How To Calculate Component Standoff Height

Phil
And welcome to Board Talk. We're the Assembly Brothers, Jim Hall and Phil Zarrow, of ITM Consulting, and we're here to answer your questions in SMT process, equipment, materials, procedures, and lots of other stuff.

Jim, what's our question today?

Jim
Measuring component height after reflow, or what we might call "standoff." This is from CS in Santa Ana, California. CS writes, "I have an SMT component that has a height of eight millimeters, a stencil thickness of 0.5 millimeters. We are using a lead-free paste. The pad size is 11" by .054" The question, is there a method of determining the overall height of the component on the board after reflow.

I have to believe that the pad dimensions of 11" by .054" is incorrect. I would have to guess if that's supposed to be .011" .110". It's probably .110"

Phil, how would you answer CS?

Phil
Well, the question, is there a method for determining the overall height of the component on the board after reflow, the answer is yes.

But from the information given, whether that one dimension was accurate or not, you don't really give us enough information. What do we need to know to do that, Jim?

Jim
Well, I don't know that I could it anyway. This is a multi-parameter mission. You've provided some of the details, the component, the thickness. You've given us the thickness of the stencil. The stencil, 0.5 millimeters, that would be a 20 mil thick stencil.

Phil
What are we soldering here? 

Jim
Stencils are usually 4 to 8 mils thick, not a 20 mils thick.

Phil
What are the parameters we're looking for?

Jim
He's given us the pad size, the thickness of the stencil, assuming we know what that is. We'd have to know what the aperture is relative to the size of the pad. But most importantly, he doesn't tell us what type of component it is.

Jim
Is this a QFP? Is this a chip? Is this a BGA?

You also want to know the weight of the component. But even then, doing an exact calculation is pretty difficult.

Most people, I believe, who are interested in this data will actually do some physical measurements for their system. If you're using it to predict the thickness of a stencil, again, you'd probably want to go on some correlation.

Jim
Also, I want to point out that he's talking the height of the component. Typically we talk about that in terms of standoff. That is the distance between the board and the bottom surface of the component. Of course, we could just add that to this eight millimeters thick component. But does that height include the leads if this is a leaded component or BGA? So more information is needed.

Phil
Well, there you go. That's a full baked answer to a half baked question. This has been Board Talk with Jim and Phil, the Assembly Brothers. And remember, whatever you do ...

Jim
Don't solder like my brother.

Phil
And don't solder like my brother.

Jim
And keep the kids away from the flux spot.

Unit standoff is -125 microns SOT devices, manifest lifted leads, is there a possibility that this will still be soldered during board mount?

The Assembly Brothers are feeling a little grumpy today?

It seems like a pretty clear question. And the answer appears to be "no, there is no way". It would be nice to have a general purpose maximum for things like collision with mechanical items that overhang the board. Also to know the pressure on TIM in the best and worst cases knowing the likely and limiting height is important. For all purposes I have always just used the component's listed maximum height from the datasheet (which appears to be conservative) and I haven't had problems yet, even without explicitly account for the standoff height. But I have only designed servers and set-top boxes where the height constraints are only important on a few components at the edges and under heat sinks.

I would think that the tolerance of the 8 mm package dimensions would dominate the overall height along with perpendicularity of the component following reflow. The stencil thickness and standoff following reflow are variables with small influence over the overall height.

JEDEC has a set of "MO" drawings that list various parameters for BGA components for the memory industry. These drawings typically list package size (including thickness), BGA ball size, and pad size (including copper defined pads and solder mask defined pads). Where applicable, they include min, nominal, and max dimensions. Search for "MO-207" (two alphas, three numerics) and you'll see dimensions for much of what was discussed above on MO-207 sheets 1 and 2.

Component height is the distance measured from the PWB surface to the TOP of the component. Standoff is the distance measured from the PWB surface to the BOTTOM of the component. Standoff is important because you want to be able to clean under parts, and for certain components the height of the solder joint itself determines its modulus of elasticity and thus its reliability (google up Engelmaier/Wild equation). Max. component height is also important to ensure form, fit, and functionality of the finished CCA or device. There is no exact way to calculate what either measurement is going to be. You can get close, but there are too many variables that you have no control over.

Whilst one can calculate what the approximate printed paste volume might be given the pad size, the aperture size, and the stencil thickness, there is no way of knowing what the two finished heights will be because there is too much variation in the PWBs etched pad dimensions, the printed paste volume, % metal content, reflow profile, time above solidus (wetting time), component solderability (the better the wetting, the lower the part will sit to the pads), flux action, wetting tension of the molten solder (not always at equilibrium between pads), lot-to-lot variation in component size, humidity, contamination levels, thermal gradients between multiple components of the same dimensions but with different connections of their pads to the internal layers (some will heatsink more than others, and thus not wet the same exact way) and about a thousand other variables that you cannot perfectly control with any kind of repeatability to ensure the calculations are correct.

I'm not sure why you need to calculate this, but I am guessing you want to either ensure cleanability, maintain a target solder joint height for reliability, or make sure you have proper clearance. The only way to know what the finished heights will be is to have a PWB house build a small sample of prototype boards and average them out across many different locations. But that will give you only a rough estimate although it might be sufficient.

I was hoping to get a little more real practical information out of this "How To Calculate Component Standoff Height" article. Partly because I've had to do this on a board packed with various BGA's and LGA's and a couple of SOIC's and QFN's that we have needed thermal Pads/heat-sink to contact the differing heights.

It seems as though there would/might be a chart that indicates height ranges for 256 pin BGA or 144 pin LGA added to the height of component. Thanks.