An important consideration when making a choice of low level measuring instrument is the range of source impedance with which it will be required to operate.

All electrical conductors have resistance at normal temperature and thus they generate thermal noise, proportional to the square root of the resistance, which varies similarly with temperature.

This noise defines the limit of measurement. EM low level DC voltage measuring amplifiers approximate to the source thermal noise characteristic for two to three decades of source resistance, and in the case of the A20 and N11, for about four decades.

When making a selection, a knowledge of the range of source impedance to be measured would be very helpful. All EM Nanovoltmeters, Picovoltmeters and Nanovolt Amplifiers are able to operate down to zero source impedance and the noise they generate is then their equivalent noise resistance (enr).

As the source impedance is increased, the source noise power is added to that of the instrument, and begins to become significant when the source noise approaches the equivalent noise resistance of the instrument. As the source impedance is increased still further, the total noise is only slightly higher than the source noise until, being the sum of the constant instrument noise and the noise from the source, which will be proportional to the source resistance. At some source resistance, the total noise begins to increase more rapidly than the source alone, i.e. at about 20dB per decade of source resistance.

The point at which the two characteristics begin to diverge is the maximum source resistance at which the instrument will operate without adding significantly to the source noise.

One of the important uses of ultra low level DC nanovolt amplifiers is the measurement of cryogenic sources. Much research is being undertaken into high temperature superconductivity. If a dc nanovolt amplifier is required to measure a source of, say, 100 ohms at 3 degrees K, it must cope with the source impedance of 100 ohms. However, if the source is at a temperature of 3 degrees K, then the noise of the source is equivalent to the noise of a 1 ohm resistor at room temperature. This makes the design of the amplifier very complicated. The measurement of this parameter is the ability of the nanovolt amplifier to cope with source resistance much higher than would normally be required for the measurement of low source resistance. EM has worked on this problem for many years and we have improved the cryogenic capability to a very large extent. The models N11 and A20 can operate with such a wide range of source resistance that their characteristics at room temperature can be extrapolated towards the source noise characteristic at helium temperature, and it can be seen that the instruments, while remaining at room temperature, can efficiently measure a source at helium temperature, over a small range of source resistance, and are useable over a wide range of source resistance.

Care should be taken when selecting a DC Nanovolt Amplifier. The model A22 has an equivalent noise resistance of about 100 ohms, or the thermal noise generated by a perfect resistor of 100 ohms. As a general rule, if the resistance of the source to be measured, including the input leads is much greater than the enr of the amplifier, a lower specification amplifier may be the more economical choice. For example if the source to be measured is 30 ohms, then the DC Nanovolt Amplifier model A23, which has an enr of 30 ohms, will give a total noise of 60 ohms or 1nV per root Hz rms. If the source resistance is 100 ohms then the total noise from the source plus A23 would be 130 ohms or 1.5 nV per root Hz rms. If the A22 were used, which has an enr of 100 ohms, the total noise would be 200 ohms or 1.8 nV per root Hz. Once the source impedance increase beyond 100 ohms, the A22 would be the better choice. A similar case is between the choice of the A23, A10 and A30. Short circuit noise performance is by no means the only parameter to be considered. It is strongly recommended that, before selection, the graphs of noise against source resistance be studied carefully. These graphs are included in the data sheets for the instrument.

EM manufacture a range of amplifier modules, some of which are miniature units for mounting on to a printed circuit board. The range of equivalent source resistance is from about 700 ohms to less than 1 ohm. The DC Picovoltmeter model P13 will detect levels down to a few tens of picovolts.

Amplifiers can be made with lower noise still for special applications.