Status of noteworthy WASP2 correlator boards

Numbers ~2750700 or ~700 are from first run for WASP2 and SOFIA.
Numbers ~6000 are from second run for CSO.

***************************************
Potential problem boards
***************************************

2750777: started working properly during repair, failure unknown.  See
below.  

***************************************
Repaired or spare boards
***************************************

2750776: dead channel at CSO, spect. 3 (18.8.03).  Swapped out, but
checks out ok in lab (21.11.03).  Now fine, in spares.

2750777: ADC#12 has no RF response, high noise.  Seems to be a demod
problem since opamp output switches +/- 3 V.  Replacing switch, opamp,
reflowing solder on synch det resistors has no effect.  Problem
disappears on probing near opamp chain inputs.  Reflow solder on input
components.  Problem goes away, but this is rather mysterious and
uncomfortable.  

2750780: ADC#7 saturated at all 1s.  Analog circuitry fine in test
setup.  Replace ADC 30.12.03, now fine.  In spares.

2750785: noisy channel at CSO, spect. 2 (24.8.03).  Replaced
multiplier on ADC#9 after verifying noise in test setup (21.11.03).
Now fine, in spares.


***************************************
Problem boards
***************************************

6061: in spect. 5, first lag on this card has "sparkles:" +/- 32k
values, at certain repeatable frequencies.  Currently column 81 in
saoimage.  Replaced multiplier but no change.  No problem by swapping
uC#5 for uC#? (spare) -- so it seems to be a micro problem rather than
an RF problem.  Perhaps marginal voltage level for ADC, but then how
does it know about frequency in cwsigs?  See notebook 22.12.03,
11.12.03.
  It does seem that this board has problems -- tests 7.3.04 show the
"sparkles" in a cwsigs, with otherwise sensible operation.  In
spectrometer #5, slot 2/7.  Changing Phytek micro and entire micro
board had no effect.  On 22.12.03 the jumpy channel moved with board
swap.  Changing the multiplier 22.12.03 had no effect.  Why it's
erratic is unclear, but it's certainly a problem channel.
  The ADC has coding problems -- works in low bits, but jumps suddenly
(to wrong polarity) at specific votages.  Removed board from
spectrometer, cooling equipment from board.  See V-81, sigsimage2.fits,
10.3.04. 

2750764: Dead half-board at CSO (25.8.03) in spect. 3.  Suspect ADC
dead or bad shift register.  Replace ADC#8 30.12.03.  Tests ok, back
into spect. #3.  
  Dead ADC, failed during lab tests 8.3.04 (zeros at and beyond ADC
88; minicom.cap), also seems to draw a lot of current.  Removed
cooling hardware.  See notebook 8.3.04.


***************************************
Boards in use, out of spare or repair category
***************************************

2750789: Unused spare, checks out fine 21.11.03.  Added cooling
hardware, tested 11.3.04.  To be installed in spect. #3.

2750779: Unused spare, checks out fine 21.11.03.  Added cooling
hardware, tested 12.3.04.  To be installed in spect. #3.