> I got the memo today... is that the fully updated one?
As up-to-date as we have so far.
> I am really torn on the right path for a fix. The solution of
> simply removing the one capacitor is really attractive. However, I am
> rather dismayed by the consequences. I keep thinking of getting up and
> giving a talk on our 1 discovered monopole and having the audience
> grumbling about why we didn't have the ability to see the full
> waveform. Also, the issues of sensitivity to smaller pulses and the
I would hate to have that happen too. Even if we go with the "small RC
fix" I think our goal is to see the whole waveform for 50 cm
pathlength (which is 93% of all acceptance for an isotropic flux and
probably even greater once one imposes two-boxes hit etc). There will
always be some acceptance where we would lose the middle (in the
small-RC-fix) and the point about catching the tail end is to say that
even for those, we can say something.
I'm hoping that for at least 2x dyon levels, PSPICE can prove that the
photostatistic fluctuations will cause us to catch all the waveform at
low beta low dL/dX. We can also try to get an Arbitrary Function
Generator and prove it in hardware.
> biggest pulses continues to worry me. On the other hand, I do agree
> with Ed's comments about not losing our main sensitivity by trying to
> chase after too many geese...
>
> OK, so let me (re)inject my own suggested solution. Why can't we
> add a zener to clobber the positive overshoot going to the
We could (a signal diode would be fine- we just need the .4-.6V diode
drop). It would help a bit, but wouldn't do the whole job (I don't
think). The typical "bad" overshoot isn't often much more than
1V.
> discriminator as well as the negative signal in combination with the
> rest of the earlier solution? I don't have a daughter card in front of
> me and I don't know the layout well enough to understand how hard that
> could be.
We'll look.
> Finally, some hybrid approach may be the best approach. Because the
> solution of simply removing the one capacitor is *so* simple, it might
> be worth doing that for short term and then get a better solution. I
One other thing to consider. Looking at the latest plots, it may be
that some slightly larger value like .02uF might help close the
missing data gap. Taking off the 1uF and putting on a 2nd .01 in its
place isn't very much harder.
Ed