Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


component baking

Ken Bliss

#21204

component baking | 16 August, 2002

I have a question for everyone about component baking. In my view this is rework. The component was manufactured and shipped correctly and should not need baking in order to use it. Some how in storing the component in a stockroom atmosphere (I assume high humidity) these components are picking up moisture.

Is it not cheaper to store them all in a enclosed cabinet with nitrogen or other system. I see on the past r-kives several discussions on the topic, but I did not see a preventative solution other than baking them.

Is there a new piece of equipment needed to reduce these costs, time, and labor of baking. If there is please tell me what is needed to solve the problem.

These costs just add to the industry not being competitive enough. Lets solve it, but how do we....

reply »

#21206

component baking | 16 August, 2002

Baking is more like damage control than a preventative measure. Baking is used when life goes wrong, as it so often does. Storing MSD is not the issue. Not storing MSD is. Central services are a thing of the past. Distributed processing is were people want be.

We have been talking about solutions to storing MSD on SMTnet since the beginning of time. Why sonny, I recall a thread back in �99 when Chris G. was talking about converting a frig to a N2 cabinet. Now those were the days, eh?

Storing MSD: Storing devices in a humidity-free cabinet until seconds before the component is required would be real nice. Goody, everyone gets his or her own cabinet. Finally, we�ve been saying that for years.

Scale of the problem * Moisture sensitivity is not a NEW problem. Years ago, we baked all of our boards prior to mass soldering. We still schedule a bake of multiwire boards prior to paste print. * Some tantalum capacitors take on moisture and we solder them after board cleaning. * 98 per cent of our MSD don�t cause us to miss a beat. * Accounting for exposure time is not a trivial problem. Firms sell bar code software to help manage this issue.

For us, �poor planning� drives to the requirement for MSD baking. Just about everyone stirs that fine kettle of soup [Marseillais fine bouillabaisse, I think].

General observations are: * �In my view this is rework.� => Could be. It does indicate imperfection. * �The component was manufactured and shipped correctly and should not need baking in order to use it.� => Sounds good. Most classes of MSD packaging have a self-life. Lots of times it�s expired before it comes in-house. * �The component was manufactured and shipped correctly and should not need baking in order to use it.� => Sounds good. Try as we might, most MSD package quantities does not match-up with product lot sizes.

reply »

Ken Bliss

#21207

component baking | 16 August, 2002

Hi Dave

Interesting points. If I understand you correctly

1. about 2% of part numbers have a moisture problem

2. Packaging frequently comes in with expired dates on them. This seems hard to believe you would accept them.

3. To really solve the problem you would need a cabinet near each area of operation so the exposure time to air is very short prior to soldering. could this be a couple of hours or literally a few minutes.

A couple more questions then.

1. Does vacuum sealed packing not stop this or does it only reduce it.

2. If there was a small economical nitrogen or equal box near each use would that solve it.

3. Any guess at the dollars worth of components per month a typical factory has to deal with with this problem.

reply »

#21213

component baking | 17 August, 2002

Clarifying: 1. About 2% of part numbers have a moisture problem <= About 2% of MSD part numbers require extra-ordinary MSD handling, like baking. 2. Packaging frequently comes in with expired dates on them. This seems hard to believe you would accept them. <= When you�re hungry, you find something to eat. 3. To really solve the problem you would need a cabinet near each area of operation so the exposure time to air is very short prior to soldering. Could this be a couple of hours or literally a few minutes? <= Yes. Some of the issues are: * Routine product build that is held-up for some reason, like the wheels falling off a machine. * First-side components waiting for second side reflow. * SMT components waiting for wave soldering. * Opened dry-packs that were not consumed, but will be used some other time. A couple more questions then. 1. Does vacuum sealed packing not stop this or does it only reduce it. <= We vacuum seal not consumed material from opened dry packs, after they are baked, prior to their return to stock. It�s rare for us to vacuum seal in-process material. A line has to be unplanned going-down hard for that to be done. 2. If there was a small economical nitrogen or equal box near each use would that solve it. <= Not too small, but economical is good. 3. Any guess at the dollars worth of components per month a typical factory has to deal with this problem. <=I�ll ponder this.

reply »

#21258

component baking | 21 August, 2002

What is the correct method to bake or cure the component?Standard baking practise by us is to place the units into oven at 125 �C for 8 hrs directly. But I was told that the correct baking practise is to cool down the Baking Oven to Room temp and only then the units be put inside the oven and let the temperature gradually goes up to 125 �C for 8 hrs in the Oven. Reason given was that direct exposure to 125 �C temperature will caused moisture trapped in the device's package to expand more rapid and may caused delamination to certain package type. Pls. advised if this is true.

reply »

fmonette

#21402

component baking | 3 September, 2002

I have never heard about a requirement for temperature ramp-up for baking components at 125C. In general the concern for delamination with moisture sensitive components starts above 200C (peak body temp).

Most important is you should bake your components for a lot longer than 8 hours. The default duration at 125C is 48 hours.

I suggest you download J-STD-033 from the JEDEC web site (Rev.A has just been released) and compare this with your internal procedure. I can send you an audit checklist to simplify this process if you are interested.

Regards,

Francois Monette Cogiscan Inc. Tel : 450-534-2644 Fax : 450-534-0092 fmonette@cogiscan.com www.cogiscan.com

reply »

Stephen

#21507

component baking | 12 September, 2002

http://www.jedec.org/download/search/jstd020b.pdf http://www.jedec.org/download/search/jstd033a.pdf

The above links are for JEDEC standards regarding moisture sensitivity.

SMT magazine has an article a few months ago regarding a company that saved money by putting radio tranmitters on all MSD trays. If I have time later I will see if the article is online.

If MSDs are exposed to atmosphere and then sealed the moisture can migrate in the package. Therefore total time exposed is not the only consideration. "high" humidity leavels are such that if your staff isn't getting nose bleeds from the dry air, then the humidity is too high for MSDs.

In theory there is no difference between theory and practice, in practice there is.

reply »

#21508

component baking | 12 September, 2002

"In theory there is no difference between theory and practice, in practice there is."

That's beautiful...is it yours?

reply »

Stephen

#21509

component baking | 12 September, 2002

I found the article online that I was talking about. It was in Cirucuits Assembly. http://www.circuitsassembly.com/mag/0205/0205cmac.html

I found the artcile quite interesting

reply »

Reflow Oven

MSD Dry Cabinets