MSL Moisture Sensitivity Level.

[Cicada]

Hello

Can someone please point me to a good source of information about MSL. Specifically a source that address the following questions:

I have several MSL 2A and MSL 3 components that I want to use in projects where I want to build just one of or in some cases batches of boards up to 100. Obviously if you build only one board of a design where you want to use the MSL 3 devices you can hand solder it.
1. Are there any risk (cracking etc) to hand soldering MSL 3 components that have not being stored in a humidity controlled environment at all for long periods of time. Is there a way to remediate such components before hand soldering that can mitigate all risk of it cracking.
2. Are there any risk to hand soldering MSL 3 components that have being stored for a long period of time in sealed bags. I.e. what does the 12 month expiry date on sealed components mean. Do you just have to bake it before hand soldering it. Will it always be ok to bake it if it has being sealed for years?

Typical devices that I want to hand solder are and possibly reflow solder are:
https://www.digikey.com/product-detail/en/xilinx-inc/XC3S200-4PQG208C/122-1339-ND/826942
https://www.digikey.com/product-detail/en/cypress-semiconductor-corp/CY8C24894-24LFXI/428-1917-ND/1206211
https://www.componentsense.com/stock/47478/K4S561632HUC75?utm_source=octopart&utm_medium=cpc&utm_campaign=octopart
Then I would potentially like to reflow this as well.
BGA    https://www.digikey.com/products/en?keywords=AT91SAM9260B-CU-ND

Then in the case where I want to build many boards of a design I will have to reflow solder. So I guess the same questions as above exist for this case. I will be using the same components as above and also possibly some BGA components that have being sealed for years and some that have not being sealed.
3. So how can I remediate devices that have not being sealed at all for years so that I can reflow solder.
4. How how can I remediate devices that have being sealed for years so that I can use them in a reflow solder process.

Or are they all just worthless for reflow soldering and/or hand soldering.

Side note
I have hand soldered FPGAs (XC3S200-4PQG208C) that have being unsealed for years onto boards without having one of them failing.

[elecman14]

You are probably looking for J-STD-033.

Some comments about your numbered questions:
2. The moisture barrier bags have a finite shelf life rating for moisture ingress. You might get lucky and get parts that solder but it is not guaranteed by the mfg. If you want to do this a humidity indicator card in the bag might help indicate if you need to bake again. If the parts are expensive enough a dry box or just baking might be a better solution. Systems to track part expiry exist if you are doing high enough volume to justify that.
3./4. Yes you can do a low temperature bake. Refer to the tables in J-STD-033 for temps and times.

[vealmike]

As above, bake.
The issue is that moisture forms on the device pin and wicks along the pin into the case. When you apply heat to solder the water boils and damages the device.
I've never seen a device (identified as a) fail(ure) for this. For onesies and twosies I really wouldn't worry. For a manufacturing process, bake the parts to dry them out before the solder process.

[eeviking]

If you work in a place with a sensible indoor air humidity (<30-40%) and don't work on something super expensive/important I would not worry too much about msl.
Just store sensitive parts in sealed bags with silica gel and an indicator card. As been said you can always bake parts if in doubt.
But another big problem with long term (several years) storage in a uncontrolled environment is that the pins will oxidise to some degree. So you can get problems with solder not sticking to them.

[Cicada]

But another big problem with long term (several years) storage in a uncontrolled environment is that the pins will oxidise to some degree. So you can get problems with solder not sticking to them.

First of all thank you to everyone that gave me some information.

A question related to the statement above. Is flux not supposed to help with the problem of oxidation at all times? I can understand that one might experience reliability problems with soldering during automated assembly processes if the pins are oxidized too much. The flux in the paste might not be enough or make good contact with the oxidezed surfaces.
But I think during hand assembly one should be able to make sure that the pins do get solder wetted properly by applying enough flux and soldering each pin twice. Or is it the case that with too much oxidation one can easily get bad connections even with ample flux applied during hand soldering.

I found IPC/JEDEC J-STD-033B and IPC/JEDEC J-STD-020C

I hope I will get a chance to read them as soon as possible so that we can continue our discussion on this thread.

[eeviking]

Of course the oxidation is a much bigger problem with paste reflow than hand soldering. But severe oxidation will also make it very difficult to hand solder.
I read your post as you where planning on using ic's in production that had been stored in open air for many years?
Then its definitely something you want to check before running a big batch and end up with a lot of rework. It may be cheaper to just scrap the parts and buy fresh ones.

[Cicada]

Of course the oxidation is a much bigger problem with paste reflow than hand soldering. But severe oxidation will also make it very difficult to hand solder.
I read your post as you where planning on using ic's in production that had been stored in open air for many years?

You are correct in your original assumption.

I also have MSL 1 components with footprints SOIC-8 and SOT-23 that I will hand solder onto some boards and some I would like to put through reflow. So you can see it will be a mix of projects with a mix of assembly methods used. I am not exactly sure how I am going to do it all.

I will inspect the pins and solder some to test them. In your experience what is the best way to evaluate whether the pins are to oxidized. Soldering it or visual inspection under microscope? I have never done the latter with the purpose of checking for oxidation.  So did not pay particular attention while inspecting the components to check for oxidation.

I am reading about the MSL topic over the weekend and I am slowly but surely forming a picture of what is going on. I will type up a post to summarize my findings.

Apart from oxidation on pins which is a separate problem to the MSL issue it seems that baking the devices at 125 degrees celcuis for at least 24 hours seem to remediate the components no matter how long they have being sealed or left in the open.
I am not yet 100% sure if baking at 125 degrees for 24 hours is sufficient in all scenarios (time stored in unsealed state, temperature and humidity) that components where stored in. I will have to look for more sources regarding this, confirming that 125 degrees at 24 hours is sufficient. I also wonder if it is possible to bake the components at a lower temperature for a longer time. And I also wonder if it would be beneficial to control the humidity to be low say 20 or 10 % while doing the baking.

I do have access to a large capacity humidity controlled oven that can do these temperatures.

In the mean time all other comments are welcome.

[eeviking]

Solder and visual inspect joints under microscope. Maybe compare with joints on new components.

The baking times are for worst case moisture levels so it will reset the floor life timer no matter how long the component have been out.
Do not bake too long as baking can also accelerate oxidation. Follow the J-STD-033 table 4-1 and watch out for packaging (tape/reel) max temperature.
E.g. a thin (<1.4mm) MSL3 component will need 9 hours at 125C (trays) or 13 days at 40C (tape/reel)