The MVD cooling system was working in the year-2 run. However, there are several things we could do to make it more reliable and easier to operate.
The existing water cooling system does not have any easy way to fill it with water. One method we have used is to pour water into the the exhaust line. Air in the line tends to prevent the water from flowing down into the system. It is further complicated by the fact that the end of the exhaust line is someplace on top of the main magnet and is hard to reach. Hubert created this page which describes a much better system to fill the water cooling loop -- but this page describes what we want, not what is there. This job could be taken over by anyone with good plumbing skills and would take a day or so.
Think about replacing the air flow monitors. The air flow sensors are on boards which look like the sketch on this web page. They work, but are hard to adjust. They are based on the tmp12 temperature and airflow sensors. The problem with these airflow sensors is that they are actually a combination of a small heater and a temperature sensor. If there is some airflow, it carries away the heat produced by the small heater and the temperature does not rise. If the airflow stops, it heats up. The temperature sensor then trips the system off. The problem is that it is difficult to set these sensors and they are affected not only by changes in the airflow, but by changes in the temperature of the cooling air. I would prefer a simpler and more reliable sensor -- like something which spins when the air is moving and stops when the air stops. This is not an urgent problem, but I would like to see it improved sometime. It would probably take several weeks of someone's time to integrate something into the existing system.
The current temperature and temperature/air flow sensors send logic signals via a sensor distribution board to a relay board which shuts down the MVD Low Voltage when there is a cooling system failure. In each air cooling loop there are 3 sensors. The system shuts down whenever any 2 of 3 sensors say the temperature is too high. However, I have noticed that whenever any one sensor stops working, the system will start shutting down every hour or so. I believe that the sensors are too noisy and occasionally send a brief "temperature bad" signal. Because 2/3 signals must be "bad" to shut down the system, this noise is not usually a problem. However, when one sensor stops working, then this brief "temperature bad" signal shuts down the MVD. It stays off until reset. I would like to find and fix this source of this noise -- probably with some big capacitors -- in order to make the system more stable. This job would probably take a few days to a few weeks and involves only simple circuits.
The RS-232 connection to "chiller 1" -- which is the chiller in the "pocket" on the south east side of the main magnet does not work. We need to either fix this connection or replace this chiller with the spare chiller -- we should probably do both. This connection is important because the chillers sometimes fail to turn on after they are shut down. When this happens, there are two ways to turn them back on -- either push the front panel "on/off" button or issue a reset through the RS-232 port. The first option requires access to a part of the IR which is hard to reach when the beam is on -- so the RS-232 option is much better. The first step is to switch chiller#1 and the spare chiller. This requires mainly a strong back (it weight 40 or 50 pounds and goes into an awkward location) and some plumbing skills (there are water lines going in and out of it).