Screaming Circuits: March 2007


Embedded Systems Conference

The conference starts next week and I'm pretty pre-occupied with all the show details. Part of our booth is missing in the tradeshow warehouse but everything else seems to be ready.

Even if the booth doesn't get found, we'll still be there. I'll be dissabointed, but quite frankly, the booth isn't all that special. What more important is the message. What's more important is just letting engineers that we are here to help with difficult problems. All the booth is supposed to is is grab you attention. Not having a booth sturcture, might even be just as effective. Who knows?

I'll also be spending some time looking for new and exciting things for those of you that can't make it to the show. Last year, we saw a lot of robotics and a lot of chips in advanced packages such as .5mm BGAs and little QFNs. Those packages are becomming a lot more common so it's a good thing we assemble a lot of parts in those form-factors.

Check back during the week and see if I've found anything really interesting.

Duane Benson

ISO 9000

Until recently, Screaming Circuits had been piggybacking on the ISO certification of parent company Logo90012000gif_2MEC   Northwest. Now, we are certified on our own as Screaming Circuits. This is just one of a number of initiatives we have underway to improve our quality. Our goal is simply to build your boards exactly the way you want them, and get them to you when you want them.

Duane Benson

China RoHS

China_rohs_e_logo We've all talked Euro-RoHS to death over the last year. Well, there's still a lot to learn but a lot has been said. Now we have the China RoHS to figure out and unfortunately it is different enough from the European version that complying with one won't necessarily get you into compliance with the other. What's a BGA to do?

Probably the biggest difference is that, starting on June 1, 2006, the European Union prohibited the banned substances, while beginning March 1, 2007, China just requires labeling of the covered substances. I ran across this FAQ article that gives a pretty good overview. Here's another one from the AeA.

Duane Benson

ESC Silicon Valley

Just a reminder - we will be down in San Jose, California on March 3, 4 and 5 for the Embedded Systems Conference. As a bonus, we'll have our friends from Sunstone Circuits in our booth part of the time demonstrating their PCB123 PCB CAD software.

Panic_time_b_rev_2600Stop by and see us in booth 3037. Take home a breathtaking 11x17 "Panic Time" poster and if you are one of the lucky ones, a pound of Screaming Circuits blend coffee from the Longbottom coffee company.

Use this link to get a free admisions pass to the exhibit floor.

Duane Benson

Mixing Leaded and lead-free BGAs

gCan you mix lead and lead-free when using BGAs? Well, no, and yes, sometimes or maybe. It would be nice if there were an easy answer, but there really isn't.

At first glance, it's pretty safe to just say no, you can't mix. The processing has to go with the BGA. Lead-free BGA means the whole board has to me lead free. Leaded BGA means the whole board needs to be leaded. That's the safe way to go, but it isn't always possible.

You can take a lead-free BGA and put it on a board with leaded solder paste and lead-free temperatures (below 217C). You will be putting the board at risk for cracks at some of the solder balls due to the stresses caused by uneven heating and cooling.

If the rest of your componentry can handle the extra heat of a lead-free reflow profile and you need to put a lead-free BGA in a leaded build, you can use either a lead-free paste or leaded paste and run it through a lead-free profile. Industry studies have shown (1, 2) that if the two alloys are allowed to fully mix, the BGA will reliably adhere to the board, at least for prototype purposes.

Going the other way (leaded BGA on a lead-free board) may not be possible. The leaded balls might over heat, over sag and end up cracking. In this case, since the board would be out of compliance anyway, it would probably be better to just run the whole board as a leaded project.

(1) Fay Hual, etal, Intel Corporation, "Solder Joint Reliability Assessment of Sn-Ag-Cu BGA Components Attached with Eutectic Pb-Sn Solder"
(2) P. Snugovsky, etal, Celestica, "Theory and Practice of Lead-Free BGA Assembly Using Sn-Pb Solder"

Duane Benson

Parts Substitution

Schottky_bottom Take a look at the two parts in these pictures and see if you can guess what the difference is. It's the same two parts in both photos, just a top and a bottom view. Click on the images for a close-up view.

I'll give you some hints. They are both Schottky diodes. The photos aren't quite clear Schottky_topenough to read the part numbers so I'll give you that one. They're both from International Rectifier. You can kind of see the "IR" on one of them in the photo, but not the other. The foot print looks pretty much identical - both SMA packages.

Enough with the suspense. I'll answer the question from three different perspectives.

One is an IR 10MQ100N, 100V, 2.1A with a Vf of 0.68volts. The other is an IR15MQ040N, 40V, 3A with Vf of 0.43V. From the perspective of Screaming Circuits process technicians, these are two completely different parts. If the BOM called for one and we saw the other, we would have to stop the build and try and call you. Doing so might very well delay your job, but we would have to do it.

So, is there really a difference? Are we right or wrong to stop the build?

Design A is a triple voltage (12, 5 & 3) DC to DC converter. It takes 15 to 30 volts input and converts it to the three outputs at a combined maximum current of 1.0 Amps. The diode, in this case, is used for reverse voltage protection at the input. Either diode will work fine in this application. The 40V, 3A would provide a little more current head room and cause less of a voltage drop for a bit more margin on the low-voltage end. The other one would give more voltage overage head room. For the purposes of a prototype, though, they are effectively equal in this application.

Design B is a free-wheeling diode application in an H-bridge motor driver. The maximum free wheeling current is expected to be under an amp at approximately 48 volts. Elsewhere, the circuit is spike protected with 72 Volt TVS diodes. In this case, the 100V part would fit the requirements and would give adequate headroom. The 40V part might very well smoke at first turn on.

For designer A, the parts are completely interchangeable in a prototype. For designer B, putting the wrong one in could smoke the part, and potentially also damage some mosfets and an expensive mosfet driver chip. We can't tell which is the case, so we have to assume the worst case and stop the build. It's the same with close but not exact value caps and resistors. We have to assume the worst and stop the build.

If you do need to sub, change the BOM before you zip it up and send it to us or at least tell us in the special instructions section of the online order form. If you make the substitution after you've placed the order and sent us the files, give our customer service folks a phone call (866)784-5887. We don't want to delay your project, but we REALLY don't want to smoke it.

Duane Benson
Where there's smoke, there's bile

Thru-hole counts too

Most of our RoHS discussions focus around surface mount parts: BGAs, GBGAs, QFNs, etc. Such components tend to be the most sensitive to process issues. It is important, though, to not forget about the lowly thru-hole part.

Thru-hole component leads are traditionally plated with solder. Some use other materials like gold or tin, but if yours are solder plated and you need RoHS, make sure you order the lead-free varieints.

Duane Benson
Pb and Jelly

Pleasant surprises

Every morning before coming to work, I make myself a Latte with my little Krupp® espresso machine. I don't make great Lattes but they are pretty good and have lots of caffeine. Often, after being in the office for an hour or two, I'll run to the local Starbucks® or the espresso shop in a Thriftway® grocery store nearby, for a second dose of caffeine.

I'm not an expert latte maker, but I have been making them for quite a while and I am an avid drinker of lattes. When I buy a commercial latte, I consider my own Lattes and I think about how close the shop I'm at comes to the taste and experience of my own product. When I make my own, I consider the same - I think about what the consumer (me, in this case) wants and I do my best to deliver that to myself.

This morning, as I walked in to the office with my home-made latte in hand, I discovered a Starbucks latte on my desk; a gift from a generous co-worker. When I opened up my email client, I also found a couple of comments from blog readers. I consider that to be a gift from generous customers (or potential customers).

We've been growing pretty fast here and over the last couple of months have been working real hard at keeping our quality and service up to your standards. We have company meetings, Lean events, put out mission statements and do all that rather glurgey stuff companies do.

I know that stuff is really important and it does produce results, but what makes me really happy to come into work at this place every day is one little questions that I regularly hear around here: "what would a customer think about that?"

In some companies, the glurge is more about be about making the company feel better about itself than it is about delivering a better product/service to the customer. From the inside, it's frequently difficult to say which is the case, but, that one question: "what would a customer think?" tells me that this group of people at Screaming Circuits really understands that the only reason we are here is because we have given customers good value. The only reason we will continue to be here is if we keep giving customers (you folks) good value and keep increasing the value we give you.

Duane Benson
"What would a customer think about that?" Think about it.

Loose parts II

Loose_parts_bent_pin A while back, I wrote a post about loose parts (read it here). Basically, we can usually/sometimes deal with them for an extra cost. This photo here does a pretty good job of illustrating the potential problem.

Of course, in addition to bent pins, loose parts also present static damage, particle contamination (dirt) and moisture absorbtion risks.

Duane Benson

Lead-free BGA shock

The electronic manufacturing industry as a whole has gotten much more comfortable with lead-free processing over the last six months. We've been offering RoHS processing for a year and a half now. Still, there are some challenges in areas such as longevity and shock performance.

We tend to deal primarily with prototypes so some of the challenges are less of an issue for our customers at our stage in the NPI process then for later stages. One such issue is that of lead-free BGA shock resistance with SAC (Tin, Silver, Copper) solder. One of the solutions being explored by the industry is an adhesive underfill.

I ran across this article by Jim Hisert of Indium Corporation that covers two different methodologies. Again, we don't perform underfill here, but when you get to production, if you have a high-stress application with lead-free BGAs, you might want to consider it as an option with your production assembly house.

Duane Benson

« February 2007 | Main | April 2007 »