[evlatests] Birdie at 30.976 GHz

[Wes Grammer]

Thanks Ken. I did confirm my earlier test results: if there's LO 
leak-through at 15.488 GHz, which seems likely, it's considerably weaker 
than at 14.720 GHz, by >20 dB. When I get a moment next week, I'll 
insert a Ku-band LNA ahead of the spectrum analyzer, so I can get an 
extra 20 dB or so of dynamic range, and maybe see it.

Regardless, I don't think there's much that can be done about it, 
without redesigning the downconverter; there's no easy fix I can see.

-Wes


On 2/15/2018 11:00 AM, Ken Sowinski wrote:
> On Thu, 15 Feb 2018, Wes Grammer wrote:
>
>> Jim, it does seem possible, since this birdie is the second harmonic 
>> of one of the L301 tuning frequencies (15.488 GHz). I did a bench 
>> test of harmonic output from the downconverter back in 2015, when a 
>> similar birdie at 29.44 GHz was found; a PDF is attached. It shows 
>> the LO-IF leakage for a wide range of the possible LO drive 
>> frequencies, but the ones applicable to the VLA are between 14.720 
>> and 16.256 GHz, as they are within the passband of the output filter 
>> after the LO tripler.
>>
>> Interestingly, this new one is right in the center of that band, but 
>> my earlier measurement showed no detectable IF output at 15.488 GHz 
>> (which would correspond to a sky frequency of 30.976 GHz). There was 
>> a substantial spike at 15.744 GHz, however.
>>
>> I'll double check this result in the lab today - I happen to have a 
>> good downconverter in the test set and the equipment set up. Rick, in 
>> your observation with this detection, was the L301 set to 15.488 GHz?
>
> L301-1 was 15488
> L301-2 was 13184
>
> Ken
>