Scheme for mounting detector, cold hardware, electronics without Soudan tower

Nick's phonon channel noise studies suggest that our noise is dominated buy Johnson noise in the SQUETs. This would be improved if the wirebonds in the SQUID card were superconducting. They certainly are not in the current setup using a Soudan tower. The SQUID card is mounted to the 600mK (Still) stage of the Soudan tower, which is heat sunk to our still by means of the 600mK rod. Heat leakage between tower stages is significant. Our still does not seem to run very cold (which has always been a bit of a mystery) and the 600mK rod may not provide a lot of cooling power.

Here we consider a scheme for mounting detectors and cold hardware without the tower. The tower provides wiring in vacuum coaxes from the MC temperature stage all the way to the 4K stage. This would basically be replaced with twisted pairs. It is pretty clear from experience that this will work just as well, for phonon channels - but maybe not for charge channels. So this scheme is proposed for phonon-only detectors. We'd use the SQUETs but not the side coaxes.

The sketches below contrast the Soudan tower scheme (left) with a possible alternative (right).

In this idea we mount the SQUETs on the Coldplate stage. Superconducting wire connectors would be needed between the MC stage and the SQUET, and also between the SQUET and the striplines at 4K.

Nader said that the SQUETs would probably dump too much heat into the Coldplate when there are SQUID jumps. He suggests mounting them to the Still stage.

There are two reasons this may not work:

1. Mechanical difficulties. We don't have a can or plate for the 600mK stage, just the rod. What would we physically attach the SQUETs to? We'd probably need to build a plate and heat sink it with the rod - but we'd have to somehow attach this plate to a different temperature stage in a thermally insulated way. Maybe stainless steel rods?
2. SQUID temperature. It doesn't look like mounting the SQUID card to the 600 mK rod will cool it below the Al transition, 1.2 K, whether or not we need to worry about leakage between tower stages. The rod probably has limited thermal conductivity for one thing. But more important - the Still itself runs quite warm according to the Still thermometer, always above 1.2 K. Which doesn't make physical sense - the actual temperature of the mixture in the still must be lower than this or the fridge wouldn't work! But as I recall in an early run I mounted the portable thermometer on the Still and it agreed with the Still thermometer. So apparently the outer part of the Still is significantly warmer than the mixture inside. This is pretty mysterious though.

tamu_tf_coldwiring.pdf