Noise hunting on the LG48 K100 fridge

Ted's notes from 1D hole wires Sept 2009

Ted's wires were not showing good steps, due to invisible noise (not visible in the lockin):

Ted's wires were not showing good steps, due to invisible noise (not visible in the lockin):

Problem: The sample-box wire-lead to the Fischer connector must be physically pushed tight against the connector, or it leaks radiation down into the fridge.

Solution: Use a stand-up clamp to physically hold the wire-lead such that it applies pressure in the proper direction against the fischer connector.

Once this was done, the sample current dropped by 15% and the steps returned!

Oleh's notes on 1D wires Aug 2009

Problem and Symptoms:

• Quantum wires show no plateaus
• Noticed that the MC temperature on the Oxford controller does not go as low as before (stays around 23 mK compared to 18 mK previously).
• Noticed : when the sample box is grounded and ground from the SMU or from the lockin is connected to the box, the MC temperature is rising.
• It is possible to reduce the “heating” by grounding the diagnostics plug to the cryostat. The “effective” temperature at the bottom of the cold finger is measured ~160 mK with CBT (when the CBT was in a copper box).

Tried many setup configurations.

Solutions:

1. Check that no pipes are touching the cryostat.
2. Make sure that the clamps and o’rings going to the insert are plastic.
3. Make sure that the sample box is grounded to the insert.
4. Disconnect the diagnostics and connect the battery box to still heater (pin 21 and 22 on the break-up box, current ~3.4 mA).
5. Make sure that the magnet power supply is connected to the mains and there is a grounding cable connecting the power supply and the cryostat.

The “effective” temperature at the bottom of the cold finger is now measured ~100 mK with CBT. The Kelvinox IGH now reads 18mK.

-- Oleh Klochan - Aug 2009

Data from Suddho's fridge run June09

Suddho measured a few atom Si:P quantum dot with:

• Standard Set Up
• Diagnostic cable unplugged
• Battery operated external still heater connected (but not grounded)
• Battery operated external still heater connected (and grounded)

Data from Oleh's fridge run Nov-Dec 09

Data to follow, but for now the network router in LG 48 near the 2 measurement computers caused lots of noise when trying to access network from one of those computers. Typical noise values were ~ 2 pA peak to peak. When accessing the network noise was exceeding 10 pA peak to peak. The router was removed and the problem was fixed.