Here is a picture of what I did yesterday.
Torturing of the dilution refrigerator is the only way to diagnose what is one of the experimentalist’s worst nightmares…
‘The Intermittent Fault’…
I have had intermittent faults with this system for about 3 years now. The main problem arises from the use of tiny stainless steel coaxial cables (0.3mm outer diameter). You can’t solder to stainless steel without using an acid flux. If you use this, the connectors fall apart due to corrosion after the first few thermal cycles. You can crimp the connectors onto the coax (it’s fiddly) but then the problem is that thermal cycling can cause mechanical stresses which then just break the inner wire, (not difficult considering it’s thinner than a hair).
Anyway back to the torturing. The faults in the wiring only occur at low temperatures <- That’s the important point by the way. You can’t check continuity of a wire when it’s deep down in a sealed dewar immersed in Liquid Helium.
There are 4 signal lines along which these awkward faults can occur. The way to find them involves dipping the lower part of the dilution refrigerator slowly into an open bucket of Liquid Nitrogen until the fault appears. The temperature gradient above and below the liquid surface is high so you can accurately pinpoint the problem area in terms of vertical position. Using a heat gun (which looks similar to a hairdryer, but goes up to 650C, hehehe) on the part just above the liquid nitrogen surface helps increse the temperature gradient. It creates pretty Nitrogen fog too. (a bit wasteful, but LN2 is cheap).
The coaxes have now been temporarily replaced with twisted pair wiring as multiple connectors on 3 of the 4 lines were found to be faulty. It’s also a really great way to find dry solder joints. NEVER rely on the fact that you can create 100 or so solder joints that will all remain robust – even if you’ve been soldering since you were about 8 years old 🙂