My father (a carpenter) always said, “A bad workman blames his tools.” That came to mind during a panel at DesignCon 2012 last week when a panel of experts I was moderating pointed out that the reason you may be having difficulty with 25Gbit/s signals on highly integrated, mixed-signal ICs and boards is that you do not fully understand what your scope can do.
“Say what!? How dare they!” I hear you exclaim. But hang on a second, they may have a point.
This year’s conference, while still focused on signal integrity in general, clearly emphasized crosstalk, which becomes increasingly problematic with 4 x 25Gbit/s SerDes and 28Gbit/s backplanes. Add in tighter chip or board floor plans, with exponentially increasing integration of wired and wireless communication with increasingly proximal power planes, and you have a recipe for cross-coupling cacophony.
The panelists, Greg Peters from Agilent, Kevin Ilcisin from Tektronix, David Graef of LeCroy, and Eric Starkloff of National Instruments, are four of the brightest minds in the industry, and when the question came to what they were doing about helping engineers keep their eye diagrams open, they professed to be doing lots. And its reflected in the equipment (scopes) they’re putting out there.
However, that equipment can do so much, and is so advanced, they agreed, that in many cases it’s a case of educating the users on the full scope of what they can do with the latest systems. The devices’ capabilities have outpaced your ability to keep up.
So, my friends, is this a case of us needing a new interface for scopes? Do we need the next Steve Jobs to stand forth for scopes?
Maybe, but, as one test engineer put it to me last week, we kind of like our scope interfaces. It’s like a car. You can go from one to the next and always know what knobs to press and turn to go where you want.
Do you know how to drive your scope on 21st century data highways? Do we need to rethink scope interfaces? Or do you need to update your license to drive?