The slew rate is the inverse of the open loop frequency response, which is given in Figure 11 of the spec sheet.
The open-loop slew rate is always increased by negative feedback, which reduces gain.
If you built the circuit shown in the application notes (negative feedback providing a gain of 22,000Ω / 680Ω), you should have a gain of about 32, which would mean (from Figure 11) that your frequency response would be over 300,000 kHz. In this case, the bandwidth/slew-rate limiting factor would be C1 and R1, the input filter.
What sort of problem are you having that makes you think that the open-loop slew rate is causing problems? Did you try putting a square wave through it and measuring the rise time with a scope? (You should do that after the C1-R1 input filter.)
The problem that I suspected was just what it sounded way worse than other comparable amps. And slew rate is a easy and comparable parameter to measure.
Yes I measured the slew rate with a scope, feeding the signal through the input filter of C1 and R1. This is how I always measured the slew rate of amps. Since C1 and R1 forms low pass filter, I don't see how this would affect the the measurements much.
Yes I measured the slew rate with a scope, feeding the signal through the input filter of C1 and R1. This is how I always measured the slew rate of amps. Since C1 and R1 forms low pass filter, I don't see how this would affect the the measurements much.
A low-pass filter will certainly effect slew rate! That's essentially what they do.
1μf and 22kΩ yields a time constant of 22 mS, which means that, with a one-volt step input, it will take 22mS to reach 0.63 volts. With a gain of 32 and an ideal op amp, your slew rate would be 0.022 / 0.63 * 32, or about 1.
You must put your input on the op-amp side of that filter if you want an accurate slew rate.
C1, R1 form a high-pass filter, though. The time constant is 22 ms (milliseconds) not 22 mS (millisiemens). Seconds is a unit of time. Siemens is the unit for conductance.
On a rapid transient, C1 will act as a short circuit for AC, which preserves the leading edge of the transient. Over time you'd expect C1 to charge, which will result in some sloping of the top and bottom of a square wave.
Tom
On a rapid transient, C1 will act as a short circuit for AC, which preserves the leading edge of the transient. Over time you'd expect C1 to charge, which will result in some sloping of the top and bottom of a square wave.
Tom
There seems to be some confusion about slew rate and rise time. They aren't the same. Slew rate is the maximum dV/dT the amplifier can produce. It is a straight number, and as long as the output dV/dT doesn't exceed that number, slewing does not occur. It is very similar to clipping - as long as the output doesn't exceed the supply voltage, the amp won't clip. When the signal rises faster than the slew rate, the amplifiers will usually behave badly. Input bandwidth limiting helps but theoretically can still be overloaded (but not easily or with ordinary signals) My 2 cents.
It is an indicator of the circuit's speed of response to a change in input signal, I think.
It can be considered an indicator of the transient response quality of the device or circuit.
Too high (more than 20) is also not good, can create issues in an audio system.
In any case, the TDA 2040 is a more than a quarter century out of production, so this question does not seem particularly relevant at this point in time.
Please clarify why you need this information, the TDA2050 and LM1875 are equally good substitutes, and the 1875 has a lot of data available for it.
It can be considered an indicator of the transient response quality of the device or circuit.
Too high (more than 20) is also not good, can create issues in an audio system.
In any case, the TDA 2040 is a more than a quarter century out of production, so this question does not seem particularly relevant at this point in time.
Please clarify why you need this information, the TDA2050 and LM1875 are equally good substitutes, and the 1875 has a lot of data available for it.
Slew rate can be as high as you like really - a fast slew-rate amp probably has a very wide power bandwidth and therefore may be more susceptible to EMI or stability issues, but very high slew-rate in itself isn't a problem, after all a DC-1MHz amplifier handles audio just fine. Amplifiers for higher frequencies than audio have to have much higher slew-limits, audio amps can benefit from simpler design with modest slew-rate of course.
A gain 30 amp with 30V/µs slew-rate limit can only handle input signals with under 1V/µs.
20kHz sine at 2Vrms slews at 0.355V/µs, so input slew rates are usually well below 1V/µs. The higher the amp power and gain the more slew-rate it needs. Slew-limiting is a form of clipping, but current clipping in the VAS stage, not voltage clipping. This is normally a result of a fixed standing current in the VAS. Increasing VAS current will usually increase slew limit. Of course you require more power dissipation for the VAS and its current-source load - so its a trade off usually - more slew rate than needed will cost more in VAS components and heatsinking.
A gain 30 amp with 30V/µs slew-rate limit can only handle input signals with under 1V/µs.
20kHz sine at 2Vrms slews at 0.355V/µs, so input slew rates are usually well below 1V/µs. The higher the amp power and gain the more slew-rate it needs. Slew-limiting is a form of clipping, but current clipping in the VAS stage, not voltage clipping. This is normally a result of a fixed standing current in the VAS. Increasing VAS current will usually increase slew limit. Of course you require more power dissipation for the VAS and its current-source load - so its a trade off usually - more slew rate than needed will cost more in VAS components and heatsinking.
James Solomon's tutorial paper (link) disagrees. He says slew rate is set by the input stage current: SlewRate = IPS_current / C_compensation . It's equation (17) on page 320.
Solomon assumes the input stage includes current mirror loads, thus during a full slewing event, 100% of the input stage current (tail current) is delivered to the compensation capacitor.
Actually yes it usually will be - it depends on the component values and gain of the VAS, compensation network, capacitance of the drivers, etc, but a Miller compensated amp is likely to be limited more by the IS than the VAS. A comparator circuit without compensation is more likely to be limited by the VAS. Anyway its a current limiting phenomenon. If slewing isn't symmetrical its an indication the VAS is the limiting factor as the IS is symmetric.
theses links about TIM and SR may be of interest:
https://miniurl.be/r-5lwn (datasheet of same family with elements on SR.
old note:
https://miniurl.be/r-5lwm
https://miniurl.be/r-5lwn (datasheet of same family with elements on SR.
old note:
https://miniurl.be/r-5lwm