To those who responded, thank you for your advice. As I suspected, it won't be a good option to perform any puncture/bake out on these components. 

Yes, I have worked with the company on a resolution, but because they have no MIL-PRF-19500 basic plant certification, they claim their best effort goes into all of their product and workmanship for ALL customers, and what we got is acceptable to their commercial standard. They indicate that the internal volume of the TO52 package is too small to permit accurate IVA moisture content measurement, though this has been refuted by the DPA labs I've worked with. 

Thanks!
Howard A. Watson 
__________________________________________
Intelligence & Space Research Division
Los Alamos National Laboratory
30 Bikini Atoll Road
Los Alamos, NM 87545
[log in to unmask]



-----Original Message-----
From: Yuan-chia Joyce Koo <[log in to unmask]> 
Sent: Tuesday, March 26, 2019 1:51 PM
To: TechNet E-Mail Forum <[log in to unmask]>; Watson, Howard August <[log in to unmask]>
Subject: Re: [TN] Internal Vapor Analysis failure on JFET

puncture can and seal with epoxy is really a bad idea.  Did you contact MFG to work with their engineer?  Recall use commercial parts at up rate temp region - option (1) build up special box control temp that ensure device work within the temp range - that means not work at extreme temp.  (2) MFG usually got few locations for device manufacture, there are bined device very close to the temp range that screen out at burn in level - you pay more to specify in contract only that location with bin X device should be used (make sure your  
project MGR understand why she/he pay more and documented in design.   
so no confusion... you might need pay special qualification cost - not cheap)... you might have to get a special run with minimum qty to meet your needs (including future repair and overhaul).  IMHO.
best of luck.
jk
On Mar 26, 2019, at 12:49 PM, Watson, Howard August wrote:

> Hello TechNet,
>
> I have a problem on a Interfet 2N6550 JFET; specifically the component 
> (TO-52) failed the moisture test having >5000 ppmv. The average value 
> was around 10,000. We purchase these devices at a commercial level and 
> then "up-screen" to a JANS level. I know, this is a terrible idea, but 
> there are no JANS qualified manufacturers for this device and the 
> designers insisted on using it. This is for space flight applications, 
> and I have 500 components.
>
> We are debating mitigation strategies and came up with one that would 
> involve puncturing the can, baking out the components in a nitrogen 
> purged vacuum chamber, and re-sealing the can with a tiny amount of 
> low outgassing epoxy. Is this too pie in the sky, or is there actually 
> a known procedure to do something like this? I might have watched The 
> Martian a few too many times, but I am getting desperate for a 
> solution!
>
> Thanks!
>
> Howard A. Watson
> __________________________________________
> Intelligence & Space Research Division Los Alamos National Laboratory
> 30 Bikini Atoll Road
> Los Alamos, NM 87545
> [log in to unmask]