Pre-baking derating

A user of one of our components exceeded the floor life of some plastic BGA components. They did do a pre-bake prior to SMT reflow but only 72 hrs at 40C and not the recommended 12 hours at 125C. I've recommended accoustic microscopy to look for popcorn/delam.

This is moisture level 3 part and we have component moisture absorption curve at 30C/60% and a desorption curve at 125C.

Is there any way of deriving a correlation between time spent 40C and an equivalency to time spent at 125C?

I was thinking of Arrhenius relationship, indeed J-STD-020 shows some kind of accleration factor between 30C/60%RH and 60C/60%RH.

reply »

| A user of one of our components exceeded the floor life of some plastic BGA components. They did do a pre-bake prior to SMT reflow but only 72 hrs at 40C and not the recommended 12 hours at 125C. I've recommended accoustic microscopy to look for popcorn/delam. | | This is moisture level 3 part and we have component moisture absorption curve at 30C/60% and a desorption curve at 125C. | | Is there any way of deriving a correlation between time spent 40C and an equivalency to time spent at 125C? | | I was thinking of Arrhenius relationship, indeed J-STD-020 shows some kind of accleration factor between 30C/60%RH and 60C/60%RH. |

Paul, just a suggestion for a long-term fix..... Why not attack the root cause of the issue? Me thinketh you make this too complicated.

Ensure that your mfg. facility has good control of ambient RH. Without it, you are "pissing in the wind" for your process, anyway. I built a new factory some years ago for a 20 Mil / year contract mfg'er, and spec'ed +/- 2 deg F and +/- 5% RH swing across the expanse of the facility--In FL!

BOY, did I have a battle with management to get that Air Handling and control system purchased. But you know what? Our yields were phenomenal. I could use RA, RMA, Water Soluble, and No-Clean with no degredation to the SMT and soldering processes--and saved the expense of using dry nitrogen.

Next, Invest in a vacuum sealing machine (<5K for a good one), some dessicant material / bags, and humidity strip indicators.

Force the humidity control back on the vendor by specifying the packaging in vacuum packaging with dessicant and a humidity strip indicator. This is where this belongs, anyway. Do not do biz with any vendor who won't comply. I remember fighting with the vendors in the early days of QFP's for this type of pkging. They didn't want to do it. But if they wanted the biz of a top 3 contract mfg'er, they did!

Put a policy / discipline out on the floor which states that once a humidity pack is broken open, you have a predetermined amount of time to use the parts before they must be repackaged in vacuum with dessicant and humidity strip indicators. In the facility I mention above, I stated max of 72 hours.

Scott Cook

reply »

| A user of one of our components exceeded the floor life of some plastic BGA components. They did do a pre-bake prior to SMT reflow but only 72 hrs at 40C and not the recommended 12 hours at 125C. I've recommended accoustic microscopy to look for popcorn/delam. | | This is moisture level 3 part and we have component moisture absorption curve at 30C/60% and a desorption curve at 125C. | | Is there any way of deriving a correlation between time spent 40C and an equivalency to time spent at 125C? | | I was thinking of Arrhenius relationship, indeed J-STD-020 shows some kind of accleration factor between 30C/60%RH and 60C/60%RH. Scott's advice is good. If the people on the floor don't have an enforced policy/proceedure to follow you are going to be popping these puppies. As far as baking time, most manufacturers will recommend 40c for 192 hours or 125c for 24 hours. The 72 (level 4) hour open time might be a little restrictive and involve extra work as the majority of the moisture sensitive devices are level 3 with a 168 (7 day) permissable open time. On the flip side, you mentioned PBGAs and some of these are level 5 or 5a with open times of 48 or 24 hours respectively. If you have a library of old Circuit Assembly Magazines, look at Sept/Oct 1996 for a good article.

reply »

| A user of one of our components exceeded the floor life of some plastic BGA components. They did do a pre-bake prior to SMT reflow but only 72 hrs at 40C and not the recommended 12 hours at 125C. I've recommended accoustic microscopy to look for popcorn/delam. | | This is moisture level 3 part and we have component moisture absorption curve at 30C/60% and a desorption curve at 125C. | | Is there any way of deriving a correlation between time spent 40C and an equivalency to time spent at 125C? | | I was thinking of Arrhenius relationship, indeed J-STD-020 shows some kind of accleration factor between 30C/60%RH and 60C/60%RH. | whoops - see also the new IPC standard J-STD-033

reply »