]Hmmm... any difference in the recommended soldering process between the two? Do either of them have a published recommended handling process for hand-soldering/proto work? -group-therapy-thread/?action=dlattach;attach=938592;image)
mnem
Nope, standard JEDEC default profiles, just like everything else.
Hand soldering 0.5mm pitch 48 pin QFNs? For masochists only and definitely something you won't find in the manufacturer's literature who would expect them only to be reflow soldered.
Hand-soldered doesn't necessarily mean with a iron anymore; hasn't for a long time. 😉
The difference I can see here being whether you're relying 100% on the process control to ensure quality of soldering on product that may never be seen by a human being vs proto work where you have eyes on the product to ensure quality.
The prior relies heavily on process control at every step in chain of custody, where the latter scenario it shouldn't be unreasonable to expect someone to ensure a part gets properly dehydrated before assembly to ensure it doesn't explode when soldered.
mnem
*whirrrrrrrrrrrrrrrr...*
You ever done any pottery? It's exactly the same problem as stuff exploding in the kiln because it hasn't dried out properly first. It has to dry out s-l-o-w-l-y before hand or it goes POP from steam pressure when it's heated up quickly. Ideally however you're reflowing a component, whether on an industrial conveyor oven or waving a hot air stick at it, you want to be following the standard reflow profile as close as you can, which typically means only taking 3-5 minutes or so to go from cold to 200+ centigrade. That's too quick for internal moisture to get out and POP!
Standard recovery treatment for stuff that's been out too long and exceeded its atmospheric moisture times, or been kept uncontrolled, is a 24 hour bake at 125ºC with a slow ramp up at the start. It's not something you can adjust for in prototype work just by taking the reflow slowly or warming the parts up for a few minutes beforehand. If you don't want to risk wasting parts or worse still have incipient part failures you can't see, you stick with the rules for any kind of reflow, oven, wand, production, prototype, it doesn't matter what because the whole part
has to get hot enough for reflow, and that's hot enough for the popcorn effect.
I was explaining this to SWMBO earlier and instead of glazing over and saying "Yes dear" in all the right places she got it straight away because she's a potter. Some clays are more sensitive to this than others, equally some moulding compounds are more sensitive than others. Similarly, geometry has an effect in both IC packaging and pottery. So, what we've got here are two packages of the exact same geometry and exactly the same moulding compound and ST says MSL 3 while Lattice say MSL 1. It makee no sensee. What works for one ought to work for the other; both vendors ought to have followed the JEDEC J-STD-20 test protocol for the parts to arrive at their MSL ratings and they ought to be the same.
The only thing that would make sense is if Lattice's dies are more robust than STs, and ST detected die related damage that Lattice didn't. Without checking, from memory the two parts are at similar technology levels (i.e. semiconductor feature sizes). The ST dies are a
little larger - 3X3mm versus 2x2mm for Lattice (again, from memory).
At the end of the day it's just an oddity, a curiosity. I'll treat each part according to the respective manufacturer recommendations - it won't hurt me any.
Home
Help
Search
Login
Register
