NEXT
User name Armstrong
Log entry time 19:42:40 on June14,2004
Entry number 125755
This entry is a followup to: 125715
Followups:
keyword=Target Fan Speed and Operating Temperature Study
Target Fan Speed and Operating Temperature Study
(David Armstrong, Jian-Ping Chen, Etienne Burtin)
- Started with 1x1 mm raster (at target, measured with spot++) to maximize "boiling"
- Target setpoint temperature = 6.6 K; this and the other inlet temperature
were constant throughout the test. JT valve at 58%, 4 K flow rate from CHL
was constant at 17.2 gm/s or so. Each of these were monitored throughout the test
to check they were steady.
- 28.2 µA beam current
- 20 cm He cell
- use PanGuin to look at asymmetry widths
- averaged the 8 Lumi widths "by eye". Non of the RMS values are regressed,
just raw.
- rough event ranges given here; but since target fan speed is in the data stream,
offline analysis should be able to select on fan sepeed.
HAPPEX Run # 2120:
Start Target fan T_Cell HP Htr Cell Detector RMS
Event # speed (Hz) (K) (W) Pres. Left Right RMS
(K) (psi) (ppm) (ppm)
---------------------------------------------------------------------
0 24 7.39 180 177 2580 2300 1950
15 K 39 7.25 140 172 2560 2200 1600
30 K 50 7.18 107 170 2480 2070 1400
38 K 60 7.16 77 170 2240 1930 1250
44 K 70 7.15 28 170 2300 1900 1070
Notes:
- clear reduction in "boiling" with increasing fan speed,
seen in detector and Lumi widths, when the raster is small
(and thus the boiling starts off large)
- Cell temperature decreases (we regulate on the input temp.,
not this one) with increase in fan speed. However, the high power
heater decreases, which means that the increased fan speed dumps
extra power in the loop. Going from 24 Hz to 60 Hz we dump an
extra 100 W due to the fan. The target density change corresponding
to the cell temperature change is negligible.
- estimate about 50 ppm measurement error - needs to be regressed
and analyzed carefully offline.
HAPPEX Run # 2121:
increased raster to 3.5 x 3.6 mm at the target.
Start Target fan T_Cell_After HP Htr Cell Detector RMS
Event # speed (Hz) (K) (W) Pres. Left Right RMS
(K) (psi) (ppm) (ppm)
---------------------------------------------------------------------
9 K 70 7.05 35 172 2080 1770 720
20 K 60 7.15 80 170 2065 1710 650
30 K 50 7.18 106 170 2105 1730 730
35 K 38 7.25 136 172 2160 1740 700
43 K 24 7.39 174 177 2050 1710 800
Notes:
- When the raster is big, so that the fluctuations ("boiling")
are reduced, it is unclear whether the fan speed has much effect.
Offline analysis with regression is needed to pull this out.
HAPPEX Run # 2122:
- Set fan speed to 48 Hz, beam current reduced to 5.4 µA
- Increased target setpoint temperature in steps: watched cell pressure and
high power heater:
Target cell Cell Press. HP Heater
setpoint (K) (psi) (W)
----------- --------- --------
6.6 169 243
6.8 180 271
7.0 192 280
7.2 203 310
7.4 215 330
7.6 225 358
Conclusion: Can gain a cooling power increase of 110 W by moving target
setpoint temperature up by 1 K. Here we have not changed density (pressure
increases, we did not vent any gas). At about 5.5 W/µA of beam power, this
corresponds to a gain of 20 µA in beam current for same input 4K helium
cooling gas.
Then we walked the beam current up under these conditions:
Beam Current Cell Press. HP Heater
(µA) (psi) (W)
----------- ---------- --------
5.4 225 358
9.1 226 327
14 227 317
20 228 270
25 228 251
28.4 228.5 212
So, with this cell temperature we could close the JT valve considerably
at 30 µA, or increase current by 25 µA and still have
headroom on the HP heater; and all this is at 17 gm/s from the CHL,
so there is another 20-25% available while staying under 24 gm/s.
I think we could get 75 µA onto this target without pushing too
hard. Of course, what the "boiling" would be like is an open question.
At this last operating point we looked at asymmetry widths again:
Left Right Lumi
(ppm) (ppm) (ppm)
---- ----- ----
2021 1816 660
Then we went back to a "bad" boiling condition, i.e. 1x1 raster (same 48 Hz fan
speed):
Left Right Lumi
(ppm) (ppm) (ppm)
---- ----- ----
2275 1950 1300
Comparing this with the earlier data with this fan speed, and beam current,
but with the 6.6 K setpoint, where we had widths of 2480/2070/1400 respectively,
there is a hint, at least, that the higher operating temperature also helps the
"boiling". A more systematic study is needed.
Finally, we went back to the original T_cell = 6.6 K, with 48 Hz fan speed,
3.5 x 3.6 mm raster for continued "production"