A couple of quick comments:
> > Finally, Chris convinced me to throw in Ioannis' suggestion for a
> > fix: Ioannis suggested to get rid of the zero-suppresion and use all
> > memory (currently split between ADC and TDC words) for ADC
> > words. This will double the avaiable time window for digitization,
> > in which case a common stop at ~160usec(?) will cover most of our
> > acceptance for slow particles. Ed, would something like that be
> > doable?
>
> Sorry- the WFD is not a computer. Half the memory is reserved for
> timeword/discriminator and half the memory is reserved for FADC data.
> Even with zero-suppression off, the timeword/discriminator is written.
> This is hardwired.
>
Erik may have mispoke here a little. Half of the memory on the board
IS reserved for time word/discriminator and can't be changed. But if
I am not mistaken, the READOUT of time and adc is performed in TWO
separate commands. Erik's point was that if we ever decided to turn
zero suppression off, we could forgo reading the time words (since the
adcs would already be at a known fixed time separation of 20ns per
four samples). In this way we could just read out all of the ADCs for
an event and get about 160usec of readout (32kB = 32 ksamples =
32k*5ns = about 160usec)
> > modified with the 1mF cap. This involves reduced pulse amplitude
> > and 2mv p-p oscillations; none of them were observed before. We
> > believe both have to do with the soldering of the cap and not
> > with a systematic effect. As a cross check we plan to check ASAP
>
> This is scary and important. It was hard enough getting the ground
> noise down with the ferrite donuts. If uncabling, removing the
> fanouts, modifying them and plugging them in is going change the noise
> environment, then we have to know ASAP and see if we can correct it.
>
We agree! But if there is an actual noise problem with electrolytic
caps, it is likely to depend on properties of the particular
electrolytic cap we choose. For the 4 electrolytics we bought at
Radio Shack and installed in 3b01,02,03,04-0 we have NOT seen such 2mV
p-p noise. So having sampled two kinds of electrolytics, we have one
kind that doesn't seem to give us a problem, and another kind that
might. We of course need to make sure that the one we order is of the
kind that definitely doesn't give us a problem.
> > dynode stages which can not supply the necessary charge in order
> > to meet the peak anode current. The PMT output at that level is
> > consistent with loss of gain (effectively you lose the last few
> > dynodes stages). They've doubled the capacitor value in the last
> > dynode chain and they thus managed to see the maximum charge
> > delivered by the PMT almost doubled.
>
> Am I to infer you are considering a base change?
>
No! Just for fun (?) we estimated the man-hours required to do this.
We guessed about 1.5 months of round the clock oil pumping. This
effect is relevant since it limits the maximum charge we have to
redesign the WFDs for. The base change Erik described above was
performed to verify that we did indeed understand the charge limiting
effect of our bases.
Chris