Run 34

Run 34 is focused on testing detector G106a - a complete 100 mm iZIP crystal from Umicore (the full run setup is described in Table 1). This run is focused on charge studies, and our next run will be full detector testing with the phonon channels. The reason for this is that we wish to replace SQUET card ZC-28 for a new card because we suspect it is the reason for the the large parasitic resistance on channel C, Dib 2. Hopefully, then, all phonon channels will be responding appropriately and allow us a completely thorough test of G106a.

The tower checkout and initial electrical checkout at the breakout box can be found here. The only significant concern was that LED 2 on Dib 3 fully conducts. However, we can use all the LEDs on Dibs 1 and 2 and that seems to be sufficient for detector flashing purposes.

Our base temperature has been ~ 51 mK. We use 350 mbar of He3-H4 mixture in our circulation, leaving 370 mbar of mixture in the dump.

In attempting to tune channels, take TC measurements, and collect data, there were several problems encountered related to the DCRC boards and the switch which have caused significant delays within the run.

Each DCRC board has its own quirks that must be managed.

There were several problems that suggested our switch was faulty. The DCRC boards would randomly lose power, though maintain a connection. Often we saw this after we stopped flashing the detector, but whether there truly was a connection between these two events are unsure. Also, when flashing the detector the temperature reached by the mixing chamber was inconsistent. This suggested that the switch was not providing sufficient power continuously. This was also supported by the individual power output readings for the ports- the port for DCRC33 was receiving less power than that of DCRC32's. When the board randomly powered off, attempts to re-establish it were also troubled by inability to turn on the FET heater.

The above problems were sometimes reduced by completely turning off the switch (unplugging it). This allowed us to finally get to the point of taking data (i.e. take Tc, IbIs, and charge breakdown measurements), but more problems were evident when we analyzed our data and found extremely poor charge collection. The collection was much too poor to accept that this is what our detector was capable of, and repeated data taking at the same voltage bias produced significantly different results.

We decided to remove the switch's capability to power the DCRC boards and only let it act as a connection to the DAQ computer. We also connected two external PoE supplies to the switch to act as power sources for the board. This has proven to be a dramatic improvement: we've not lost power with the boards yet, flashing produces a consistent temperature of 180 mK, and we have collected very good charge data, which will be described below. We probably will buy a new switch to replace our current set up, but we are now able to take data and that is a huge relief.

Charge breakdown studies were conducted by looking for breakdown across the crystal as well as on individual sides between the Qi and Qo electrodes. These tests were done with both polarities. The results can be found below. To summarize: this crystal shows little evidence of charge breakdown. Few hints were detected around 9 and 11 V between charge electrodes on one side, and it only became very obvious at 12 V.

Charge-only data has been collected using DCRCs 33 and 32 with Dibs 1 and 2, respectively. A complete inventory of data taken thus far can be found here. To briefly summarize the data taken thus far, see Table 4. Examples of the charge spectrum at different biases are shown in Figs. 1-9. The biases are displayed for +/- 2 V, 4 V, and 8 V with the Ba-133 source and without. Biases of the reverse polarity with the Ba-133 source are also displayed.

Work in progress

IbIs data was taken for all phonon channels. Channel C2 again has a very large parasitic resistance, which is similar behavior to what we saw when testing G103a. Figure 10 shows a comparison of C2's IbIs curve to C1's curve, C1 being a well-behaving channel. The SQUET card (ZC-28) is suspect, so we are going to do another run with a new SQUET card so that all channels will be fully functional for phonon studies.

Figure 10: C2 IbIs curve in comparison to C1

Remaining figures of the IbIs curves are located here. Table 5 shows the values of the normal and superconducting resistances for the phonon channels. The question marks are estimations due to a not ideal fitting algorithm. The deviation in the biases use to take data from the calculated QET bias is due to the observation of the pulses under the calculated QET bias. In effort to improve the pulse shapes, some of the QET biases were slightly modified. Dib 3 is much noiser than Dibs 1/2, and almost all of its QET biases were modified.

Updated phonon settings as of 2013/02/16

All Tc's were measured using 10 mVpp triangle wave at 100 Hz except for Ch. C2, which used 100 mVpp triangle wave at 1000 Hz. The temperature of the mixing chamber was slowly raised using the mixing chamber heater in order to record when transition occurred and observe when the signal went normal. Then the mixing chamber was allowed to cool and Tc measurements were again taken. Ch. C2 was was slightly more difficult to record because it required a different triangle wave to be used, and we were only able to record the Tc as the temperature of the mixing chamber increased. Ch. B3 was also somewhat difficult to get the Tc's for when initially taking measurements on Dib3. It's signal was smaller and noisier and we couldn't increase the output gain to enlarge the signal. Increasing the output gain would cause the baseline to go out of range and the driver offset adjustment was not sufficient to bring it back within range. Furthermore, when trying to adjust the gain and offset on Dib 2, the LabView program would freeze again.