Steve,
Great information in your reply - I will dig into this when I get back to work Monday (snow day today). I believe the package is TO-48. Regarding the processing of the epoxy, I totally agree and this might be where I hit the wall as far as data sleuthing (got that from Sherlock!) I have the data pack and there is a process traveler, but still, I don't know what they really did.
Thanks,
Howard Watson
-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Steven Creswick
Sent: Thursday, December 13, 2012 6:58 PM
To: [log in to unmask]
Subject: Re: [TN] Component issue - epoxy die attach
Howard,
Sorry, this will be a bit long.
I am glad you stated that the packages passed hermeticity testing. That would have been the first item to check.
I assume this is TO-5 or smaller style package. The smaller volume packages can be problematic if there is a leak during puncture of the package because the available tested volume is so small compared to a potential leak.
Different labs used to test for RGA in different manners, with different style equipment.
On one style of test equipment, the sample package would be placed into the evacuated test chamber and punctured/broken. The sample gas would then flow into the chamber and subsequently be taken into the mass spec.
Other equipment relied upon placing the package up against the sampling port, sealing it via an o-ring [of sorts]. Once fixed against the evacuated sampling port, a needle would protrude through the center of the o-ring and puncture the package lid, allowing the test gas to enter the mass spec.
Both methods have pro and cons. The chamber method is most convenient for all-ceramic style packages where there is no metal lid to puncture. A problem with it is that the entire exterior of the package must be thoroughly cleaned, but can still carry ad/absorbed species into the test chamber. Plated surfaces also can contain a great deal of trapped hydrogen put down during plating. Some systems would see Hydrogen, and log it as H2O.
With the puncture method, if the lid is too robust, the act of puncturing could displace the sample from the seal and allow atmosphere to be introduced, thereby squirreling up the data. Most hybrids had 0.010-0.020"
thick lids and what WE often did was to take a small end mill [~0.050" dia] and mill a small recess in the lid, leaving only about 0.005" of metal thickness. The lab would than center this in the seal. Generally, this gave us more consistent data.
Variance of data - either approach is prone to variation if everything is not absolutely perfectly cleaned and performed. You did not say how many samples you ran at each lab. Hopefully, you ran 4-6 at each. As a minimum, 3, so you could throw out the low and high, and keep the middle. I am sorry, but to test 1 or 2 is almost fruitless due to variance.
To the adhesive - Two major potential issues come to mind.
First - if you take a great adhesive and improperly process it, you end up with garbage. The 84-1LMI is a very good adhesive which has been used by many firms in Space, Mil, and Implantable medical applications. It can meet the requirements of Mil Std 883, TM 5011 when properly processed. That is a very good material to use - if properly processed!! Depending upon the date of assembly, it was THE material to use. [I am not in any way associated with Ablestik/Henkel, but I have indeed used this material on many hybrids, including many space applications, one of which is still in the Saturnian system]
Second - No matter how well you process the adhesive, if the pre-seal conditioning [pre-seal bakes, vac bakes, package [and lid] cleanliness [and bakes] is not adequate, one will end up trapping trash inside a hermetic package.
Yes the limit is 5000PPM. The real issue is [if the RGA testing is accurate] what other ionics and corruption do you have inside the package to combine with the water??
Whether adhesive is allowed or dis-allowed is generally addressed by the detail specification for the device. I do indeed know of quite a few Space level parts that have adhesive inside.... properly processed adhesives...
No one can really give you a definitive answer, at arms length, such as this, but here are my free two bits ...
I would have to go back and review my dew point knowledge, but I strongly suspect the 28000 PPM is bogus [unless device assembly is really sloppy - see below]. You should be able to achieve well under 2000 PPM without too much grief if it is a TO-5 or smaller package [properly processed].
My fear is that you are buying a commercial grade device, that was assembled and processed as a commercial grade device, having no RGA requirements.
Therefore, the adhesives were likely not processed in a manner which would give you good RGA results. To make a poor analogy, it is like you are buying a standard 75W incandescent bulb and banging it around as though it were a Rough Service bulb. Yes, some will survive the abuse, but most will have the filament destroyed almost immediately because they were not constructed in a manner which would offer a much greater likelihood that they would pass testing. That may be the situation you are in. Sorry.
Hopefully Inge is eyes-open and he can add a few comments as well.
Steve Creswick
Sr Associate - Balanced Enterprise Solutions http://www.linkedin.com/in/stevencreswick
-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Watson, Howard A
Sent: Thursday, December 13, 2012 6:24 PM
To: [log in to unmask]
Subject: [TN] Component issue - epoxy die attach
Hello 'netters,
I debated posting this, as it seems to me to be an obscure problem, but then, I'm always amazed at your knowledge base. I have a JFET 2N6550 component to be used for space application. Unfortunately, it is not manufactured at JANS space quality, so we buy the commercial grade, plus an option 2 screening, which upscreens the part to "like" JANTXV. I think this is called re-branding. Then, we send it to a lab for further upscreening to JANS. The problem is that the parts are failing the moisture test of the residual gas analysis (RGA). I found out that epoxy is used for the die attach, and likely the epoxy is outgasing during subsequent baking as part of the testing. My first question is who knows of a standard for die attach of this component type stating that epoxy is forbidden for military and space use? The epoxy used by the manufacturer is Ablestik p/n 84-1LMI; Material # 1119570. I just found out today that they do have the capability of eutectic die attach, and I'm pursing this option, expecting a huge expense and lead time.
Secondly, I had two independent labs perform the RGA. The first lab had results averaging ~28,000 PPM. The second lab results averaged ~5600 PPM.
The standard is no more than 5000 PPM. They both performed the testing to the same MIL-STD-750. I can't understand the wide range of results, but my second question is who knows of any studies related to the negative effects of excessive (>5000 PPM) moisture inside hermetically sealed devices used in space? By the way, they all passed the seal tests. Perhaps some of you are knowledgeable in this area. Thanks in advance for your help.
Howard Watson
______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] ______________________________________________________________________
______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] ______________________________________________________________________
______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask]
______________________________________________________________________
|
