Opamp slew rate question

[CaptDon]

I want to use a 6SL7 vacuum tube as a deflection amplifier for an electrostatic C.R.T.  I would like to drive the grids in a push-pull fashion using a dual opamp. One half of the opamp will be a simple unity gain buffer and the other half will be used as a unity gain inverter. I see the 5532 listed as 10MHz GBP with a 9V/us slew rate. The other choice was a 1458 which I think was listed as both 1MHz GBP or 3MHz GBP depending on manufacturer BUT both listed 13V/us as the slew rate. I am puzzled how the second choice has a faster slew rate but lower GBP? I would like my grid drive circuit to have a minimum 4MHz bandwidth in an effort to past reasonable looking 400KHz s square waves. The board is already wired but the only opamps I have on hand are TL072 and TL082 and they seem to not have the bandwidth I need. I am sure they are authentic because they are probably 20 years old and bought from a reputable U.S. supplier. I also now have the simple requirement of 8 pin DIP since that is the sockets on my board. What opamp would you recommend and why does it seem slew rate doesn't equal bandwidth or GBP? My supplies are + and - 12VDC and I only need around 6 volts of signal differential to the grids of the 6SL7 to get full deflection. Thanks!!

[Obin]

Higher slew rate should go hand-in-hand with higher BW. Your intuition is correct on that front. TI lists the slew rate of the 1485 as only 0.5V/us, which would be in line with the lower GBP. https://www.ti.com/product/LM1458


For high slew rate PDIP-8, i've used https://www.analog.com/en/products/AD8056.html in the past, a 1300V/us beast but your +-12V PSU would not work for that one.

The TL082 would be just below your 4 MHz spec, the 5532 is popular and likely a good choice in your situation, just be sure to have the correct datasheet/version

[Uunoctium]

For such a stage i think low distortion paired with high linearity is recommended. AD817 comes to mind. High GBP and high slew.

[Sensorcat]

... why does it seem slew rate doesn't equal bandwidth or GBP?

Slew rate and GBP are two different ways to express how fast an opamp is. They cover different aspects of speed, which means that both are needed and there is no way to convert one into the other. GBP tells about the low-pass response at a given gain, assuming that the signal level is so low that it does not limit the output ('small signal model'). But if you have large signals, another limitation is the maximum change of output voltage per time unit, the slew rate. The output simply cannot change at a higher rate, regardless of the frequency. In design, you have to test against both limitations: Is the BW OK to have my signal pass not attenuated? Is the required output signal slew rate always below slew rate of the opamp?

[RoGeorge]

Unitary Gain Bandwidth tells the max frequency at small signal.
If you need a large signal, you have to make sure the output can swing fast enough.

For a sinusoidal waveform, the max rate of change in the signal is during the zero crossings.  Minimum required slew rate is dictated by both the frequency and the amplitude of the output signal.


For example, if you need an amplitude of A=1V and f=1MHz, the slew rate has to be at least Smin​=2πfA = 2*3.14*1MHz*1V = 6.28 V/us (the fastest swing appears during the zero crossings).  Now, if you need the same 1MHz, but at an output of 10V peak to peak (so A=5V), then the slew rate of the opamp has to be at least 31.42V/us.

Notice how both signals have the same frequency, 1MHz, but because the second signal is 5 times larger, it will require a 5 times faster slew rate.  An opamp with GBP 100MHz but only 10V/us SR, won't be able to output a 10Vpp at 1MHz.

[MariuszD]

A faster amplifier will have both parameters better, but the slew rate is not strictly related to GBW. It is possible to make an amplifier with a high GBW and a low slew rate. So there's nothing strange about such amplifiers. Old audio amplifiers built on transistors had this flaw, which caused TIM distortions. The cause was the input stage current being too low.

[Andree Henkel]

you want 400kHz reasonable square wave, 6V amplitude,
"reasonable" is lets say 1/10th of period can be rising/falling edge, but you have to specify what is "resaonable"
so you would have 1/400kHz/10= 0,25µs ; 6V/0,25µs would mean your OP needs equal/better than 24V/µs slew rate
for the 6V swing you need at least +/-5V dual or 10V single supply

I made a quick search on digikey.
Looks like TLE2072 or TLE2082 are two nice yellybean options for you, close to double the 24V/µs.
If you want faster it gets expensive.
https://www.digikey.de/en/products/filter/linear/verst%C3%A4rker/instrumentierung-operationsverst%C3%A4rker-pufferverst%C3%A4rker/687?s=N4IgjCBcoGwJxVAYygMwIYBsDOBTANCAPZQDaIALGGABxwDsIAuoQA4AuUIAyuwE4BLAHYBzEAF9CAVmqIQKSBhwFiZEACYAzPU1TGhMFLjr6EAzQowK6kAbgUaMAAy2NhzVVfrvHsxs009PQUXvQ0YPau8DA0NgYUTnR%2BYA5GNK7uRlIZsTROmhlwmt76bgEw2SwgHFy8gqISBupwENDyaFh4hCSQ5M2acE7ZhP0tLiNFLXEa9k7NXvYUlgtLFSv0UpVsnJA8-MJikuBwdHIKSl2qvSAuTEfqTpFt550qPX3M21wAcgCuALa4QQoe4wMAuNo1XYAVSEAnYAHlUABZXDobC-Pi4Vy-LgANVc-2EXHBjRAAFobM8oPxfm81JVxEdspCoGA2Kw2U4nEcYHIBAATLjk8F%2BKEgVwAR3YAE8uMMQLLWNjduiQeIgA