Notes on PACT - John Schaapman - May 31th, 2001 

	
  The best way to describe it from a control point of view is that it is a
modified PAD module. The monitoring, test pulse amplitude setting and test
pulse trigger and enable are all done through an unmodified PAD control
module. In fact, PACT and PAD modules can coexist in the same crate. The
only thing that is different is that the PACT has no thresholds to set
just the test pulse amplitude. The threshold readback line is conveniently
used to monitor the 24 v supply instead. The only thing we didn't consider
is that there is no direct way of identifying whether a module is a PAD or
PACT. Perhaps a quick test to see whether a module responds to threshold
setting changes would be good enough.
  Each PACT module has three front panel Lemo connectors which accept NIM
level signals for Fast Clear, GATE 1 and GATE 2 respectively from top to
bottom. Fast Clear can be left disconnected to enable most normal
operation at the rates we are likely to use. There is an internal jumper
on the PACT module which is normally set so that the GATE 1 input becomes
a common gate for the whole module to save on fanout requirements. They
will only be used seperately for the scintillator readout to allow a
different gate for upstream and downstream.
  The PACT requires a gate signal in order to read anything including the
test pulse. On top of that, the gate has to be closed before the test
pulse ends. So the easyest way to arrange this is to feed the gate signal
to the test pulse trigger input on the control module as well ( or the
test pulse to the gate, depending on your point of view ). The delay
through the control module, crate backplane and PACT circuitry ensures
that the gate will be open before the test pulse arrives and closed before
the test pulse falls.
  The gate signal should be in the range of 100 to 200 nsec. I've been
using a 170 nsec gate in my setup in order to fully capture my VTX test
pulse. The rate I've been using is 20 KHz. because my preamp individual
input test pulse has a long tail. You could easily double that rate when
using PACT test pulses.
  When there is no input signal or test pulse enabled the PACT will output
a pedestal signal for each gate. It will be about 300 nsec with 100 nsec
gate, 200 nsec with 170 nsec gate and 50 nsec with 200 nsec gate. There is
less variation from channel to channel with the shorter gates. Beyond 200
nsec some channels will stop giving pedestal output and at some longer
gate none will give output.
  With a 170 nsec gate and test pulse width and no preamp cables
connected a test pulse amplitude setting of about 1,600 mv will give a
Full Scale output of about 2,200 nsec ( subtracting the pedestal of about
200 nsec gives signal of 2,000 nsec - 4000 counts in TDC ). With the
preamp cables connected a setting of 2,040 mv will give a 3/4 Full Scale
output. Ideally, the test pulse output should not vary with gate width but
it does in this case because the gates are too short to allow the test
pulse to decay before the gate closes ( a compromise in order to get
decent signal levels ). So, I guess you have to keep track of what the
gate width was when taking testpulse data. ( also when long vs short
preamp cables are used since a reflection will arrive before the gate
closes on a short cable and after the gate closes on the long one. )
   As far as regular signal input goes, the gate should arrive at the
module at least 8 nsec before the signal. A maximum amplitude preamp pulse
through a short 9 m cable gives near full scale output. The longer delay
cable will reduce this slightly.
   For overscale inputs ( such as test pulse amplitude setting of 2,040 mv
with no preamp cables connected ) the output will keep getting longer (
about 3,000 nsec in this example ). It just isn't spec'd beyond the 2,000
nsec Full Scale.
   The PACT module will be used with the MQT300A set for the MID range
which is done with shorting jumpers on the first and third pair of pins
on the front panel 8 pin connector.

 
John Schaapman                     Ph: 780-492-3043
Centre for Subatomic Research     Fax: 780-492-3408
University of Alberta            NOTE: new area code
Edmonton, AB
CANADA       T6G 2N5