Resent-Date: Tue, 15 Sep 1998 14:34:45 -0400 (EDT)
Every once in a while, the TD system seems to go into a mode such that when you TDDAC the whole TD system from DUI every channel fails to TDDAC. There is no need to break out into a cold sweat. (I'm the guy in charge of cold sweats...) When this occures I run TDDAC from bnlku7 which seems to clear up the problem. The problem typically happens when the power is cycled on some system. Master crate, fera's etc.
Ok, so now for a quick tutorial on TDDACing from bnlku7.
I have two TDDAC scripts which I use often. One that TDDAC's a crate's worth of TD's and another which will TDDAC one board in a crate. The latter one also allows for further diagnositics. What I do is TDDAC a whole crate which tells me which are the bad channels. I then go and TDDAC the board with the bad channel and use the high level diagnostic routines to see what is wrong.
For example, if I replace TD #3 in crate A01, (the top right TD crate, 4th one from the right) I do the following after I power up the crate.
cd ssp.dir
sspar
exit
cd $TDDAC_PROGRAMS
do_tddach_ch_crate a01
If there is a suspected bad TD, say channel #2 in board #4 in td crate A13, then I do the following after loging into bnlku7 as online.
cd ssp.dir
sspar
exit
cd $TDDAC_PROGRAMS
do_tddac_ch_board A13 4
paw
exec tddac
.
. (hit return a total of 12 times)
.
quit
What this does is TDDAC the 5th board from the right of TD crate A13 which
is the top middle TD crate. This single board TDDAC will save the data
from TDDACing and then I have a paw tddac.kumac file which will read in the
data and display it in a graphical form. You can then see what is going
on with all 4 channels on the TD board. What tddac.kumac does is the
following.
For each channel on the TD board it plots 3 displays. The first page
of display are two graphs ploting the dac value vs adc value. The top
plot is the dac vs adc for one of the Flash ADCs, on the bottom the
same data for the other Flash ADC. You should see a nice streight line
on both of them. For a nice example click up
ssadler.phy.bnl.gov/~adler/e787/tddac1.ps
The next page takes both lines and plots them on ontop of another. This
is a quick visual check of how well the two ADC's match up. (i.e. both
Flash ADC's digitize the same number for the same input voltage.)
See:
ssadler.phy.bnl.gov/~adler/e787/tddac2.ps
Next plot is where I plot the difference between Flash 1 and Flash 2 for
each DAC value. This shows you how well the to flashes are really calibrated.
The critical value here is that the absolute value of the difference
should be less than 2. do_tddac_ch_board and do_tddac_ch_crate will
flag the channel if this abs(difference) is > 1.5. But Marty's version
will not flag it until the abs(diff) is > 2.
See:
ssadler.phy.bnl.gov/~adler/e787/tddac3.ps
You can see all the plots from one execing of tddac.kumac in
ssadler.phy.bnl.gov/~adler/e787/tddac.ps
There are 3 plots for each channel.
Going back to the 'All the TD's are &*#&'ed up!' issue. When tddac
from the DUI window fails, I login into bnlku7 and run TDDAC for
which ever crates seems wonky. (as I showed you above) 99% of the
time this clears up what ever problem the TDDAC/DUI got into. If
this does not seem to fix the problem, then you should try and contact
one of the TD experts on hand to help you out....