On Tue, 23 May 2006, Vivek Dhawan wrote:
>
> Ken is compiling a history of what modules were where/when.
As promised, I provide, entwined in Vivek's narrative, what
I have been able to learn of the HW state on various dates.
It seems to make a case for the glitch we are chasing being
associated with L302s using "new" software but does not hint
(at least to me) at a mechanism.
Ken
> ----------------------------------------------------------------
I was unable to find a record of it but Matt confirms that the
ea18-l302-1 synth was the "new version". I was unable to
determine whether ea13-l302-2 was old or new software and
assume it to be old.
> APR 27: dropouts on 18 AC, NOT on 13 AC or BD.
May 3: synthesizer with "new" software installed as ea13-l302-1
to test the fix for the "sawtooth" problem.
May 4: synthesizer with "old" software installed as ea18-l302-2.
> MAY 04: dropouts seen on 13 AC AND 18AC
>
> The magnitude of drop on 13 to 18 is the same
> as the drop on 13 or 18 - not doubled or zeroed.
>
> Dropouts can miss a beat or 2 - not every 10s
> has one. The pattern is identical on 13 and 18.
May 9: Rob reports that ea13-l302-1 sometimes shows out-of-lock
and proposes replacing it. I concur, thinking it may
fix the 13AC glitches seen by Rick and me.
May 10: ea18-l302-1 ===> ea13-l302-1 "new" software
ea18-l302-2 ===> ea18-l302-1 "old" software
> MAY 12: 13AC (~50% drop) AND 13BD (~10% drop)
>
> Not visible on 18
>
> Not always there on every 10s tick.
(Walter reminds us that this is because of
undersampling)
> All IFS drop when 1 drops.
>
> (Walter reported on this same data earlier)
May 19: Swap the L302s in antenna 13 so that:
ea13-l302-1 is "old"
ea13-l302-2 is "new"
> May 19: moved from 13AC to 13BD with L302 swap.
>
> not visible on 18.
>
> ----------------------------------------------------------------