[evlatests] Re: EVLA 700 KHz spurs, Dec 22 2005

[Vivek Dhawan]

Re: the 700KHz birdies (reputedly from the switching power supply)

As Ken says, it is certainly worst in 14A, about 100*(front end noise)
in one 3KHz channel. Is this easy to see on a scope at the backend?

It does NOT appear to correlate across other antennas, though 16 has
weak signs of it in autocorrelation- the vector average on baseline
14-16 over 140 sec of data shows much less than the 10s records.

 From the data available, it requires suppresion by a factor ~1000 to
be within spec on 14A for autocorrelation data (i.e., 0.1*FE noise)
and a factor ~50 suppression on 16A.

V.



Ken Sowinski wrote:

>                          December 22, 2005

> Birdies

> I thought it was time to see if the strong birdies that were 
> seen in the past are really gone.  The remedy tried a while ago was
> to crank up the deformatter output to drown the spur and then reduce
> the signal with pads to the T4 input spec.  I don't really remember what 
> pads are used on what signals, but perhaps someone can fill this in.
> The only spurs seen were at around ~700 kc and look like what we have
> in the past attributed to switching regulators on the deformatter cards.
> The presence of birdies was checked only with 10 second integration.  
> It is likely they would be seen on all IFs with enough integration.  
> In 10 second integrations nothing was seen in 14B, 14D and 16B.  If
> anyone really wants to look at the data I can reconstruct the times.
> 
> 14A
> This one was very strong and could be seen in autocorrelation even
> in 25 MHz total BW (about 800 kc/channel).  In 3 MHz total BW it
> could be seen in correlation with 16 and in autocorrelation in 16 as
> well.  At 1.5 MHz total BW it could be seen on a baseline to a VLA
> antenna: 9 x 14.
> 
> 14C, 16D
> These could be seen only when the BW was 6.25 MHz or narrower
> 
> 16C
> This one was visible only in 3 MHz total BW.
> _______________________________________________