--------------- REGULAR WFD SYSTEM UPDATES? ------------------------------
I will try to send out updates about what is going on a little more
often now. I'll concentrate on getting word out about what is happening
on my end, but I will also try to include mention of other work that I
am aware of. If you would like to send me something about what you are
working on, try to get it to me by the weekend, since these reports will
go out on Saturday or Sunday. If you have a long summary to report, you
can always post to this list yourself.
These updates may be a little stream-of-consciousness at times; I may
have to trade off tracking down details in favor of getting the news out
quickly. I intend to specifically mention items that require further
planning so that you can give me feedback. For example in this issue I
point out that we are reopening the question of how much to fan-in the
signals. My point is that just because it is mentioned here, doesn't
mean that it is written in stone. Specific and non-specific questions
can always be referred to me via email.
--------------- CATCH UP ON WFD NEWS -------------------------------------
I'll start by catching up from around the time of the collaboration
meeting. At that time, we had identified the "rollover" problem at BU,
and I described proposed solutions at the meeting. After the meeting, we
were busy accumulating the handful of revisions to the motherboard and
getting them to the layout designer. We also were improving the design
of the analog daughtercard: we settled on adding an amplification stage
to the input to the comparators (discriminator) to solve the problem of
reliably discriminating in the 1-6 mV range. We built a prototype of the
STOP/START/CLOCK fanout board and tested it briefly at Gran Sasso.
We were also in the midst of the first round of WFD usage at Gran Sasso.
Beginning in mid- September, we had accomplished WFD readout via the
VaXELN based mini-DAQ. Hong, Walter, Lu and Sanzgiri worked out the
fundamentals of reading out via the existing CAMAC list directives. Lu
and Sanzgiri added control functions that may be linked into COMMAND,
including waveform display capability. They used this to make good
progress in solving the power supply switching noise problem that occurs
when the WFD is connected to the PMT fanouts.
By January, Ash was finishing up his noise studies and preparing to
return home, and I was starting to get into the logistics of parts
ordering and sending out for a pre-production prototype of the
motherboard (as well as the latest version of the daughtercard).
This was underway by mid-January, and I took that opportunity to go
to Gran Sasso and see about implementing "time-window" readout
commands that could be incorporated in the CAMAC lists. For more
about this, you can see the recent MACRO memo.
--------------- RECENT STUFF ---------------------------------------------
We built the pre-production prototype, and it works without modification.
Unfortunately, we discovered that the proposed rollover fix (using
NIM_CYCLE to switch zero-suppression off temporarily) does not work as
well as it needs to. It seems this signal cannot always be used as a
fast control signal, because it is not latched to the 200 MHz clock.
This is similar to the rollover problem, but different enough that we
did not pick up on it when designing or testing it.
We are still trying to gather all the information we need to evaluate
this. But the bottom-line for standard MACRO running is that it is not
an obstacle. At low thresholds, we will have enough data to always
occupy the 327 us rollover period. The time-readout algorithms have
already been rewritten to count time correctly without actually finding
a rollover cycle (and it seems without much penalty in algorithm speed).
The new version of the daughtercard has a few bugs that need to get
worked out. Also, I hope to shortly try some tweaks to the design to
catch more post-samples.
We started ordering parts for 110 boards. The parts will be ordered by
BU and TAMU- Ann David was a big help in getting some orders placed
faster than the BU bureacracy could permit. The longest leadtime part is
due in mid-May (F100325 ECL-TTL converter). It may be useful to assemble
the boards without this part because most of the VME functions can be
tested without it- we'll have to see how much time we actually can save.
--------------- GRAN SASSO TESTS IN MARCH --------------------------------
I went to GS in March, in principal for an Offline meeting, but also
interested in trying a few more things with the wfd. Chris Walter was
already there with a prototype LIP trigger, and Ash Sanzgiri arrived
ready to help with more tests. Together we assembled a working system
using the LIP trigger, two WFD cards and the new CAMAC list readout
routines. We took an overnight run of muons (where the LIP trigger
required a coincidence between streamer tubes and scintillator) as well
as higher-rate runs of just scintillator coincidence. The data is on
disk in the proposed equipment format- ask Chris for details.
One small test: we put in some optimizations to the time-readout loop
of WFD_RDTIM and measured a very small increase in readout speed.
(Where *are* those numbers...???)
Pisa repaid an equipment debt by buying us our very own VIC-8250.
Ash figured out the switch and jumper settings and that is what is
being used for readout now (replacing the VBR8212 with the VMDIS
acting as arbiter/system clock).
--------------- RANDOM MUSINGS --------------------------------------------
The budget is still tight; I haven't redone it yet using the latest
numbers. But it may be necessary to build 80 cards instead of 110.
This would change the fanin of horizontal tanks to 4:1 instead of 2:1.
We would save on parts (not as much as if we had decided this long ago)
and also save money and headaches on trying to handle the problem that
our +5 V power supplies are not beefy enough.
We should talk seriously about our readout options. There is talk of
having a reprise of the readout meeting that we had after the CIT
collaboration meeting. Doug is especially interested in this, and is
in the US for two weeks now, so maybe we can get some video conference
going on short notice. Stay tuned for more information.
In January, I noticed that one of the two WFD boards was not working
(would not "stop", or for the literally minded: I couldn't make green
light turn red). After my March trip, I brought back this board (# 004).
Three channels were not functioning, and we spent some of this week
looking at it and eventually replacing an ASIC (which seems to have
fixed that channel). I am concerned that what was once a working board
now has 3 of 4 ASICs go bad. This has not happened on the other board at
GS nor on the other 3 boards here and at TAMU. If anybody has anything
in their notes about when this board ceased functioning, please let me
know. I am very worried that there is some failure mechanism that blew
this card out- and we must see that it never blows out an entire crate
of cards!
----------------------------------------------------------------------------