Sonsivri

Hi

i'm looking for a NE5534 replacement in order to:

work with +/- 18 or 19Vdc for the opamp supplies.
Input signals from 0 to 50KHz

Do you have any advice?

I would make a daughter board so you can use two dual opamps. Then you have a huge selection to work with. Assuming audio use just google for "NE5532 replacement"

It really depends on your application. The designer chose the part for a particular attribute in that application and that's what you should consider with a replacement.

For example, two dual op amps would be a disaster if matching of transistors in the amps is critical.

Well, the application is not "audio" related
I have some sensors on which i perform a frequency sweep and a fft analyse with a mcu (for one is a 1-100Hz for other is 20K-50K sweep).  The input signal must be in the order of +/- 16V (this is one of the reasons why the first choice was the low noise ne5534)
Gains are in the order of 1 (only buffers) to 8-10 from a DDS.

It will always boil down to what you can get, like hold in your hand. Another thing is supply voltage. It is bad design idea to say use a opamp with absolute supply voltage 15 volt if your system supply voltage is 15 volt. What you can do is to compile a list of components you know you can get. that way it will be much more easy to help

I have used the MC34071, which is pin compatible with 5534. Has fairly good specs. It will run up to +/-22v   

for some variants is +/-20V  supply max before frying :)
https://www.onsemi.cn/pub/Collateral/NE5534-D.PDF

the choice of 16V stimuly is not mine, btw...  but i gotta deal with it.

You can use ADA4096-2. It has has 32 volt supply range with low offset and high open loop gain parameters.

^If only an AFE was that simple.

Do you need a pin compatible drop-in replacement or is this something that you build. And therefore have much more degree of freedom

What is the objective here? If it is just operating voltage range then you can use selective search functions most producers and suppliers provide. Just select the max and min values.
If it is the slew rate, you are talking about a very achievable range (up to 50kHz). If it is some audiophile considerations, it is mostly subjective; you will find lots of people who would be happy with the specs of NE5534. There will always be some for whom even the best chip simply won't cut it. If it is about driving a certain load then it is also matter of searching the aforementioned sites this time specifying the load.

Let's take a look at the specs of OPA1678. Anything the producer boasts about on the front page is potentially something to use for comparing this to other chips. A good audio op amp must offer low distortion, low quiescent current, low noise, high open-loop gain, high common-mode rejection, if you have a high Z input it must feature a low input bias current. If it is to be run on batteries it must have low quiescent current and low current consumption versus output. A high slew rate is desirable. So is a very high signal to noise ratio which is also affected by how the system is put together.

OPA1678 (not necessarily meant for your needs)
Low noise: 4.5 nV/√Hz at 1 kHz
Low distortion: 0.0001% at 1 kHz
High open-loop gain: 114 dB
High common-mode rejection: 110 dB
Low quiescent current:

    2 mA per channel

Low input bias current: 10 pA (typical)
Slew rate: 9 V/µs
Wide gain bandwidth: 16 MHz (G = 1)
Unity-gain stable
Rail-to-rail output
Wide supply range:

    ±2.25 V to ±18 V, or 4.5 V to 36 V

Dual-channel and quad-channel versions
Small package sizes:

    Dual-channel: SO-8, MSOP-8, SON-8
    Quad-channel: SO-14, TSSOP-14, QFN-16

He already said it was NOT audio. Which you missed and that miss sent you off into the weeds.

He already said it was an AFE for a sampling system, though he didn't give full details on the sampler or Nyquist filtering which makes the AFE component selection next to impossible.

He already stated the high voltage input requirement and unity gain which is an ADC headscratcher.

He already stated some channels were merely unity gain buffers, where others had modest gain.

He already stated he's doing frequency domain analysis which adds its own headaches.

So, despite useful info in an audio application that somebody here might find applicable to their own application, take another stab at it. I know you're good, you merely missed some critical info.

@sideshow bob, is a custom board to communicate with a stm32 and another board with sensors.  They need a +/- 16V sine stimuli in order to minimize SNR ( also a 0-8V step signal is needed for another test) and frequency sweeps dirfferents for the sensors.

@solutions, yep i gotta replace the actual testbench that consist on a picoscope ...  i ported most of the calculations performed from python to c for the stm32 but i gotta redo it later for callibration (recovering the reference signal at the input and verifying local gains before even launch the algorithm :p

The other part of the circuit recovers signals in order of +/-2 V   and +/- 20mV scale...  and the former +/-16V sine...   i'm using some AD8227 for the instrumentation part in order to simplify and some low pow rail to rail for the rebiasing and out gain...      I'm afraid if i show what  the schema looks you will hit me with a giant hammer :)

Rubber hammer ready...go for it, lol

Hammer time

for example this goes for the input  20mV range signal