Screaming Circuits: February 2011


Hot Time in the Small Chips Tonight

Years ago when I worked for a local projector company, we were introducing a relatively compact projector lighted by a Halogen lamp. Cooling such bulb in a big, wide open projector wasn't a problem, but we barely had a few cubic inches of open space around the 400 Watt "heating element" in our projector. Our engineers had to dig up old and nearly lost information about cooling high-powered vacuum tubes in constrained places.

We don't have to worry about cooling glass devices, and big processors, big regulators and other big power components have needed heatsinks and fans for quite a while, so cooling isn't really a new science. But the science of cooling is changing. We're seeing more and more tiny components needing advanced power dissipation techniques.

With our projector bulbs, just sticking a fan next to the bulb wasn't good enough. These new tiny power components, like the MCP1726 regulator, have a similar issue. You can't just stick a heat sink on them and call it good. You need to engineer the cooling system with thermal planes, thermal vias and other layout considerations. Some, like the CMLDM7484, a dual MOSFET from Central Semiconductor, in a 1.7mm x 1.7mm package, ask for aluminum or ceramic core PCBs to survive its maximum power dissipation. Using the PCB for cooling can be a lot more complex than cooling with heatsinks and fans. Anyone remember how to cool a vacuum tube base area and PCB surrounding it?

Duane Benson
Kelvons have feet, but photons can fly.

Possible Weather Cause Delays

BULLETIN:

The snow is accumulating in the Portland area and we have declared today (2/24/11) to be a snow-affected day. We're expecting delivery and shipment problems and many employees are experiencing difficulty getting in to the shop. We will still do our best to ship your jobs, but unfortunately because of the weather, we can not guarantee shipment turn-times right now. Weather related delays do not count toward your turn times.

I'll keep you posted.

NOTE: For those of you that live in blizzard-ville and can't see what the problem is with such a tiny, by your standards, amount of snow; we don't have the equipment around here. Yes, it's nuts that this little bit should bother us, but nothing ever happens out here so the cities, counties and states really don't have the facilities to make it all better right away.

 

The Most Important Thing

Actually, there are other things that may very well be more important. Oxygen and food certainly are. Roads are pretty important as well. Life would be difficult without them. But, for a job, a career, an  100px-Three_IC_circuit_chips Industry, I can't think of much that is more important than the very last sentence of the linked article here.

It's not difficult to find literature discussing the challenges facing the technology industry this country. There's always talk of the burdensome regulatory climate, unequal labor costs, government subsidies, dumping and on and on. But, what if we didn't focus on what's making things difficult for us? What if we, rather, focus on what we can do to make things better? Again, look at that last sentence (read the rest of the article too) and think about it.

Duane Benson
Zero - Ye ha, 500 blog posts!
(Is that a lot?)

Outsourced Out of a Future?

Alert! - Self-serving Content Ahead: I suppose I'm supposed to be self-serving now and then. I mean, I don't personally design and build things for a living. I do that for fun and dim hopes of robot world domination. These days, I tell people about things for a living so that makes it my job to be mostly self-serving ("self" being defined as "my company"). But self-serving isn't always bad. I couldn't get gas for my car here in Oregon without being self-serving. I could get hamburgers though, so I'll have to conclude that it's a 50-50 proposition.

We understand outsourcing here at Screaming Circuits. That's what we do - take people's outsourced prototype and short-run production assembly work. Being located in Canby Oregon, USA, we see both sides of the outsourcing discussion. As I said, on the one hand, we do assembly for other people. On the other hand, being a North American manufacturing company, we've seen a lot of work go offshore. Not that I'm against things not in the U.S. - I firmly believe that most of this country's success is due to the fact that we're from here, there and all over the world. But, I do want to have a job and I want my friends and family to have a job. We have to be worth something in this country and the rampant pace of off-shoring sometimes makes me wonder if we're just all deluding ourselves in that regard.

The prompt for this post is this article in the Los Angeles Times about Boeing's outsourcing in the 787 Dreamliner program. The quick summary of the article is to speculate that Boeing went way too far in their outsourcing and have put at risk not only the financial success of the program, but also the company's future engineering prowess.

Now, here's the self-serving part. The article outlines how they went wrong by over-outsourcing, but it also points to the value of specialty companies like Screaming Circuits:

"That's not to say that outsourcing never makes sense — it's a good way to make use of the precision skills of specialty manufacturers, which would be costly to duplicate." [from the above linked LA Time article]

That's us. We specialize in prototypes and small volumes. We specialize in new and difficult component package sizes. We see such a variety of different types of designs here in our shop that we get good at things like QFNs and Micro BGAs sooner than anyone else. (Hyperbole, perhaps, but I do believe it none the less). We've built things that go under water, up into space and everywhere in between. We don't specialize in one or a few specific vertical markets, like medical or consumer, we specialize in the prototype phase of the development program across virtually all market categories.

So, outsourcing: I'm in favor of intelligent outsourcing. My advice to you: Outsource where it make sense. Don't outsource where it doesn't. Look at the true cost of such decisions, not just the surface image. Keep some value add in your company and don't just become a marketing shell.

Duane Benson
Time to make my oatmeal.
One

Drip. Drip. Drip. Drip. Drip...

MSD logo Okay, water doesn't have to actually be dripping down on to your parts to be a problem. I don't know if you've heard or not, but water also comes in vapor form. Weird. I wonder if there's a way to harness water in that vapor stage and do something useful with it. Hmmm. Ponder material for another day.

Here is something not so useful about vaporous water: It can get into your parts and make them unhappy. I was recently asked about opening and resealing moisture barrier packages for moisture sensitive parts. This old post has a link to the JDEC standards document (J-STD-033B.1) covering this subject. It's can be a complicated subject and the document is worth the read.

In sort, the best thing to do is just leave the parts sealed in their original moisture barrier packaging. If that's not practical, look at the MSD classification. There are eight levels with one being least sensitive and six being most. (Levels are 1, 2, 2A, 3, 4, 5, 5A and 6). Most parts seem to be level 3, which can be open and exposed for a cumulative time of 168 hours. The actual safe time may vary based on your local humidity.

If you want to open and reseal, you'll need the humidity indicator card that came with the parts, desiccant and a thermal sealing gizmo. Open the package, take out the parts you need, put the remainders, desiccant and card back in the package and reseal it. Sealing it with tape won't do the trick. Then you would count the time that the components were in the open air toward the cumulative open time.

If there's any doubt, just let the assembly house know that the parts need to be baked. It will probably add some time to your job, but it's better to add a bit of time than have bad parts.

Duane Benson
Add walnuts and chocolate chips.
Two

Hello... Any Good Part Shipping Ideas?

Ever have one of those situations where there really aren't any good answers? There may be right answers, but not necessarily any that fit real well to the specific situation or are all that convenient. The question at hand relates to packaging and shipping small quantities of big ICs to your prototype house.

If you buy them in small quantities, go ahead and use that packaging to ship to Screaming Circuits. That's not a problem. The problem (and this question) comes in when you need to purchase a bunch of them but only need to use a few at a time. This is primarily an issue for parts that come in trays.

Let's say you have a 23 x 23mm BGA or QFP that came in a tray with 60 parts. You need to ship five of the parts for an assembly order at Screaming Circuits. We tell you not to ship them loose. You don't want to risk in-transit loss or damage to all 60 parts by shipping the full tray. What do you do?

For leadless parts like BGAs, LGAs and QFNs, you can VERY carefully pack each one in a small individual anti-static bag and then wrap them in bubble wrap. Make sure you don't damage any of the solder balls. You can't do that with leaded parts like QFPs. Don't do it. Bent leads don't solder well. And, don't put more than one BGA in an antic-static bag. Missing solder balls won't solder either.

For all types of chips, you can find someone that sells JEDEC Matrix IC Trays and see if they have any that meet your size and capacity requirements. www.Topline.tv or www.practicalcomponents.com are good places to start. That won't help much if it's Tuesday, noon and you need to make a shipping deadline of 3:00pm. But it will likely help for future projects. The other problem with this approach is that without the proper tools, it may be very difficult to get the parts out of one tray and into the other without bending a few pins.

The best option may very well be to just send in the full tray and have us send them back after we've assembled for that job. If you do choose to send in your full tray, we will treat all of the parts with care, bake if needed, properly reseal and pack them for the return trip to you.

Duane Benson
Three

How Many Spins?

The other day I wrote about my failure to follow my own advice. Obviously, advice is only for someone else. Just like the best standards are double. Right?

Hmmm. It got me to thinking about board spins. Years ago, I remember products produced by the company I worked for often coming out with double-digits worth of mod wires in production PC boards. I think with the ability to turn PCBs in a day or a few, that rarely happens anymore. But what about in the prototype stage?

Here at Screaming Circuits, surprisingly few repeats show up other than from people using us for small-lot production. We do see a lot of layout issues here, but likely we see a lot because we see a lot times a big multiplier of different designs here.

For my little dohickeys, I seem to need about one board spin due to design or layout problems for each five designs. Of course, mine are pretty simple. Most of my boards spins are due to me coming up with better ideas after using the thing for a while.

If my supposition is true that mods are required less often now, is it because designers are better now, tools are better now or components are better now? How many times do you typically re-spin a PCB due to design or layout problems?

Duane Benson
Four

Hi Ho. Hi Ho. It's off to Scone I go

I found this video over on the adafruit blog. Asimo, by Honda, is cool. Mars rovers, by JPL, are cool. Packbots, by iRobot, are cool. But, for robots to really take off, they need to be able to run down to McDonald's and buy some Big Macs or saunter on down to the Latte shop and pick me up a 16 oz non-fat Latte while I stay at my desk and keep working. Maybe hop out of the car and get one while I drive around because I can't find a parking spot. Donuts would be good too.

Now, that is a real robot. The only problem with the bot in this picture is that it needs a place to put the drink. Scones can be pretty good, but they tend to be kind of dry for my tastes without a drink to go with them.

Duane Benson
Watson - come here! Find Juan Valdez and ask him where I can get some coffee.
Five

Nightmare on BGA Street

I seem to be in a bit of a BGA mood lately. I do that sometimes - pick a subject and talk it to death before moving on. Well, maybe not quite talk it to death, but at least talk it to the pain.

9x13 via in pad BGA land Take a look at this land pattern for a bluetooth module. Anyone see anything odd? Yeah. All of those really big open vias. I know what the designer was trying to do. A good number of the vias are ground connections of one sort or another that need to be connected to an internal ground plane layer.

Given that is is a 1mm pitch BGA, there is plenty of room to put the vias between the pads and not cause any trouble. That would be one recommended approach. The other would be to have the vias filled and plated over at the board house. No matter what you do, though, the vias can't be left wide open like this. It's a real bummer.

Duane Benson
B.V.O.U.S.'s? BGA Vias Of Unusual Size. I don't think they exist.

BGA Woes

Quite a few of the new chips I see coming out stick to the BGA or QFN form-factor. Sometimes they'll be referred to as WSP (wafer scale package) or CSP (chip scale package), but those are still just little BGAs. Some do show up in larger packages, but many of the really new designs seem to stick to these form-factors.

A few years back, we tended to see a lot of design problems related to regular, big BGAs (0.8mm or greater pitch). Things like black padmicrovoids and via in pad cropped up to cause proto-headaches. While those problems still show up from time to time, they have become much less frequent. No, we're seeing issues with the tiny ones - 0.5mm and 0.4mm BGAs, CSPs and WSPs.

With a big BGA, you can route to vias in between the pads. That's easy. With the small ones, especially 0.4mm, you can't. You have to put the vias in the pads. Of course, you have to fill and plate over the vias. Big BGAs tend to prefer non-soldermask defined pads (NSMD) while some of the 0.4mm BGAs require soldermask-defined (SMD) pads. A really flat surface is more important for the tiny parts too. Don't fear extra small parts, but you may need to do a bit more homework and relearn a few old rules-of-thumb.

Duane Benson
I'm solderin, I'm solderin, I'm solderin for you

« January 2011 | Main | March 2011 »