Starting point: new sample and new tip are loaded in the STM. Tip-sample distance is “large” (i.e. few millimeters). STM is in lowest position, inside the centering pin and 1K-shield-window is open. 1K pot flow is set for normal cooling operations (~6-7mbar). All 3He is inside the system and V1 is closed.

1. Position and focus USB camera such that you can see clearly sample and tip holdres (start software “MicroCaputre” on Labview/T-control computer)
2. Femto preamp: check back side of RHK, switch to BNC. Start Labview VI: STM Conductance. Set correct gain.
3. RHK premp: check back side of RHK, switch towards D9.
4. Prepare RHK controller: XYZ offsets in the middle (5.0). Set initially: time constant ~0.2, gain ~0.3. Set appropriate bias and current set points.
5. In RHK software, set Waveform “Parabola” or “Sawtooth” and “Approach steps = 1”. Click “Approach” button.
6. Use gain as Z-piezo motor speed control. When the tip and sample gap is very large, set gain ≥ 0.5. Observe if the Z motor moves fast enough. Keep the motor approaching “fast” until the sample gets close enough that the tip-sample gap can't be seen anymore (either via the camera or by eye looking into the view port).
7. When the gap gets small enough (less than 1 mm), slow down the Z motor by decreasing the gain. Decrease the gain till it's very hard to see the motor moving. Now the motor has to move very slowly, or the tip can crash. This is the fine approach.
8. When the fine approach is finished successfully (software will pop-up a window), set gain~0 and time constant~2 before releasing the tip via the software.
9. It is not advisable to scan when the STM is in the pin. Retract therefore 3 x 100 steps (Waveform “Sawtooth” and “Adjust steps = 100”, press -) and send the STM up with the “lift”.

It is possible to approach immediately after the STM has reached the top position (measurement position in the middle of the magnet), but there will be initially a lot of thermal drift as the sample/tip are initially at room temperature. They will thermalize rather soon, within 1 hour, at about 40~50K (there will be less drift then).

1. Before approaching: set the maximum scan area (300×300 nm^2). If needed, change the XY Gain with “Range”. Do this out of tunneling, before approaching.
2. Follow the fine approach procedure as above, with 1 step and low speed.
3. Scan at relatively low tip speeds (high scan rate, 1.5 or 2 sec per line). There will be a lot of drift as sample and tip are cooling down.

At this point (when tip and sample are loaded) the cooling procedure can start. The STM will be cooled first to about 2K by desorbing and condensing the 3He that is now inside the sorb pump.

1. Set very slowly the sorb-pump temperature to 20K-25K.
2. When the 3He temperature starts to drop rapidly towards 20K, perform a careful fine approach (now 2 steps might be required for “Approach steps”).
3. Remain in tunneling if possible (signal is quiet, no one is using the crane).
4. Otherwise retract 1 x 100 steps.
5. When the 3He temperature starts to drop rapidly towards 10K or less, perform again a careful fine approach (again 2 steps might be required for “Approach steps”).
6. As above, remain in tunneling if possible or retract 1 x 100 steps.
7. At 2K, perform again a fine approach. Remain in tunneling if possible.
8. Slowly rise the sorb-pump temperature to 35K.
9. Retract 1 x 100 and adjust if necessary the 1K-pot flow to achieve 2K on the 1K-pot. Do not adjust the 1K-pot needle valve in tunneling, as it can crash the tip!
10. As the sorb pump is heated to 35K, the signal (in and out of tunneling) will be noisy.

1. After 1 hour or 1 hour and 15 minutes, stop condensing by switching off the sorb heater: T sorb can be set to zero.
2. Initially the temperature will stay at ~ 2K for a while before starting to drop very rapidly to base temperature.
3. Perform a careful fine approach when the 3He temperature is still at 2K, and remain in tunneling down to 300mK (to avoid approaching at 300mK, as it gives some heat load).
4. After ~ 45 minutes the 3He temperature will drop to 300mK and the STM will follow (T-STM~20 KΩ)
5. Normally there is no need to re-approach, as there's very little drift from 2K.