MCPs 80mm-II

The roentdek documentation on the MCPs (see p.7) says that the nominal working voltage in chevron configuration is around 2400 V through the plates. So far, we where around 1850 V, thus it is normal to have low gain.

When I tried to increase step by step (100 V per 100 V) the voltage up to 2200 V through the plate (the ion gage is always off when I ramp up) , the dark count rate started to increase up to 1kHz or more… I quickly went down to 1800 V, but the current was twice higher. This meant that there was a drop in resistance down to 20 MOhm. It may be due to current leak through the plates ?.... Anyway I let it on for a while at 1800V until everything was normal...I ramp up again, and point out a limit (see arrow in the picture) where the problem started and I took a picture of the pulse shape. I continued to work for a while close to this limit at least to check the signal on the anode.

There was a terrible noise when the acquisition system started. The acquisition system was frozen around 1kHz or less. After one or two hours to understand where the noise was coming from, it turned out that it was necessary to avoid the NIM-bin as much as possible and to feed directly the MCP signal into the DS16. Hence, we directly got in ECL format the MCP signal and we used the level translator to have a trigger. Doing so, it seemed to avoid ground loup or antena effetcts and we could get the acquisition system working up to 25 kHz. A short study of the image with the gage on, as a function of the MCP voltage is done. In the table we give the number of events as a function of the MCPs voltage. At 2300 V the gain is enough for the MCP signal, but on the anode we still get a better hit distribution 3 pixels at 2400 V (around 2200 V through the plates).

U (V apply)	2200	2250	2300	2350	2375	2400
Count rate (kHz)	8	15	23	24	25	26

One should try to user a little higher voltage but there was a software bug… memory allocation problem due to some FIFO full??? In any case the distribution of the hits on the DS16 looked the same than yesterday. The signal on the cards A, B, E, F, G, H was about the same intensity which is much better but the distribution was still very chaotic… In the rush before leaving, I checked with a pulse generator whether some DS16 channels were dead. Hopefully everything is fine, but I should double check !! Is it a real physical signal ?.... It could be because A & B looked mirror image as well as E-F & G-H

posted by Mathieu Gisselbrecht at Aug 2, 2006 2:00 AM