 to 1v, the output goes from 0v to 2v in time t1. But slew rate is usually given in voltage per second or voltage per microsecond. So because this output went from 0 to 2 we have to divide the time t1 by 2 to get the slew for a 1v level change.

In other words, if the data sheet shows 0.5 volts per microsecond and our output when from 0 to 2v, then it would take 4 microseconds for our output to go from 0v to 2v. For the first microsecond it goes from 0v to 0.5v, then for the next microsecond it goes from 0.5v to 1v, then the next microsecond from 1v to 1.5v, then the next microsecond from 1.5v to 2.0v, and then it stops the slewing process.

So we got from 0v to 2v on the output but it took 4us. With a higher speed op amp, it would have gotten there faster. If the slew rate was 2v per microsecond, it would have went from 0v to 2v in only 1 microsecond. So you can see how the slew rate affects the ability of the op amp to respond quickly to a changing input signal.

For an AC signal the slew rate is fastest at the zero crossing of the test input sine wave. This allows us to calculate the required slew rate for a given AC application that has to respond to a certain maximum frequency and max amplitude. Note that we need to know both the max frequency and the max amplitude, because they are both required in order to calculate the slew rate at the zero crossing of the sine wave.
The slew rate at the zero crossing is just the ramp time where the voltage changes from just under zero to just over zero, divided by the voltage delta. So if the ramp went from -0.25v to +0.25v in 1 microsecond, the slew there would be 0.5 volts per microsecond. If that sine was at the max frequency and max voltage then we would need an op amp capable of slewing 0.5 volts per microsecond or faster.

Since higher speed op amps are often more expensive we usually try to pick one that will do the job but isnt an overkill for the application. For example, we would not need a 10v per microsecond op amp in an application that only required a 1v per microsecond op amp.

Always keep in mind that for AC signals the max output amplitude is just as important as the max frequency when dealing with slew rate. If either one is too high for the op amp it will cause output signal distortion.
Luckily if the effect is small distortion will only show up near the zero crossing which has less of an overall effect on the total THD than near the peaks, and near the peaks the slew is lower for sine waves.

thanks for that it helps alot, slew just seems a odd word, i did a kind of test just now, a simple square wave 1kH then zoomed right in on the scope and bingo the square wave at realy fast scope speeds isnt square it slopes up (as i would expect) but now i know that the time from bottom of slope to the top is the slew rate?? wich is kinda neat and now useful to me

i have to learn 4 new things a day minimum and today has been a good day for learning thanks all

Hi again,

Well, the time from bottom to top is approximately known as the "rise time".

The slew rate is the voltage from top to bottom divided by the rise time.

So if you have the edge of the square wave going from 0v to 5v in 5us, then that is 5v in 5us which is a slew rate of 5v/5us and simplified that is 1v/us. So the rise time is 5us and the slew rate is 1v/us.

MrAl said:

Hi again,

Well, the time from bottom to top is approximately known as the "rise time".

The slew rate is the voltage from top to bottom divided by the rise time.

So if you have the edge of the square wave going from 0v to 5v in 5us, then that is 5v in 5us which is a slew rate of 5v/5us and simplified that is 1v/us. So the rise time is 5us and the slew rate is 1v/us.

Click to expand...

perfect now i get it!! sorry it took so long but sometimes i need it in a certain form i can relate too makes loads more sense if i can relate it to something i am familiar with like scope square waves

Note that a high frequency response does not necessarily mean a high slew rate. Some op amps can have a high gain-bandwidth frequency but can generate only a small output voltage at high frequencies because they also have a relatively low slew-rate.

A lousy old 741 opamp has a slew rate of only o.5V/us. Its output cannot produce a 28Vp-p sinewave higher than 9kHz.
A lousy old LM358 dual or LM324 quad opamp has a slew rate of only (the datasheet does not say it). Its output cannot produce a 20Vp-p sinewave higher than 2kHz.
A TL07x audio opamp has a slew rate of 13v/us. Its output can produce a 24Vp-p sinewave up to 100kHz.
Audio goes to 20kHz.

Little Ghostman said:

i have to learn 4 new things a day minimum and today has been a good day for learning thanks all

Click to expand...

A very good goal to aim for. Keep it up and you'll be the most knowledgeable person on here in no time at all.

Mike.

Mr Al said:

The slew rate is the voltage from top to bottom divided by the rise time.

Click to expand...

I was always taught to use the 10% to 90%, not 0 to 100%

KeepItSimpleStupid said:

I was always taught to use the 10% to 90%, not 0 to 100%

Click to expand...

hi kiss,
I agree 10% to 90%
E.

KeepItSimpleStupid said:

I was always taught to use the 10% to 90%, not 0 to 100%

Click to expand...

Hello there,

Well, you will notice i used the word 'approximately' in the first reply that talked about this.
That's because for one i wanted to keep it simple, but really it doesnt matter if you use 0 to 100 or 10 to 90 percent when calculating the slew rate for a perfect ramp. You can use 0 to 100, 10 to 90, 20 to 80, or even 10 to 20 percent. As long as you use the time period between those two states and also the voltage levels of those two states.

Rise times are sometimes defined differently depending on the application or what is being measured or what information we are looking for.

For example, for a ramp that goes from 0 to 100 percent over a voltage of 0 to 10 volts and it takes 10us, then to go from 0 to 50 percent it changes by only 5 volts and that only takes 5us. So if you calculate using 0 to 100 percent you get 10v/10us and if you calculate using 0 to 50 percent you get 5v/5us, either of which results in a slew rate of 1v/us.
So i tried to keep it simple and approximate by using the description of "bottom to top" or even "top to bottom".

What happens in real life however is much more complicated because the wave may not be a perfect ramp. In that case you would want to find the max slew rate which would be the place where it changes the fastest. That might be from 40 percent to 60 percent for example.

So rather than call it "rise time" even though that is accurate, we should probably just call it the "time it takes to rise" (from one level to another). This way we leave the more common convention alone.

One thing we dont want to do though is make the mistake of using the full voltage change while using only something like the 10 and 90 percent rule. In that case we might measure a time that is shorter than what we really needed. So if the wave changes from 0 to 10v where 0 is min and 10v is max 100 percent and we use the 10 and 90 percent rule, then we would be using a time value that is too short because the voltage only changed by 8 volts between those two limits.

Why is 10% to 90%? Is this to allow a safe error margin?

that explains why on my anologue scope it has a 10% line and a 90% line marked out

audioguru said:

A lousy old 741 opamp has a slew rate of only o.5V/us. Its output cannot produce a 28Vp-p sinewave higher than 9kHz.
A lousy old LM358 dual or LM324 quad opamp has a slew rate of only (the datasheet does not say it). Its output cannot produce a 20Vp-p sinewave higher than 2kHz.
A TL07x audio opamp has a slew rate of 13v/us. Its output can produce a 24Vp-p sinewave up to 100kHz.
Audio goes to 20kHz.

Click to expand...

thanks AG that is a good way too why it matters and explains some problems i had with a LM324. every ones info has gone in my book i have asked for a big hardback ledger type book for xmas so i can keep my notes in that, had to be a xmas present because £28 for a writing book is alot, but i want to keep it, because it has all the stuff i have learnt in it and i refer back to it alot