Application Note 7
Differences Among Different Model 100 Series Picoammeter Measurements
Glenn Fasnacht, President, AMETRIX Instruments
Introduction
There is not a one-to-one relationship among the user-requested measurement mode (data rate), the displayed measurement data rate, and the logged measurement data rate. This paper explains the differences and provides guidance on choosing the most appropriate measurement mode. There is a dark side to these TIA ammeters, however, and this paper delves more deeply into those dirty little secrets.
The actual measurements are raw ADC counts that have had calibration coefficients applied. They (sort of) arrive at the sample rate that the user selects.
Modes
There are two modes; one that offers higher accuracy over longer measurement times and wider temperature swings, and one that provides faster update rates. These are very general comments and don’t always hold true, so some further explanation is required:
1. Continuous
When one of the continuous modes is selected (7.7, 900, 1800, and 3600 samples per second (SPS)), the analog to digital converter (ADC) is continuously connected to the current measurement and each measurement is returned at the selected rate.
2. Background
When one of the background modes is selected (1.9, 39, 45, 150, and 300) the ADC takes extra measurements (backgrounds) in between current measurements. These background measurements include measurement circuit common, the voltage reference, and a few others. These background measurements allow for the correction of some time and temperature induced drifts.
While this provides higher accuracy over wide temperature and time, the switching of signals to obtain the background measurements does induce some noise, so if one is operating in a fairly stable temperature environment, and taking measurements only for a few hours or days, quieter measurements may be obtained by using one of the slower continuous modes; 7.7 SPS returns the quietest measurements.
Mains Frequency Rejection
Some of the modes within both the continuous and background groups also integrate the input signal over one power line cycle to provide improved rejection of Mains induced noise. Selecting 1.9, 7.7, or 45 SPS will reject 60 Hz noise, while selecting 1.9, 7.7, or 39 SPS will reject 50 Hz noise.
If one has well routed and shielded cables and is not suffering from Mains noise pickup, it is not necessary to stay in one of these modes. How can one know? Set the data rate to 300 or 900 SPS, log a few line cycles worth of data, and plot it, or perform an FFT on it; if no significant Mains noise is seen then you don’t have to use a mains rejection sample rate. This will help gauge the need for mains rejection. This will also help detect whether the slower measurements are contaminated with a waveform that is partially over driven during part of the measurement cycle.
In General
The slower the data rate the less noise one will experience. The background modes tend to add some small measurement-to-measurement noise but reduces long-term and thermal drift.
Displayed
Except for the slowest two modes, the data displayed on the soft front panel (SFP) is decimated for readability purposes, otherwise the display would update so fast that the least significant digits would be an unreadable blur.
The number of displayed digits is reduced at the higher sample rates because the high sample rate measurements have higher noise and the last digits are of no value.
Logging
The measurements that are logged are always at the user requested data rate. The logged data always have the calibration coefficients applied and return calibrated amps. If one chooses a background mode, the background measurements are used to correct the measured result and only the corrected, calibrated current is logged, not the background measurements.
No matter what the sample rate, the full measured value is returned; that is it is not truncated in the logging window like it is in the main display window.
Summary
Here is a graphic representation of actual measured data (figure 1). These data were taken with a capped input, so this is about as ideal as it can get.
Some observations from figure 1:
-
There may be no advantage to operating slowly on the higher current ranges
-
The background modes are not quieter; if you are in a stable environment you may choose not to use these modes
-
The Y axis is logarithmic so the differences are much more dramatic than they appear
