Funky Footprints

Parts with uneven solder pads cause headaches on a fairly regular basis when matched with PCB land patterns that don't also follow the uneven dimensions.

The actual layout, as seen on the left, won't work too well. The surface tension of the molten solder will act on the bigger surface area of the pad on the left and cause the part to shift to the left during reflow. That can lead to reliability issues. In some similar cases, it can also lead to shorts with nearby components or mechanical structures. This part will end up pretty close to the cut-out edge and something structural there could interfere or hit and stress the mechanical bond.

Pad_causes_part_to_shift_in_oven_pr

In the simulated, more betterer landing pad on the right in the red box, both pad sizes on the PCB match in size both of the solder areas on the part. This way, surface tension will act equally on both sides and the part will stay right where it's supposed to.

Duane Benson
Keep off the mask, ya young whippersnapper...

Solder Lug vs. PC Mount

Solder_lug_vs_pc_mount_2 It can happen to the best of us. This is just a quick little cautionary tale about package form-factors. The CAD part footprint is most likely as intended here. But, the part is clearly a solder lug part where a PC mount version was called for.

It's also possible that a relay socket is supposed to be mounted onto the PCB so the relay can later be plugged into it. In either case, a quick form-factor vs. land pad check could have prevented this sad situation.

I think this is an LY suffix relay. I'm not totally sure without the part or part number in front of me, but close enough for this discussion. If it is, I found the PC mount socket in Digi-Key for $1.48. The same catalog page also had a PC mount version of the relay so either way, socket or different part, this is a fairly easy, but still annoying problem to fix.

Duane Benson
Can you fit a slot lug into a round hole?

Via in Pad - Right and Wrong

Here's a pretty simple via-in-pad example that has both good and bad. It's an SOIC-8 so it's a pretty big part. It's not super roomy, but there is plenty of space on the PCB to do things a little different.

Via_in_pad_right_and_wrong Note pin 1. It does have a via connected to the lead land pad, but it also has a little strip of soldermask - a dam between the contact area of the land pad and the via. That little mask dam will stop solder from flowing into the via and everybody will be happy. Well, some people just have a sour disposition so they'll probably never be happy, but at least the assembled PCB will be happy.

Pins 5 through 8, though, are a different story. They have an unobstructed metal path from the land area to the via. It's quite likely that the capillary action of the via will suck the solder off of the lands leading to an unreliable connection. Since one side is all okay and the other is all not okay, you might see a tug-o-war of sorts with the surface tension of the good side pulling the chip too far that way.

My advice to you: Move the vias a hair further away from the pad and put a little soldermask dam in there, just like was done on pin 1.

Duane Benson
Soldermask is as soldermask does.

Let's Get Small

150pxmc68451_p1160081 I was just doing some research about the character of the boards we assemble here at Screaming Circuits and some of the data is pretty interesting. I looked at work we've done from 2005 until now and split out the types of components that we put on the boards.

100pxthree_ic_circuit_chips We don't ask a lot of questions, so there is some information that we can't get. Like, for example, we count all discrete SMT parts that aren't fine pitch or leadless as the same thing. So, an SOIC-24 is lumped in with an 0201 resistor. It would be interesting to track the size of passive components over time, but we just don't ask that question.

The character of the orders is pretty similar. The average job in 2005 had just about the same number of individual PCBs to put parts on to as did the average job so far in 2008. They both had virtually the same number of line items on their Bill of Materials.

Lmx9820a_bluetooth_14116bga_150 The number of thru-hole components per board has dropped by 13% over that time. That's understandable. Thru-hole are getting more expensive and things like reliable connectors are starting to become more common in SMT form factors.

Fine pitch parts - leaded parts with 20 mil or less lead spacing, like QFPs and those thin flat-pack flash chips - have gone down more that any other type of part - by 20%. At first that reallyQfn1sm  surprised me, but after thinking for a bit, I began to suspect that these are moving to BGAs and QFNs. The flat pack flash chips do seem to be going to BGAs and microcontrollers that used to show up as thru-hole or fine pitch QFPs seem to be coming more commonly as QFNs.

Leadless parts, like BGAs and QFNs are up by nearly a factor of three! This is very much to be expected. More and more parts are moving to these smaller footprints. I wish we could split out BGAs, microBGAs (.5mm or less ball pitch), QFNs and CSPs.

Last year, the QFN seemed to be all the rage. ZigBee radios are all QFNs. Micro controllers seem to be going to QFNs and even dual row QFNs. We're seeing a lot more .5mm pitch BGAs too. The ARM32 processors all seem to by .5mm BGAs. Bluetooth modules are mostly small BGAs also.

The surprise to me is the power components. Things that used to be in big hulking TO-220, TO-263 and similar packages are now showing up as QFNs and Chip scale micro BGAs. I ran across a little 1 Amp motor driver in an 8 lead QFN package a while back. Amazing. A lot of the new Buck and/or boost chips are showing up in chip scale micro BGAs too. Same for serial flash chips.

Duane Benson
We're aware it's a small world after all. Are you?

Spock Help Me Find A Product Manager

Well, it wasn't actually Spock. It was Leonard Nimoy and the TV show "In Search of..." sought out things much more interesting then new employees. Well, unless you happen to be a good product manager and are looking for a new job at a fast growing company.

If you are such person (we're looking for a few other folks too), jump over to our jobs page and see what's up.

Maybe if I can get some more help here, I'll be able to get back to more regular blog posting. That would be a good thing. Even better, if any of you have any cool and or interesting and useful tidbits, try and write up a guest blog post. If you want to give that a try, email it to marketing@screamingcircuits.com. I'll take a look and see it if fits.

Duane Benson

Footprint Mismatch, Part II

A question just came in about the "Footprint Dimension Mismatch" post below:

8mm_on_a_30mil_pitch_footprint_sm "One thing that always makes me wonder is that, despite the footprint mismatch, does this still fall within IPC Class II standards? As far as I know, that's what you guys assemble to, and I heard somewhere that as long as 50% of the lead is aligned over the pad it falls within Class II spec. Is that true?"

It is a good question and does a bit to illustrate the challenge with standards. According to our engineers here who are very familiar with the IPC standards, it does meet IPC Class 2 (IPC-A-610 Rev D, Section 8.2.5 pg 8.41-8.53). So, this does in fact comply with the IPC Class 2 standard.

Our engineer states: "Technically, yes, these parts would pass the IPC 50% rule, assuming the wetting is good and there is wicking / a fillet where lead is on the pad. That being said, the picture indicates a pad pitch issue that amplifies as you move out from the center pads in either direction. I'm not a structural expert, but have to believe this pad-to-part mismatch cannot be good for mechanical integrity. It also increases the chance of shorting between pins if placement is slightly off or the part shifts as the solder reflows." We will still notify the customer in a case like this, but since it does comply with the standard, we would go ahead and build it.

In a prototype environment, that's most likely okay. Since it meets the IPC standard, we can defer to the judgement of the customer and go ahead and build it. Again, if this were a production environment, it would not be okay even though it meets the standard.

Order Status Emails and eMail Filters

Hi all;

When you place an order with us, we send out periodic status updates via email. We also send you any questions that way. If we don't get an answer to these questions, your job might be delayed, so they are pretty important emails.

We've been finding more and more cases where these emails are blocked by corporate spam filters. We send these notifications from orders@screamingcircuits.com. We only used this address for your job status updates to you, so if you are doing business with us, you should make sure that this address is whitelisted or not blocked by your company spam filters.

Don't worry about getting other mail from us on this address. Our newsletters come from a different address: sales@screamingcircuits.com and you can always opt out of that from your "My Accounts" page on our website or from the bottom of the newsletter.

Duane Benson
but I don't like Spam!

« May 2008 | Main | July 2008 »