In a previous research we discuss the design and performance of J-FET discrete input ultra low noise preamplifier with voltage noise less than 0.3nV/ЦHz . The circuit has significant benefits compared to the simple op-amp use but it is not suitable for some low scale and low power devices. Front transistor used in our prototype discrete input amplifier was the J-FET BF862. Unfortunately it was discontinued by NXP and stopped from manufacturing at the end of 2017. The good news is that ON Semi  still produces a replacement part, which uses the same footprint with signature number 2SK3557. In the previous research  the focus was on the parameters of the existing transistors suitable for low noise applications. Here we will discuss the op-amps suitable for the low noise applications.
The aim of this paper is to model the inductive source and to analyze the factors that produce the noise in that kind of high impedance inductive sources. The input sensor is a magnetic head that generates voltage proportional to the magnetic flux recorded on the magnetic media. Still there is lot of applications that uses that kind of media, for example the payment terminals using magnetic strip. We will investigate one of the hardest tasks – the amplification from the magnetic tape head. We will use the same source with different op-amps to analyze the signal-to-noise ratio at the output…..
The details about the introduction and the test bench can be found in the full text paper at the end of this article.
The first priority is to find out the physical limits that set how low the noise can be. With a purely resistive source it is easy to calculate the Johnson noise from the input source resistance. With a noiseless amplifier this would be the equivalent input noise (EIN), but real amplifiers have their own noise, and the amount by which the source/amplifier combination is noisier is the noise figure (NF). The inductive source (Fig. 2) is modeled as a resistance Rgen in series with a large inductance Lgen, and is loaded by the resistor Rin…..
The details about the model and the simulation can be found in the full text paper at the end of this article.
The simulation model shows that the AD797, which is rated as one of the best ultra low noise op-amp, is not suitable for high source impedance. The measurements from Table 3 confirm that it is worst for the schematic on Figure 1. The comparison with the OPA2140 confirms that the AD797 is 9,1dB noisier than the OPA2140 for the input impedance of 1550Ω+650mH. On the other hand the simulation model with the same input parameters confirms that the AD797 will be 8.41dB noisier than the OPA2140. The 0,69dB difference between the simulation and the real experiment is caused by difference of the bandwidth. The simulation covers the frequency range from 15Hz-20kHz, while for the measurement is used wideband true rms microvolt meter and the frequency range was extended up to 25kHz. The practical measurement confirms the simulation very close. It can be concluded that the model is correct with its dependences.
If the design requires best noise performance then the described synthetic loading can be used or several JFET op-amps can be used in parallel reducing their input voltage noise.
 Low noise preamplifier suitable for high impedance sources, Sozopol, Bulgaria, 10.1109/ET.2016.7753510, 2016
 ON Semi, 2013,
 Analog Devices, 2017,
 Linear Technology, 2013,
 AN-104 Noise Specs Confusing, Texas Instruments Incorporated, April 2013
 Douglas Self, Small Signal Audio Design, Focal press, ISBN: 978-0-240-52177-0, 2010
If you are interested in this research you can read the full text paper by the following URL. The paper describe precisely the test bench, the practical noise measurement from the real circuit and the simulation results. Also the paper investigate the noise floor of the high impedance source. You can follow the paper by clicking here: PID5503409