Checking For Problems; Fixing The Problem - HP E1419A User Manual

Checking for
Problems

Fixing the Problem

Chapter 3
Thus far in the discussion, we've assumed that the low-level channel
measured after a high-level channel has presented a low impedance path to
discharge the A/D's stray capacitances (path was the thermocouple wire).
The combination of a resistance measurement through an HP E1501 Direct
Input SCP presents a much higher impedance path. A very common
measurement like this would be the temperature of a thermistor. If measured
through a Direct Input SCP, the source impedance of the measurement is
essentially the value of the thermistor (the output impedance of the current
source is in the gigohm region). Even though this is a higher level
measurement than the previous example, the settling time can be even
longer due to the slower discharge of the stray capacitances. The simple
answer here is to always use an SCP that presents a low impedance buffered
output to the HP E1419A's Range Amp and A/D. The HP E1503, 8, 9, 10,
12, and 14 through 17 SCPs all provide this capability.
The method we'll use to quickly determine if any of your system's channels
needs more settling time is to simply apply some settling time to every
channel. Use this procedure:
1. First run your system to make a record of its current measurement
performance.
2. Then use the SAMPle:TIMer command to add a significant settling
delay to every measurement in the scan list. Take care that the sample
time multiplied by the number of channels in the scan list doesn't
exceed the time between triggers.
3. Now run your system and look primarily for low level channel
measurements (like thermocouples) who's DC value changes
somewhat. If you find channels that respond to this increase in
sample period, you may also notice that these channels may return
slightly quieter measurements as well. The extra sample period
reduces or removes the affected channels coupling to the value of the
channel measured just before it.
4. If you see some improvement, increase the sample period again and
perform another test. When you increase the sample period and no
improvement is seen, you have found the maximum settling delay
that any single channel requires.
5. If the quality of the measurements does not respond to this increase in
sample period, then inadequate settling time is not likely to be
causing measurement problems.
If your system scans fast enough with the increased sample period, your
problem is solved. Your system is only running as fast as the slowest
channel allows but if its fast enough that's OK. If on the other hand, getting
Programming the HP E1419A 105