During the summer 2004 running, it
became apparent that the temperature sensors for the LH2 and LD2
cryotargets may have been faulty. This was evident from the fact that
it took a LOT more cooling power than usual to regulate the LH2 target,
for example, at its usual 19 K.
After the summer run, James Brock from the target group checked the
calibration of the CERNOX's used on loops 1 and 2 (Loop 1 was LH2 and
Loop 2 was LD2).
Below are plots for each loop of the temperature as measured by the
sensors used during the running (y-axis "Ch.1" and "Ch.2") vs a well
calibrated, reference sensor. The top plot is Loop 1 and the
bottom is Loop 2. Overall, one can see the agreement is pretty good for
Loop 2, but off by several degrees for Loop 1.
The data files can be found here (Loop1_calcheck.txt
For the hydrogen target (Loop 1), the PID was locked to channel 2 at 19
K. From the calibration table, it looks like 19 K on Ch. 2 corresponded
to about 14.78 K. At 19 K, our nominal target density is 0.0723
. At 14.78 K the density is 0.0763 g/cm3
- so about 5.5% thicker.
On June 28, Chris Keith changed the PID to regulate the temperature
(according to CH.2 CERNOX) at 21 K - this means the real temperature
was 16.90 K. At this temperature the density is 0.0744 g/cm3
or 2.9% larger than usual.
The deuterium target was locked to channel 1 at 22 K. James' table
indicates that the calibrated temperature is then 21.95 K, smaller than
the 0.1 K error we typically assign to the temperature measurement. For
the record, a 0.1 K change in the target temperature would change the
target density by about 0.16%. The nominal density at 22.0 K is