From: Vincent Smith [Vincent.Smith@bristol.ac.uk]
Sent: 18 September 2009 10:48
To: Dave Cockerill
Cc: Ken Bell; Alexander Iskrenov Hristov; Wolfgang Funk
Subject: Excess current in -NQL

Attachments: EEHV(1).xls


The attached spreadsheet shows the variation of current with voltage for anode and dynode of -NQL, 
taken yesterday afternoon.

The columns are: anode volts, dynode volts, anode current, dynode current (all these from the lemo 
sockets on the Torbet input card), then the CAEN anode current, the CAEN dynode current, and finally 
the ratio of Torbet volts to Torbet currents (expressed in Megohms.)

NB The Torbet values are 'as read' in mV.
For voltage readings: 1mV corresponds to 1V.
For current readings: 1mV corresponds to 100nA (or 10mV = 1uA).
For the calculation of effective resistance, I have subtracted the current offsets.

Readings were taken up to 800V (anode and dynode together), then the values as used in real life, A=800V, D=600V.

Finally, since Ken reports that the excess current appeared after the resistors were changed on the input
 cards, which sounds very suspicious, I tried using a different input card. The excess current is still there.
(Actually, I think we tried exchanging input cards at the time we first noticed the excess current, didn't we Sasho?)

So the final column shows a small change of effective resistance  of the dynode with voltage, but this probably
reflects the lower accuracy of the readings at lower voltages. In any case, the trend is opposite to the anode.
It certainly looks like a short circuit in series with the 22M + 1.5M in the dynode supply line to each VPT.

Vince.

PS I have put the CAEN firmware release numbers, etc on the new EE HV Wiki page. Now we have to find what
version the Barrel CAENs use...