It’s a good thing I wore my most comfortable shoes during my recent visit to HP’s test lab in Ft. Collins, Colorado. If I hadn’t, my feet and legs would still be sore today. I joined a small group of reporters at the lab and spent 3 hours walking miles through this huge facility stopping at each specialized test lab for very in-depth guided tours.
We toured the Hardware Test Center, which I call the ‘Shake & Bake lab,’ where workstations and other hardware are physically stressed to the max; The Electromagnetic Compatibility (EMC) labs where hardware is bombarded with radio frequency waves to gauge how susceptible the units are to the effects of RFI; the Model Shop where the future is imagined; the RF testing area where hardware returned from the field is investigated to identify the cause of the failure; to a demonstration room where we learned about HP’s latest and greatest new hot stuff; and to the Material Science Lab, which is the focus of this installment.
I’ve elected to break my experience into installments that focus and expand on each of the departments I visited so I can ‘bring’ you with me on the three-hour tour.
The “CSI” Lab
Material Science Lab
Ft Collins, Colorado
Paul’s a Happy Man
CSI, the acronym for ‘Crime Scene Investigation,’ is the perfect name for Dr. Paul Mazurkiewicz’s well equipped Material Science Lab here in Ft. Collins. When Jim Schinnerer, in the ‘Shake and Bake’ lab next door, drops a workstation from ten feet high onto a concrete slab, or exposes one to high humidity or extreme and prolonged heat it’s up to the good Dr. to reconstruct and investigate any failures caused by this rough and tumble treatment. And, when equipment fails in the field, Paul uses his investigative skills and advanced tools to find the cause and make recommendations so this failure won’t occur, ever again.
Paul is a very happy man. Smiling, with a zealous, almost religious glint in his eyes he proudly showed our small group of reporters around his high-tech sanctuary. And, once I got a glimpse of the lab and its bevy of high-tech tools I understood why Paul was smiling. Neat technology is embedded in my genetic coding, and I was smiling too.
Dr. Mazurkiewicz welcomed us as we filtered into his well equipped lab. “I’m Paul, chief scientist here at the Material Sciences Lab and my job is to help the engineers figure out complex problems they encounter during development, and I also help the customer support engineers solve problems out in the field. The best way for me to show you what we do here is to take you through the materials lab and show you some of our tools and how we use them.”
The Real Time X-Ray: Seeing Through Metal
Paul first showed us how he uses the real-time x-ray machine to perform failure analysis. “The real-time x-ray machine was specifically made for the electronics industry and includes artificial intelligence routines that help us make decisions about the quality of such things such as solder joints and electrical connections. “
He kicked off the demo by passing around a 3” diameter analog pocket stop-watch. He then placed the watch inside a small shielded safe-like box closed the heavy door and turned on the real-time x-ray machine. “Opening the pocket watch doesn’t look like it’ll be easy,” he said, “and there’s a chance we might break it while opening it. So, it’s always handy to do non-destructive analysis where you can take a look at something without ripping into it.” Pointing to the two LCD screens he continued, “This is a real time video feed, and we can easily scan over an area and are able to zoom in and see the inner workings of the watch. You can see little gears working here, and if one of these cogs was out of place or a gear tooth was broken off we ‘d easily and quickly see that. “
Paul fiddled with some knobs and zoomed in on the image so that only the spring mechanism was centered on the screen. Then, he moved in even closer and pointed out that, “Most failures in spring wind watches occur in the spring itself,” and as he zoomed further in we could see a small, ‘accident waiting to happen,’ hairline fracture at the base of the spring.
Paul explained that “The idea is to solve a problem very quickly, but very accurately at the same time. If somebody’s having problems it doesn’t make a difference whether it’s in development or out in the field, no one wants wait and we need to solve this person’s problem fast. “
Next he put a hard drive into the x-ray machine and explained that this was his personal drive and that it had started making a squealing noise. He showed us how he could zoom into the tiny microprocessor and its even smaller controller chip. “We can see right through the metal top and plastic casing – I can go right through and look at a whole bunch of features on this chip. As an example, I’m going to zoom in a little closer and then adjust the histogram to get a nice gray scale range”
He pointed to the screen, “You’re looking at all the solder connections. These solder points are basically what adheres the chip to the printed circuit assembly. We’re actually looking right through the package at copper traces that go to little gold wires, and if we pan over a bit we’re now looking at the shadow of the microchip. You can’t actually see silicon in an x-ray machine, it’s not dense enough, but we can see the silver epoxy that’s used to glue the chip down to the surface. We’re able to really zoom in, and since the x-ray doesn’t use light we can magnify up to thousands of times very quickly. “
“Here’s something we can investigate,” he continued, “These little soldered connections beneath the chip carry heat away from the chip – it’s basically a heat sink on the other side of the chip -- and you’ll notice that these have big holes in them right here. These holes are air voids and these voids could result in less heat being transferred away from the chip than was intended. So, one of the reasons that this drive may have failed is that there may not be enough heat being transferred out from this particular area. And, we were able to drill down and find the problem without having to take anything apart and see what’s going inside the component. You just throw it in the x-ray chamber and away we go.”
He then adjusted the focus to look at the drive’s platter. “This platter spins really fast as you write data to it. Pointing to an armature he said, “This arm floats across the surface of the platter and writes and retrieves data from the spinning disk. Normally these arms park off the drive platter so that when you transport the drive the arm doesn’t bounce on the drive’s surface and cause damage. But, as you can see, this one is awfully close the spindle here.
He zoomed in to show the tiny pickup heads resting on the disk platter.. “Each one of these heads is a tiny copper wire-wrapped electromagnet which pulses the electromagnetic energy. This one is awfully close to the spindle, and there a possibility that a bearing wore out or might have vibrated and knocked that head over to the side and it’s stuck there. The screeching noise I heard may have been this head rubbing while the device spun down from 15,000 rpm to 0 in a second. This would obviously make some noise. This configuration might have been by design, and normally we’d have a schematic in front of us that tells how it should be so we can interpret what we’re looking at. Using the real-time x-ray is the first step in the analysis and allows us to learn as much as we can, quickly and non-destructively to see if we can get to the heart of the problem.”