From the Sept/Oct 2010 tests, I found that new, sensor-specific, calibrations of E.lambda and Tau63_ref from the agar test produced results with less scatter. Thus, the plan is to use these for data processing. Since I will need to rely on these values, it seems like a good idea to do agar post-cals on these probes. In addition, I might as well double-check the probe resistances, so Chris is now making a test cable that I'll use with the HP5006 plug-in 4-wire meter. With these revised calibrations, I'll go back and look at the PCAPS data when the soil wasn't frozen. Finally, we'll rerun some tests in the FLAB yard. I'm not planning a new set of keg tests.
Resistance values measured 15 Apr 2011 using 4-wire meter:
ID |
Sensor S/N |
R_measured (Ohms) |
R_manufacturer (Ohms) |
Difference (Ohms) |
---|---|---|---|---|
1 |
200589 |
16.14 |
16.1 |
0.04 |
2 |
200590 |
15.40 |
15.4 |
0.00 |
3 |
200236 |
14.79 |
14.8 |
-0.01 |
4 |
200591 |
16.70 |
16.7 |
0.00 |
5 |
200239 |
16.37 |
16.2 |
0.17 |
6 |
200592 |
15.53 |
15.5 |
0.03 |
7 |
200241 |
16.27 |
16.2 |
0.07 |
8 |
200242 |
16.10 |
15.9 |
0.20 |
9 |
200243 |
15.76 |
15.6 |
0.16 |
10 |
200234 |
14.72 |
14.7 |
0.02 |
11 |
200238 |
15.50 |
15.4 |
0.10 |
|
200235 |
sliced |
14.8 |
|
|
200240 |
lightning fried |
15.0 |
|
All are within 0.2 ohms of manufacturer values. Except for one at 0.01 ohms, our values are higher. <Might> be connectors, but 4-wire measurement should compensate.
Agar results (first row Sep 2010, second row Apr 2011). Lambda values use original Hukseflux coefficients in this table:
ID |
Sensor |
Vheat (uV) |
Vpile.on (uV) |
Vpile.off (uV) |
V63 (uV) |
Tau63 (0.01s) |
Lambda (mW/m K) |
---|---|---|---|---|---|---|---|
|
Expected |
1200000 |
|
0 |
|
1900 |
600 |
1 |
200589 |
1236927 |
375 |
0 |
139 |
1943 |
666 |
|
|
1236416 |
375 |
0 |
139 |
1921 |
665 |
2 |
200590 |
1194264 |
314 |
0 |
116 |
1940 |
668 |
|
|
1194652 |
314 |
0 |
116 |
1985 |
669 |
3 |
200236 |
1238958 |
313 |
0 |
116 |
1877 |
578 |
|
|
1237230 |
313 |
0 |
116 |
1892 |
576 |
4 |
200391 |
1240328 |
313 |
0 |
116 |
2073 |
686 |
|
|
1239126 |
313 |
0 |
115 |
2042 |
684 |
5 |
200239 |
1239719 |
470 |
0 |
174 |
1965 |
591 |
|
|
1238975 |
470 |
0 |
173 |
1972 |
590 |
6 |
200592 |
1237441 |
407 |
0 |
150 |
2125 |
655 |
|
|
1237288 |
407 |
0 |
150 |
2170 |
655 |
7 |
200241 |
1241515 |
376 |
0 |
139 |
1892 |
596 |
|
|
1239874 |
376 |
0 |
139 |
1867 |
595 |
8 |
200242 |
1236806 |
407 |
0 |
150 |
2089 |
600 |
|
|
1236332 |
407 |
0 |
150 |
2105 |
600 |
9 |
200243 |
1236372 |
438 |
0 |
162 |
1746 |
598 |
|
|
1236324 |
438 |
0 |
162 |
1755 |
598 |
10 |
200234 |
1237511 |
281 |
0 |
104 |
1679 |
603 |
|
|
1237304 |
281 |
0 |
104 |
1673 |
603 |
11 |
200238 |
1238329 |
439 |
0 |
162 |
1773 |
586 |
|
|
1236740 |
439 |
0 |
162 |
1792 |
584 |
All of these have very close agreement between the Sept 2010 and Apr 2011 values. Tau63 differences are 0-0.5s, Lambda differences are 0-2. Thus, best approach is to take the average of these values to generate new Tau63_ref and E.lambda values.
Thought I'd start by scribbling down the Agar recipe:
Add 7g agar flakes to 1.75l water in 2l beaker.
Place on 1000W burner of 2-burner hotplate.
Rig up a shroud and lid around beaker to reduce heat loss. I borrowed a chunk of pipe and PVC from the shop this time (see photo). The spot I used in the cage had a draft, so I used the blue foam to block it.
It took about 20min on high to get to a boil. I stir every couple of minutes using a glass rod Chris found to the left of the cal lab sink.
Reduce heat to medium-high and simmer with lid off (packet instructions said for 5min, TP01 said 20min, I did 15min this time). I noted that the volume reduced to 1.6l after all this time, which is still fine.
Let cool until milky white (several hours).