The home lab of Darryl Phillips
My buddy Darryl Phillips has a home lab that is a labyrinth of steel, alarms, and cameras. After the success of my recent posts about Barrie Gilbert’s home lab, I wanted to get Darryl’s lab in the blog. I met Darryl when he was a power apps engineer at National Semiconductor. He now does consulting on analog and power electronics design and testing. He writes:
- I’m the engineering equivalent to the sadistic doctor in the war movies and my job is to mercilessly torture analog ICs to find all the bugs before the customers do. I have become an expert in functional testing of analog ICs. It’s like being a poultry inspector; it’s an obscure but essential role that nobody cares about until things go terribly wrong. My view is that poor functional quality is a major hidden factor in weak analog product sales. It’s quite insidious. Often when customers have a bad experience they discreetly take their business to the competition rather than complaining and waiting for a fix. The result is that business weakens mysteriously. A certain analog company burned Apple with a product defect. After visiting to introduce a new-product, one of the FAEs (field application engineers) noted; “The Apple engineer was very polite but as I turned to leave, he picked-up the eval board like a dead mouse and swiveled to drop it in the trash.”
- This industrial-strength home-lab is built for testing analog ICs across the military temperature range and for strategic-customer design-win support. It features an environmental chamber straddled by independent desk and rack-mounted test-benches, backed by CAD workstations and a CNC milling machine. Coming into electrical engineering from a construction background, I’m a firm believer in the importance of owning the tools of one’s trade. It’s more than a matter of pride; it’s a basic tactic for success and survival. The engineer who owns his own tools is better positioned to sustain a career through both opportunity and adversity.
- The heart of the bench setup is a cluster of meters and an oscilloscope. This lab is heavily influenced by Cal Poly. Rather than multi-tasking a single meter, a whole set of meters is used for simultaneous Kelvin-style measurements of essential circuit parameters, such as input and output voltages and currents for a dc-dc converter. When testing new-products, it’s very important to have instruments with all the important parameters in a single field of view. This arrangement ensures that as one is observing one meter, bugs flashing across the scope or other meters will be observable and not be missed.
- The environmental chamber is a Thermotron S1.2C. This is a nice, reliable oven for general use. I like the version with the older 2800 controller. These are available used for as little as $3,500 but research the purchase carefully; There is a lot of crusty junk marketed as S1.2Cs, and some old versions use obsolete refrigerants.
- The rack setup is for automated test. I’m a practitioner of the philosophy; Have people do what people do best and machines do what machines do best. Automated test is good for repetitious collection of bulk measurement data. But, it’s no replacement for human effort, experience and judgment when searching for and evaluating analog IC bugs.
- On the opposite side of the lab sits an LPKF PCB (printed circuit board) milling machine. I use Pulsonix for PCB CAD (computer aided design) and Spice simulation. Pulsonix is very productive and has great LPKF support. In a crunch, I can design, mill, assemble and test a custom analog IC evaluation board in as little as a day and a half. However, chip-scale packages require service from a professional board vendor.
- One of the unusual details is that Macs run this lab; I like computers and software tools that simply work, like a hammer.
So there you have it, another great home lab. Darryl once gave me a riddle about what color the coax cables of a good engineer should be. He was trying to demonstrate that hi-temp coax is brown, not black, and if you were going to run the coax into your test oven, you should use the brown stuff. I forget the designation; I will ask him to post it in the comments.
Mark F. commented:
I worked with Darryl at Maxim in the late 90's. He taught me a lot about DC/DC converters. He had an impressive lab back then. But nothing like it is now. I can confirm that Darryl is the best at torturing ICs to find their weaknesses.
RosinSmoked commented:
Oh so envious! Smart move on buying test equipment; maybe not so much on RE.
It's amazing what you can find on eBay these days!
Comment on your rack of meters. "it’s very important to have instruments with all the important parameters in a single field of view. This arrangement ensures that as one is observing one meter, bugs flashing across the scope or other meters will be observable and not be missed." NASA used to take that approach, and their ground controllers were overwhelmed. I don't know how far they got with it, but the idea they were moving towards was to digitally monitor many, many parameters, and present out-of-tolerance parameters to the human for decision-making. With the proliferation of PC-based instrumentation, I would think that would be the direction to take if you had it to do over. Often lets you move the signal conditioning closer to the UUT too.
- Jim
Paul P. commented:
That’s a beautiful setup. As a power device and prototype lover myself, I have a great appreciation for all the power equipment and the especially the PCB mill! I’m actually in the process of building my own CNC mill. But alas, like so many of my work-in-progress projects, it’s got a nice layer of dust on it out in the garage. But my “lab” in the house is nice and clean as it’s used frequently. Of course, my “lab” is but a fraction of your setup. You’ve got a very useful, albeit substantial, work place there!
Jennylee commented:
Your asnwer was just what I needed. It's made my day!
The home lab of Darryl Phillips commented:
A homelab with a clean desk is not reality to me,
normally something is going on in the lab, at work or at home there is just a very little room what is free on the desk...
Jac Smeets PE1KXH
Darryl Phillips commented:
Great write-up, Paul! The brown BNC cables are typically RG316 or RG142. The distinction is that these are made of FEP (Teflon). In contrast, the black BNC cables are typically RG58, made of polyethylene, and if they are used in environmental chambers at high temperatures (125C) the dielectric will melt and ground braid will short to the center conductor. One distinction between the companies that know what they are doing and those that don't is the existence of such cables in the lab, which implies that new-products are being evaluated over their rated temperature range prior to production.
In the roaring 90s, I worked for a company that out-grew its resources, resulting in a corporate lab that was like "Lord of the Flies." As I cashed out stock options, I put 1/3 in equipment and 2/3 in real estate. The resulting lab is a result of many trips to the Foothill Flea Market, Halted, Weird Stuff, Test Labs, Spectra, Metric, etc. But its nothing compared to Larry's.
Jimelectr commented:
Wow, that's a lot of stuff, but then again, Darryl is running his own consulting biz. Not sure how he afforded all that stuff. After seeing Barrie Gilbert's and Darryl's home labs, I'm kinda depressed about my own measly setup (a couple scopes, function generator, logic analyzer, FFT spectrum analyzer, and some power supplies).
I had to laugh about the high temp cables being brown; we have some RS-232 serial port cables around our lab at work that have turned brown and crispy from being used in the temp chambers! The lab manager is in the process of having higher temp teflon-wrapped extenders being built. We also had some digital audio plastic fiber cables melt on us and switched to glass fiber.
selinz commented:
That's some pretty expensive stuff for a home lab. Borrowed from work?
Bob G. commented:
Waaaay too neat! Darryl, must be from another planet. Or, maybe his brother Darryl is the sloppy one.
