Differences

This shows you the differences between two versions of the page.

--- stm:manual_cryostat [2008/10/22 14:07] – galli
+++ stm:manual_cryostat [2009/05/18 12:36] (current) – galli
@@ Line 1: / Line 1: @@
 ====== Un-official Manual of Heliox_UHV (cryostat and 300mK fridge) ======
 ===== First cooling down =====
@@ Line 12: / Line 13: @@
   IMPORTANT!: Do not pump the Helium tank if ALSO the OVC is NOT pumped!
 ==== Pre-cooling with Liquid Nitrogen ====
@@ Line 24: / Line 29: @@
   - With capillaries and helium tank connected to helim, close the recovery speedy-valve located on the top of the helium tank. The helium tank is full with helium gas.
   - Pump on the 1K-pumping line.
-  - Open the 1K-NV (with Oxford Objectbench software connected to ITC503 MAIN) 50 or 80 %. Watch the pressure on the 1K-pot P-gauge. Close the valve to the 1K-pot pumo to see the pressure change (increase) faster. Do this 2-3 times to rinse the 1K-NV.
+  - Open the 1K-NV (with Oxford Objectbench software connected to ITC503 MAIN) 50 or 80 %. Watch the pressure on the 1K-pot P-gauge. Close the valve to the 1K-pot pump to see the pressure change (increase) faster. Do this 2-3 times to rinse the 1K-NV.
   - CLOSE 1K-NV again (0%).
   - Do same rinsing procedure with UHV-sock NV: open NV 1/2 or 3/4. Watch pressure on flow meter. Repeat 2-3 times. CLOSE uhv-sock NV again.
   - After cleaning all lines with dry helium, open 1K and UHV lines to recovery. Keep MAIN recovery valve closed and local recovery pressurized by the helium Dewar boil-off during the entire nitrogen transfer.
   - Close the recovery speedy valve on the helium tank!
-  - Quickly insert the nitrogen transfer sifon inside the helium tank, till it touches the bottom.
+  - Quickly insert the nitrogen transfer siphon inside the helium tank, till it touches the bottom.
   - Quickly fit the restrictor onto the blind flange of the helium recovery.
-  - Start transferring LN2 from 2 separate LN2 Dewars inside both helium and nitrogen tanks. It can be done at the same time. Start first slowly (~300-500 mbar) than you can speed up (0.8 to 1 bar).
+  - Start transferring LN2 from 2 separate LN2 Dewars inside both helium and nitrogen tanks. It can be done at the same time. Start first slowly (~300-500 mbar) than you can speed up (0.8 to 1 bar). Normally this can take up to 2 hours. You can monitor the magnet temperature with a multimeter: ~ 300 Ω at RT, ~ 400 Ω at 77K.
+  - Leave the nitrogen inside for 1 night. After 1 night the liquid nitrogen can be expelled from the helium tank:
+  - Use the gas outlet of one of the 2 nitrogen Dewars to pressurize the helium tank and kick the liquid out. Connect a long hose from this gas outlet to the exaust of the helium tank - in the meantime you have replaced the constrictor with a conventional "X-mas tree".
+  - The long hose that was used for the initial refill should be left to air or better to an empty LN2 Dewar, so that you don't waste the LN2 coming out of the cryostat. It is very important now that the siphon is really touching the bottom of the helium tank, otherwise some LN2 will remain.
+  - Open the gas outlet of the LN2 Dewar, slowly first than increase the pressure. LN2 should immediately start coming out of the tank. When no liquid comes out anymore, flow some dry N2 gas for 5-10 minutes to be sure that no liquid remains. The magnet temperature might/should rise a bit too (380 Ω or so).
+  - Remove transfer line and close the tank. Remove "Xmas tree" piece and blind the tank with a flange. Open the valve to recovery, after having switched to the helium pump (1K-pot pump). Pump the tank and keep an eye on the pressure (use 1K-pot P gauge): if the recovery tube freezes and the pressure does not drop below 10mbar, STOP PUMPING. It means that there is still LN2 in the tank, and it should NOT be removed by pumping but by flowing more warm gas longer. In any case, if the siphon was lowered fully till the bottom in the previous steps and warm nitrogen gas was flown for 10 minutes after all nitrogen came out, it's very unlikely that liquid nitrogen is still inside.
+==== Cooling the magnet to 4K ====
+  It is very important to flush the helium tank, the capillary and rinse the needle valves with dry-clean helium gas before cooling down, or the capillaries and needle valves will freeze. The whole tank should be pumped and flushed several times in order to pump away all Nitrogen gas which will otherwise freeze during liquid helium transfer.
+  - Be sure that there's enough helium overpressure in the recovery and in the tank. Then insert liquid helium transfer line (the half that goes in the cryostat) with the usual procedure that avoids transfer line blockage.
+  - Transfer helium initially slowly - the magnet takes time to cool and it should not be subject to thermal shocks (NbSn magnets are brittle!).
+  - At the beginning of the transfer, open 1K-needle valve (something like 80%). You might also open the UHV-sock needle valve a bit. The 1K-pumping line and the UHV-sock pumping line should be open to helium recovery.
+  - The magnet temperature can be monitored by reading the thermometer resistance: ~400 Ω at 77K, when liquid helium starts to collect at the bottom the resistance immediately rises to 1 KΩ and finally when the magnet is fully submerged in liquid helium the resistance should read ~ 4 KΩ.
+  - Initial transfer normally requires ~ 100 liters (12 Kg).
+  - Set the correct flow in the UHV-sock, by opening the UHV-sock needle valve and pumping on it with "Saskia". Set the needle valve accurately such that the correct flow (see flow meter) is achieved.
+  - Check for 1K-NV flow: pump on 1K, (open 80%), wait a little while (the 1st time can take up to few minutes) and see whether the pressure in the 1K line increases. Now close 1K-NV.
+===== Cooling the STM to 2K =====
+  - Cooling of the 1K pot can be started after the system has reached ~ 100K by radiative cooling from the UHV-sock, i.e. does not make sense to cool the 1K capillary and pot above 100K (very helium wasting and not really useful.) After first liquid helium transfer it takes 2 up to 3 days to reach 100K on the He3 pot (STM).
+  - At T-He3~100K, start cooling 1K capillary, sorb pump and finally 1K-pot: pump on 1K line and open 1K-NV. Initially open ~80%, than shortly after close to 40% and finally to 30-28%. It might need extra flow (~40%) initially to cool down faster.
+  - When 1K-pot is cold (~4K or better less), wait few hours for the whole system to thermalize (better 1 night).
+  - Now you can open V1 from the dump. V2 should be closed if all He3 is in the dump. Otherwise is half of the He3 is in the smaller dump, also open V2.
+  - Now wait for the pressure in the dump to drop below 400mbar at least (can take several hours).
+  - Close V1.
+  - Set T-sorb ~ 28K slowly (use "rampsorb.mac" macro, or if by hand, ramp slowly such that it will take ~2 hrs for the sorb to reach 28K). When enough He3 starts to desorb and the pressure in the He3 line rises (never get above 2.5 bar - then the He3 will be dumped back in the dump via the overpressure valve!) the He3-pot temperature and the STM temperature will start to drop below 100K. This will warm up the 1K-pot (because the 2 pots, 1K and He3, are in thermal contact now) quite a bit. Increasing the flow in the 1K capillary does not always really help and just increases helium consumption.
+  - Below 20K the temperature will finally drop a lot faster, finally to about 2K. In general measurements can be performed at this temperature rather easily. The advantage is that with continuous flow in the 1K-pot (T-sorb=28K or less), 2K can be kept continuously as long as there's enough liquid helium in the tank. NOTE: the 1K-capillary and the UHV-sock start to run empty when liquid helium level is less than ~20%.
+===== Cooling the STM to base temperature =====
+  - IF the sorb is already cold (below 5K) open V1 to condense all the He3 left in the dump, if any.
+  - Close V1.
+  - SLOWLY warm up the sorb again to 28K-30K.
+  - Adjust 1K-pot flow so that the 1K-pot temperature is 2K or less, and the flow is as quiet as possible (adjust it as close as possible to the 1K-flow indicated on the flow meter).
+  - Wait at least 2 hours (waiting less decreases the base temperature hold time to about 3-4 hours or less). All He3 will condense in the He3 pot.
+  - Set T-sorb=0K. The He3-pot temperature and the STM temperature will drop rapidly (less than 30 mins) to base T (usually ~280mK on the He3 pot, 6 to 20 KΩ on the STM thermometer). Base T hold time with the STM ranges between 3 to 30 hours, depending on the amount of He3 condensed.
+===== Useful "tricks" and recommendations =====
+  - Approach with the STM (go in tunneling) at 2K. Approaching at 300 mK can warm up the He3 pot from slightly to significantly (depending on the amount of condensed He3 and how long it takes to approach).
+  - Setting the magnetic field at high rate causes severe warming up of the He3-pot due to Eddie currents. Use the recommended rate of XXX.