Re: MuCap electronics cabinet

[Tom Banks <tbanks@socrates.Berkeley.EDU>]

> You have to be careful how you look at the ECL signals, so that your
> observation does not introduce the distortion.

I began to suspect as much, when my replacement NIM/ECL converter module
output had the same distortions as before.  Aart Dijksmann was kind
enough to lend me his ECL scope probe, and I am using that now.

Using this new scope probe, it became clear that the NIM/ECL converter
was actually OK.  I therefore went ahead and ran several clock tests at
varying duty cycles: 25 MHz at 25,50,and 75% duty cycle, and then 50 MHz
at 25,50, and 75% duty cycle.  Claude ran his analysis program on these,
and for the 25 MHz signal, the 50% duty cycle was significantly better
than the others, although there were still some residual time structures.
The 50 MHz 50% duty cycle was actually inferior to the 25%, which supports
the idea that 50 MHz is just too fast.

What concerns me now is how we can guarantee the perfect operation of the
CAENs.  It seems we need to empirically "tune" the clock's duty cycle
until the artifical time structure disappears in the CAEN data, which is a
rather distasteful arrangement.  It's also possible that the 3 CAENs would
require slightly different tunings--which is nearly impossible to
achieve--in order to satisfy their individual circuit setups.  I should
mention that the ECL signal is still not a nice, sharp, square signal.  I
will try running it through another discriminator and see if improving the
signal shape improves the CAEN performance.

> Once you made some progress, we could talk on the phone one of these
> days to discuss the technicalities. and possible solutions.

I'll stop badgering you guys with e-mail updates, since I usually have to
retract my previous assertions anyways.  But I wanted to let you know that
my previous feelings about the converter have changed.  I'll try to
summarize my findings sometime next week, and perhaps I should talk with
you (Peter) then.

Tom