Interpreting GFV High Vacuum System Calibration of Hot Cathode Ion Gauges
High vacuum measurement
systems consist of a tube and a controller.
The tube is mounted on the vacuum system at the point where the pressure
is to be measured. A cable connects the tube to the controller unless tube and
controller are combined in the same package.
The controller provides the voltages necessary to run the tube and
displays the signal from the tube in appropriate pressure units.
Hot cathode ion gauge tubes consist of a heated
filament maintained typically at +30 Vdc external to a grid enclosure maintained
typically at +180 Vdc. Electrons
from the filament pass into the grid enclosure and ionize gas molecules.
A central rod maintained at ground potential collects the ions.
The ion current is proportional to the gas pressure in the tube.
The relation between ion current and pressure is different for every
tube, even for tubes of the same construction.
The relation is termed the sensitivity or "gauge factor" of the
tube. The calibration process
measures this relation.
Hot cathode gauge high vacuum system calibration
Figure 1. Typical high vacuum calibration. The straight line is zero error.
Figure
1 shows a calibration curve for a hot cathode high vacuum measurement system.
The curve is typical of those we have observed over hundreds of
calibrations and scores of systems.
At
low pressure, 10-7 Torr range, there may be considerable discrepancy
between the zero error line and the calibration points;
In
the mid range, 10-6 Torr and 10-5 Torr range, the
calibration points fall close to the zero error line.
Mid range result
The
mid range result is the easiest to explain: the output from hot cathode gauges
in this range and below is intrinsically linear with pressure.
If, as required by our procedure, the control is set at one point in this
range as close to zero error as possible, other points taken in this range will
show close to zero error.
Some
controls do not provide for an adjustment for the sensitivity of the tube.
The calibration curve in mid range for these systems will typically
resemble Figure 2. The curve is
linear but displaced from the zero error line because the control assumes tube
sensitivity different from the actual sensitivity of the tube.
Figure 2. High vacuum calibration resulting from incorrect setting for tube sensitivity.
Low range result
Although
these gauges retain the linear relation of output to pressure in the low
pressure range some randomness results because of the way in which the output is
calculated. The calibration vacuum
system is pumped down to an end vacuum, typically displayed as 1×10-6
Torr. In order to determine the
system response to nitrogen, the end vacuum indication or "background"
is subtracted from the signal observed when nitrogen is bled into the system.
Lack
of precision in the control's display of pressure produces random departures
from zero error. Typical controls
show only one figure beyond the decimal point in the pressure mantissa; analog
meters are no more precise. Thus if
the background indication is 1.0×10-6 Torr and the first measurement
point shows 1.5×10-6 Torr the difference is 5×10-7 Torr.
Jitter in the last figure is typically one unit, which allows, after
subtraction, for a variation in the result from 4 to 6×10-7 Torr for
the same pressure, resulting in an error of ±20%.
Controls that display two figures after the decimal point will show significantly less randomness in the results for the low range.
High range result
At
pressure greater than 1×10-4 Torr significant departure from
linearity of hot cathode tubes occurs, particularly for Bayard- Alpert type
tubes with a slender collector rod. The
electric field that attracts the ions to the collector is concentrated around
the collector; ions formed some distance from the collector travel very slowly.
The space charge built up at high pressure deflects the ionizing electron
stream in unpredictable ways that change the sensitivity of the tube.
How to use the calibration curve
Your
controller may have no function, which allows the controller to accommodate
tubes with various gauge factors. Neither
the gain of the electrometer amplifier or the emission current used by the
tube may be set to compensate for different gauge factors.
When
used with typical ion gauge tubes the pressure readings provided by the
controller will be significantly different, as much as 50%, than the true
pressure. However, readings are
off by a fairly constant percentage; they may be corrected by applying a
correction factor characteristic of the tube and filament in service.
To
calculate the correction factor, average the percent errors reported in the
"% Error" column of the calibration data sheet for the tube for mid
range pressures. The correction
factor will be 1/(1+E) where E is the average percent error divided by 100.
Thus for a typical tube filament the average percent error is -30%.
The correction factor is then 1/.7 = 1.43.
The correction factor is then multiplied by the pressure reading from the
control. Thus if the control reads
1.7×10-5 Torr, the true pressure will be