Screaming Circuits: July 2007

Spotlight: The Centroid File

We recieve a lot of questions about the Centroid file that we require. The file just tells our machine where each part goes on the board and what direction to point the part when putting it on the board. It's not the only file we need, but it is certainly one of the most important.

There are two ways that we can end up with a good Centroid file; a) you can provide it, or b) we can generate it from the other CAD and BOM files that you provide. If you want us to make it, we'll charge $50.00 to $100.00 for a typical board, sometimes more for really complex boards.

Most CAD systems will automatically output the file for you. It might also be called an "Insertion file", a "pick and place file" or an "X-Y file."

If you use:

  • PCB123: If you use PCB123 and tell them that you are sending the boards to Screaming Circuits, they will send us the Gerbers and Centroid without an extra charge.
  • Eagle: Download the User Language Program (ULP), described here on our web site, and run it on your board.

If you want to create it yourself, of just understand it better, download our PDF describing it and our sample Centroid file from our website.

Duane Benson
A = sd1
V = Ad2

CAD Library or Substitution Issues

We recently ran across an issue out on the shop floor that was most likely due to the components library inQfn_on_qfp_land  the CAD package. A QFN part was sent for this assembly but the land pattern on the board was just a bit too big. The QFN pads would not reach the pcb solder pads.

Most likely, the designer accidentally selected the library part for a QFP variant of the chip instead of for a QFN. The parts are pretty similar in size if you exclude the leads on the QFP package. It's also possible that someone accidentally purchased the QFN variant when a QFP was called for.

Just one more thing to add to your design-check.

Duane Benson
It's just over that next hill...

Sponsored Spotlight - OpNeAR

Our sponsorship program recently added the OpNeAR group from The Erik Jonsson School of Engineering & Computer Science at the University of Texas at Dallas (that's a mouthful of a name). If you want it to be even longer, go ahead and spell out The Open Networking Advanced Research Lab at The... Just go to their web site.

These guys are focused on advancing wireless and optical networking technology and the group we are sponsoring just won a design contest put on by Texas Instruments. Take a look at this page for specific details on the project.

I love the tag line for the engineering school: "FEARLESS engineering" That phrase is their trademark so don't go trying to use it yourself. They beat you to it. It makes a simple, yet dramatic and powerful statement. You can't do much better than that.

Opnear_board_500Speaking of doing better... The guys won the design contest and the system they built works so well that additional students are coming in just to use it. Click on the thumbnail for a larger view. If you look closely at the board, you'll note that our board-fab partner Sunstone (AKA PCBexpress) also helped on this project.

Congratulations OpNeAR! and keep up the good work.

Screaming Circuits

BATTLE OF THE BLOGGERS - Screaming Circuits Style

Join our first Guest-Blogger challenge! That's right, try for fabulous (fabulous by my definition, anyway) prizes and maybe get your story published on the Screaming Circuits blog.

PCB-Layout engineer 100 We try to be helpful and informative here in our blog, but it's always been missing something: you!

First place will receive a $100.00 gift certificate and become Screaming Circuits very first Guest-Blogger! The runner-up will receive a $25.00 gift certificate. Honorable mentions may or may not by published on our blog. See details below.

We want this to be fun and informative for everyone, so please submit a story with any one of the following themes:

  • Cool and unique applications: We're looking for the biggest, smallest, fastest, newest, farthest...
  • Design disasters / visits from Murphy: Yes, it happens to all of us.
  • Tech notes: Easy to read and covering common problems with difficult solutions.

The challenge begins now and will end when we find the person and story we've been looking for! We are flexible in choosing who the lucky Guest-Blogger will be, because we may (or may not) be willing to award more than one first place winner! Judging will be based solely on our opinions and our choices are final. If you disagree, we'd certainty like to hear from you, but we are the final word.


Submit an entry by:
Email your entry to, subject "Blog Challenge".

Terms: Entries should be about 350 to 1,000 words. You can add in an illustration or photo as long as you have the rights to them. Entries must be your own material. The contest ends when we choose a winner. If we find more than one winner, we may award as many or as few first and second place prizes as we feel necessary. However, if we don't find a winner, we may end the challenge without awarding any prizes. Sorry but employees of MEC and Screaming Circuits are not eligible this time. We'll have our own contest soon though.

Speaking of "Not Happy" PCBs...

Pic_uc_bd_blank Speaking of boards with issues, this one (click the image to enlarge) used in my prior post has a few more problems than just holes that are too small. Other than the holes, it is serviceable, but it really isn't a great PCB. It's a immersion silver board from an off-shore manufacturer. This house is not a place we would ever send our customers.

One of the most obvious issues is the registration of the silk screen. For example, Look at the nomenclature for "C6". Half of the "C" is right on top of the pad for pin 18 Yucky_brd_c6 of IC2. You could scrape it off, maybe or you could hope that the enough solder would flow under to make a good connection, but it isn't a good situation. In the CAD package, the label was not on the pad but the board house used a slightly larger font size and the registration was off a bit. One more thing for your design check - make sure that you have plenty of clearance in case the board house does what this one did to me.

Yucky_brd_i2c Sometimes, it's the library again. Looking at X6, the I2C connector, one of the silk screen lines trims the top of all four solder pads. The library part in the cad package does that so I should have caught it.

The last comment I have about this board concerns the drill holes. Look at the ridge around the holes here on. All of the drilled holes had that ridge to a certain degree. It's not catastrophic in a prototype environment, but it speaks to a few processing steps left out at the board house

Duane Benson

Drill Sizes in your CAD package

We recently ran across a sticky situation related to the drill size for a thru-hole part. Most of what we do is machine SMT work, but most boards have one or two thru-hole parts on them as well. Occasionally, we get boards with a lot or even all thru-hole.

Just the other day, we got one in that, in total for the job, had several hundred thru-hole resistors. This wouldn't be an issue at all except that the hole size was just a little bit too small for the leads. In almost all cases, there pretty much isn't anything we can do in a case like that other than send the boards back to the customer to be remade. These were customer supplied, so as much as we hate to do it, it was all on their dime and we still had to charge for the labor that we had already put into the project.

Too_sm_holes_2This customer was using Protel. I had the same thing happen to me some time ago with a board I laid out in Eagle. It was a PIC micro board designed to run a couple of motor driver boards. After getting the boards fabbed, I started to assemble them and discovered that none of my Molex keyed connector blocks would fit. The drilled holes were too small.

At first, I was upset with the board house, but when I went back into Eagle and looked a little closer, something didn't look right. I thought that I had used the correct parts library - and upon double-checking found that I had - but the holes sizes used on the package were too small.

Looking at my board (click image to enlarge), "A" shows the spot for the three-pin Molex connector. Comparing the hole size forPic_uc_bd_w_callouts  that, look closely and you can see in my call-outs that the holes are the same size as the holes for IC1 (call out C) and are slightly smaller than the holes for my breakaway connectors (call outs B and D).

The drill size used for the Molex connectors in the library package in Eagle  was .032 while the datasheet specifies a .040 hole size. That's not gonna work. The breakaway connectors call for a .045 inch hole size while the library package used a hole size of .036. Oddly, because of drill size variation, I could wiggle the breakaway connectors in.

In both cases, the real solution was to modify the parts libraries to increase the hole size. These may have been user created libraries or they may have came with the original install. I don't know for sure, but with so many new components coming out, it's quite likely that you'll have to use custom libraries no matter what CAD package you are using.

What should you do?

It's always a good idea to look at things like footprint, SMT pad size and thru-hole hole size in your design check - especially if you are using third-party libraries. I didn't really pay much attention to that while originally laying it out, but once I went back in to check it out, it was obvious at a glance. An extra few minutes spent looking at those things would have saved me (and this recent customer) a lot of time and money

Duane Benson


Obviously, SMT = "surface mount technology", TH = "through hole" and COB  = "chip on board." But, what about COG? It could be "cost of goods" but I think I like "candy on graham" best.

Run over to this article on about an interesting and new technology for the assembly 768600012_428c3e19eb of electronic circuits.

I would like to note that when I checked out the color coding on the resistors, I interpreted them as orange, yellow, black and gold for tolerance. It's really brown, black, yellow which actually does make sense. (The image came from Evil Mad Scientist via flickr under the Creative Commons license)

Duane Benson
Mmmmm... Chocolate...

Trade show Planning

In September (yes, that seems like a long time from now, but the summer will be over before you know it), we're doing to two tradeshows.

First, on the 12th, we going to a local show, called OctoberBest. For those of you in the Portland area, stop by. We'll be in a little booth but we're also speaking during one of the technical sessions on using advanced packaging.

Then, if you happen to be in the Boston area, we'll be at the Embedded Systems Conference on the 19thEsc_boston_sm  and 20th in booth 1224. We'll be featuring some cool projects from organizations that we're sponsoring.

We'll post more details on each show as the date closes in.

Duane Benson
Planning ... good

A bit more on Thermal Mass

Thermal_mass My post from a few days ago discussed a problem that can potentially make some RoHS parts not work in a RoHS application. I forgot to mention, that since this is a prototype, we solved the problem by hand soldering the parts. You could never afford to have that done in a volume production environment, but for a prototype, it worked. Still, the best solution is to change to a more robust part or re-layout the board to give more spacing between components with above average thermal mass.

Duane Benson

We're Closed on the 4th

Just a reminder - We're closed today for the Independence day holiday. Today is not counted as a "business day" when we calculate your ship date.

Happy 220th birthday, US Constitution!

Constitution_centenial_2 (The Constitution birthday is not exactly today, but on September 17th of this year)

« June 2007 | Main | August 2007 »