[Pafgbt] Purple Plague

[Rick Fisher]

Thanks, Matt.  That's good to know.

Looks like the immediate task is to verify the noise and gain performance 
of the amplifiers currently in the Dewar.  Bob, Steve, is there an easy 
way to do this without the PAF back-end and fiber link?

Rick

On Wed, 10 Apr 2013, Matt Morgan wrote:

> Hi Rick,
>
> No, the ones with packaged transistors shouldn't have this problem, unless 
> they are bonded incorrectly inside the package, which as David said is highly 
> unlikely considering it is a commercial part.
>
> So perhaps in this instance the Purple Plague was actually a Red Herring.
>
> Unfortunately, then, It still begs the question to me why so many of the amps 
> appear to have failed at this time.
>
> Matt
>
>
>
> On 4/10/2013 2:21 PM, Rick Fisher wrote:
>>  Matt, I cannot recalled whether you said that Sandy's packaged amplifiers,
>>  like the ones in the PAF, are subject to purple plague.
>>
>>  Rick
>>
>>  On Wed, 10 Apr 2013, Steven White wrote:
>> 
>> >  Figure 2 is the same as the amplifier designated as the spare.  When 
>> >  Mitch
>> >  tested this amplifier, he encounter oscillations in both and could only 
>> >  test
>> >  one amplifier at a bias of 1.2 V.  This may explain the high noise
>> >  temperature measured ~ 13-14K versus Figure 3 & 4 measurements. 
>> >  Yesterday,
>> >  the amplifier was tested in GB and did not oscillate, but the gain was
>> >  closer to 33 dB than 39 dB of S21 measurement of the document.  *IF* all 
>> >  the
>> >  amplifiers in the dewar are the packaged design, the situation may not 
>> >  be as
>> >  dire as if the chip version was used.  We are planning to test the spare
>> >  amplifier cold next week for noise and S parameters.
>> > 
>> >  Also, a swap of IF paths at the output of the dewar is suggested to 
>> >  verify
>> >  the problem, a good channel with a low gain LNA channel.
>> > 
>> >  Steve
>> >  ?
>> >  ?But we definitely need to take some steps to resolve this.  Mainly 
>> >  because
>> >  it is
>> >  ?urgent to let Toney Minter know if we can meet the 2nd week of June
>> >  ?schedule, or need to fall back.  ?He has been alerted to the fact that 
>> >  we
>> >  have
>> >  ?this issue.
>> >  ?
>> >  ?Anish said he was confident that the high noise temp channels were "in 
>> >  the
>> >  ?dewar" and not in the backend.  But it would be good to do something to
>> >  ?prove that irrefutably.  Steve, you suggested a cable swap test.  Or, 
>> >  is
>> >  there
>> >  ?any bias monitoring on the amplifiers themselves?
>> >  ?
>> >  ?Beyond this further diagnostic, we should decide whether there is any 
>> >  more
>> >  ?software/hardware to test (it may be useful to resolve the
>> >  ?attenuation/sensitivity issue), while the receiver is in OTF and cold, 
>> >  or
>> >  ?whether to bring the receiver back to the lab.
>> >  ?
>> > 
>> > 
>
>