Do We Still Need SMT Adhesives?

Phil
Welcome to Board Talk. We are the Assembly Brothers, Jim Hall and Phil Zarrow of ITM Consulting. We're here to attempt to answer your SMT process assembly questions. 

Jim, what's today's question and who's it from?

Jim
Well, it's from D.A. The question is: If double-sided surface mount assemblies are the way of the future - weren't they the way of the past - should we anticipate a growing or declining need for surface mount adhesives?

This is interesting because there could be two questions here.

Surface mount - the question is the interpretation of the term surface mount adhesives. Are they asking, because we're going to double-sided surface mount, will I be gluing fewer chip components on the bottoms of my boards and wave soldering and therefore I will need less surface mount adhesives?

Or are they asking because of double-sided assembly, am I going to need more adhesive to glue the parts on the second side of the board which ends up on the bottom during the second reflow pass.

Phil
I would say much to the chagrin of the adhesive companies, the answer to that is yes and no. Yes, you will be using less because running surface mount components through the wave solder process has become very out of vogue. It frankly was never a good idea and never defect-free by any means, and that's been kind of frowned upon for a long time.

And as we see the increase in the mass pallets and more and more select soldering systems, wave soldering surface mount components has gone away. So, using adhesive for that purpose certainly is on a decline.

Jim, differentiate what we mean by side A and side B, first side and second side.

Jim
There is absolutely no universal terminology. So, let's define it for our conversation today. We say the bottom side of the assembly is the side we reflow first and the parts end up on the bottom side during second reflow. The top side is the side we reflow second.

Phil
So, on those surface mount components that are being re-reflowed, the old theory was that they would fall off when they would be re-reflowed. However, what people soon discovered, the truth of the matter is they don't most of the time.

Because the surface tension via the molten solder is a heck of a lot greater than wet solder paste. So, what they found was it really depended upon the ratio of the surface area of the cumulative leads and the mass of the component.

There were some experiments done and the formula we generally use is you take the mass of a component in grams and you divide it by the total pad surface area in inches and if you come out with, on the conservative side, a ratio of 35 grams per square inch, that part's not going to fall off.

We've even gone higher. We actually did some connectors that were working in excess of 44 grams per square inch. But at the 35 grams per square inch, you're typically talking about just about any QFP and certainly fine pitch.

Jim
It's counter-intuitive to the people who first looked at it. They'd look at the small parts and say well they don't weigh very much, therefore they won't fall off. But the big parts will.

And it's really the opposite because the big parts, for the most part, tend to have many leads and each one of those leads and pads adds its own surface tension, so the cumulative surface tension is more than enough to hold the part on.

The only exceptions, and there are some, tend to be large heavy parts ceramic parts that only have a few leads.

Phil
In most design schemes your IC's are going to be on the second side, the top side. But where you are faced with them being on the bottom side, take a look, do this ratio and don't do any unnecessary process steps, particularly gluing them.

I think we answered that question. This is Jim Hall and Phil Zarrow, Board Talk, saying whatever you do, wherever you go don't solder like my brother.

Jim
And don't solder like my brother.

Informative as always, and I really do appreciate the reference to Tom & Ray at the end.

As with the case of the double sided board from the previous comment, there are still countless circumstances where a bonding adhesive is the best solution.

Light-curable materials offer the desired bond in seconds and will not damage components. These results are what will keep adhesives in the SMT market for years to come.

I have a hot air machine which I use for most of my assembly work. I recently have had to build a double sided board which has 2 IC's, one on each side and they both have a center ground plane which covers the majority of the package.

The problem I have had is, after placing down the smaller IC and turning the board over to place the larger IC, the smaller IC is dropping to some degree and loosing some contact with the pins and center ground. I have switched to do the larger component first then turn the board over and do the smaller IC. Same results are happening.

I have taped the first IC into place after soldering it down and turning the board using Kapton tape and it still slides around when placing the second IC on the board. I have never worked with any of the bonding material and am wondering whether this will take care of my issues.