Professor Jim Connell
I met with Dr. Jim Connell on November 2 to discuss the use of silicon detectors in the work he does.
Dr. Connell works in Morse Hall in the Space Science Center using silicon detectors to study particles in space. I began by asking him about simulations they run and he said that they do not require very high precision in their instruments and do not simulate things with as much detail as is included in the NIM 544 paper (included in Relevent Papers section). The reason that they do not need as high precision is because they are concerned with just the type of particle that is incident on the detector rather than the position at which it hits. However, being able to tell its position can be very useful in some cases to refine measurements when particles are not incident at normal angles.
The main detector setup is one that works on a circular surface with a single set of 222 strips oriented parallel across the disk of silicon. Each of these strips is attached at one end to a single readout that reads out the energy of the entire array (E) and the other is attached to a series of resistrs and an amplifier that reads out the associated energy on a certain strip (P). To a first order approximation, this setup allows for the position to be determined as x~P/E. On the Ulysses project there are two stacks of three detectors oriented at 60 degrees to each other. The redundancy of having three per stack is to ensure that at least two make it to space. If all three survive, then the third acts as a check for the first two to give a measure of the precision. Having two stacks allows for two (x,y) measurements and thus the path of the particle which is approximated at a straight line. In addition to the redundancy, another big difference in his work is that he must take into consideration the power output of his detector and usually a detector runs on ~4W of power where as our detectors have nearly unlimited power sources.
The next issue we adressed was that of the the thickness of the detectors. I asked him why 300microns is a usual choice of thickness. This is because of the balance that must be made between signal size (SS) and capacitative noise (CN) while taking into consideration that a thicker detector is more detirmental to a beam and takes more energy out of it. The main focus is to look for a good SS:CN ratio. The relations are SS~thickness and CN~1/thickness. Thus, with a smaller detector, the capcitative noise goes up and the signal goes down. To fix this, the detectors are made larger. However, there is a limit at which the beam is impeded too much and hurts measurements too much. This problem can be remedied by dividing the anode into strips. This reduces the effective capcitance and thus CN for each strip while keeping the SS the same for each strip as the amount of charge deposited is constant. Doing this then increases the SS:CN ratio and also allows for coordinates to be extracted more readily. With experience, the best balance for many cases has been found to be ~300microns.
He and I then discussed the advantages of having these detectors in his work. On the Ulysses project, these are useful in that they allow for the incident path of the particle on the energy detectors to be determined. This is important as the energy deposited in a layer is proportional to the path length and increases as you go away from normal. Knowing the incident angle well allows for corrections to be made and so you can use the detector to tell the different between elements better.
This was a very effective talk and he invited me to look on when they tested some of their detectors. This is something I would like to do and I hope to meet with him again.