Hello skelec,

This note is to update you on the status of the front-end QTC modules,
particularly with emphasis on my recent visit to Super-K and the
upcoming December DAQ tests. You should soon have a copy of the
collaboration meeting talks which contains some plots and more
information- I'll also try to get some of it onto the WWW.

The quick summary is: the daughtercard design is finished and is going
into production; the mothercard has a handful of simple changes
requiring a new revision; and we should have our first look at the
front-panel card in a week or two. Estimated delivery of production
modules is in February 1996.

For the near term, we have two modules of 16 channels each. One is at
Super-K right now; the other is being fixed up to 16-channels at
B.U. (Jeff Wilkes has had this at Washington with two channels to play
with). These modules have the following errata:

1) The gate is fixed at 200 ns (will be adjustable in final version).

2) There is a bug in the mapping of input signals to output signals that
will be fixed in the final version. In the prototypes you will find each
group of 4 signals is in reverse order. So assuming the inputs are
numbered sequentially, the outputs are:

  4-3-2-1-8-7-6-5-12-11-10-9-16-15-14-13

3) The PMT threshold value read at the back is 2x the actual threshold,
not 10x as desired.

4) I have cobbled together a pcb+connector to use until the front-panel
card is available. The one at Super-K right now only handles 12 signals
(one ribbon-coax worth). Due to mechanical constraints, these 12 go into
channels 3-14.

5) The module at Super-K currently has 50 Ohm termination. (This is good
for pulser generators.) The ribbon-coax is 100 Ohm, however the mismatch
did not seem to cause a noticeable effect.

Other than these caveats, the modules for the December tests are pretty
much plug-and-play. Just turn the crate on and get a negative going
pulse into one or more inputs. For the DAQ test, the most convenient
mode of operation is to put a NIM pulse into the PEDESTAL input on the
front panel. This will initiate a gate and QTC measurement on all
channels (presumably pedestal). This is a quick way to get TDC edges
into the data stream. I also left with Mika a LEMO gadget to connect a
pulser to a single input. And when it is dark, you can use PMT signals
with the ribbon-coax.

Here are some notes on the ribbon-coax that you may find handy. The
pin assignment on the paddle card, as viewed from the front is drawn
below. Note the alternating signal pins surrounded on 3 sides by ground.
The input pin assignment to the QTC front-end is *reversed*. This is
deliberate, as it is the only way to connect the ribbon cable so that
channel 1 remains channel 1. You may need to stare at the boards and
a ribbon cable to convince yourself of this. Anyway, the proper way to
connect one of the coax ribbon cables is with the PCB on the RIGHT side
at the paddle card and with the PCB on the LEFT side at the QTC modules.

 PADDLE CARD        QTC MODULE             X = GROUND
  OUTPUT              INPUT            1..12 = INPUT
    X X                X X     
    1 X                X 1     
    X X                X X     
    2 X                X 2     
    X X                X X     
    3 X                X 3     
    ...                ...     
    X X                X X     
   11 X                X 11    
    X X                X X     
   12 X                X 12    
    X X                X X     

We tested the board several times at Super-K to verify that it was
working properly. 1) We generated a NIM pulse and put that into the PED
input to see that all output channels were working. 2) We put a PMT-like
pulser signal into each individual input and used a scope to observe
that the delta-t of the output pulse increased with Q. 3) With the
pulser signal at -3 V we observed that there were no crosstalk induced
pulses on other channels. 4) We took advantage of dark tank conditions
to put PMT pulses in and observe the output. 5) We also observed that
the hitsum was clean and made sense during these tests.

A note on the PMT observations: typically, the width of the output
jitters by a few 10s of nanoseconds reflecting the random height of the
dark rate pulses. Visually, it resembles a PMT pulse turned on its side.
It seemed to me that the jitter of the output pulses was much less than
I observed with the B.U. dark box. Basically, the most probable signal
was too steady near the pedestal level. (Sorry if this isn't clear- it
is easier to notice than to describe. Ask Mika or Wojtiech for their
impression as they were also there). This was noted with a PMT threshold
of 25 mV. After raising it to above 100 mV I believe I saw the expected
delta-t variation. One possibility is that some noise was triggering the
QTC and causing a gate, but the noise is bipolar giving a charge
measurement of zero. At any rate, I have no reason to believe the QTC is
doing anything unexpected and I am just recording these notes so that
someone can follow up in future observations.

A final note to the December DAQ testers. You have probably been told
that you should bring everything you need and not count on getting parts
or equipment there. I can vouch for that. There are only a couple DVM's
and they may be needed by other workers. There is one U.S. analog scope
and it is in lousy condition. Fortunately, we were able to borrow a nice
Tektronix digital scope from the Japanese but you can't count on it. We
also had to borrow: a pulse generator, LEMO cables, a scope probe and
some handy variable delay modules. For the December tests you will
probably be concentrating on digital data and will not need much test
equipment. But be warned that you should try to think of *everything*
beforehand. Bring rack mounting bolts. Bring your own Radio Shack
pocket DVM, pot turning screwdriver, penlight etc.

Best regards,
Ed